Magnetic stratigraphy

MAGNETIC STRATIGRAPHY NEIL D. OPDYKE AND

JAMES E. T. CHANNELL Department of Geology University of Florida Gainesville, Florida

This is Volume 64 in the INTERNATIONAL GEOPHYSICS SERIES A series of monographs and textbooks Edited by RENATA DMOWSKA and JAMES R. HOLTON A complete list of books in this series appears at the end of this volume.

w ACADEMIC PRESS San Diego London Boston New York Sydney Tokyo Toronto

Contents

Preface

Table Key

1 Introduction and History 1.1 Introduction 1.2 Early Developments 1.3 Evidence for Field Reversal

2 The Earth's Magnetic Field 2.1 Introduction 2.2 The Dipole Hypothesis 2.3 Models of Field Reversal 2.4 Polarity Transition Records and VGP Paths 2.5 Statistical Structure of the Geomagnetic Polarity Pattern vii

7

3 Magnetization Processes and Magnetic Properties of Sediments 3.1 3.2 3.3 3.4 3.5

Basic Principle Magnetic Minerals Magnetization Processes Magnetic Properties of Marine Sediments Magnetic Properties of Terrestrial Sediments

4 Laboratory Techniques 4.1 4.2 4.3 4.4

Introduction

7.1 Oceanic Magnetic Anomaly Record 7.2 Numerical Age Control

8 Paleogene and Miocene Marine Magnetic Stratigraphy 8.1 Miocene Magnetic Stratigraphy 8.2 Paleogene Magnetic Stratigraphy 8.3 Integration of Chemostratigraphy and Magnetic Stratigraphy

Resolving Magnetization Components

9

Statistics

Cenozoic Terrestrial Magnetic Stratigraphy

Practical Guide to the Identification of Magnetic Minerals

5 Fundamentccls of Magnetic Stratigraphy 5.1 5.2 5.3 5.4 5.5 5.6 5.7

Late Cretaceous-Cenozoic GPTS

Principles and Definitions

9.1 9.2 9.3 9.4 9.5 9.6

Introduction North American Neogene and Quaternary Eurasian Neogene African and South American Neogene North American and Eurasian Paleogene Mammal Dispersal in the Northern Hemisphere

Polarity Zone and Polarity Chron Nomenclature Field Tests for Timing of Remanence Acquisition Field Sampling for Magnetic Polarity Stratigraphy Presentation of Magnetostratigraphic Data Correlation of Polarity Zones to the GPTS Quality Criteria for Magnetostratigraphic Data

10 Jurassic-Early Cretaceous GPTS 10.1 10.2 10.3 10.4

Oceanic Magnetic Anomaly Record Numerical Age Control Oxfordian-Aptian Time Scales Hettangian-Oxfordian Time Scales

6 The Pliocene-Pleistocene Polarity Record 6.1 6.2 6.3 6.4 6.5

Early Development of the Plio-Pleistocene GPTS

Jurassic and Cretaceous Magnetic Stratigraphy

Subchrons within the Mafuyama Chron

11.1 Cretaceous Magnetic Stratigraphy 11.2 Jurassic Magnetic Stratigraphy 1 1.3 Correlation of Late Jurassic-Cretaceous Stage Boundaries to the GPTS

Magnetic Stratigraphy in Plio-Pleistocene Sediments Astrochronologic Calibration of the Plio-Pleistocene GPTS wAr/39ArAge Calibration of the Plio-Pleistocene GPTS

182 196 202

Contents

X

12 Triassic and Paleozoic Magnetic Stratigraphy 12.1 12.2 12.3 12.4 12.5 12.6

Introduction Triassic Permian Carboniferous Pre-Carboniferous Polarity Bias in the Phanerozoic

Preface

13 Secular Variation and Brunhes Chron Excursions 13.1 Introduction 13.2 Sediment Records of Secular Variation 13.3 Geomagnetic Excursions in the Brunhes Chron

14 Rock Magnetic Stratigraphy and Paleointensities * 14.1 Introduction 14.2 Magnetic Parameters Sensitive to Concentration, Grain Size, and Mineralogy 14.3 Rock Magnetic Stratigraphy in Marine Sediments 14.4 Rock Magnetic Stratigraphy in Loess Deposits 14.5 Rock Magnetic Stratigraphy in Lake Sediments 14.6 Future Prospects for Rock Magnetic Stratigraphy 14.7 Paleointensity Determinations

Bibliography Index

250 25 1

Magnetic polarity stratigraphy, the stratigraphic record of polarity reversals in rocks and sediments, is now thoroughly integrated into biostratigraphy and chemostratigraphy. For Late Mesozoic to Quaternary times, the geomagnetic polarity record is central to the construction of geologic time scales, linking biostratigraphies, isotope stratigraphies, and absolute ages. The application of magnetic stratigraphy in geologic investigations is now commonplace; however, the use of magnetic stratigraphy as a correlation tool in sediments and lava flows has developed only in the past 35 years. As recently as the mid-1960s, magnetic polarity stratigraphy dealt with only the last 3 My of Earth history; today, a fairly well-resolved magnetic polarity sequence has been documented back to 300 My before present. An increasing number of scientists in allied fields now use and apply magnetic stratigraphy. This book is aimed at this expanding practitioner base, providing information about the principles of magnetostratigraphy and the present state of our knowledge concerning correlations among the various (biostratigraphic, chemostratigraphic, magnetostratigraphic, and numerical) facets of geologic time. Magnetic Stratigraphy documents the historical development of magnetostratigraphy (Chapter I), the principal characteristics of the Earth's magnetic field (Chapter 2), the magnetization process and the magnetic properties of sediments (Chapter 3), and practical laboratory techniques (Chapter 4). Chapter 5 describes the essential elements that should be incorporated into any magnetostratigraphic investigation. The middle of the book is devoted to a survey of the status of magnetic stratigraphy in Jurassic through Quaternary marine (Chapters 6-8, 10, and 11) and terrestrial (Chapter 9) sedimentary rocks and describes the integration of the biostratigraphic and chemostratigraphic record with the geomagnetic polarity time scale (GPTS). Recent developments in extending the magnetic polarity record to pre-Jurassic sediments and the emergence

xii

of a magnetic polarity record for the Triassic, Permian, and Carboniferous periods are chronicled in Chapter 12. The record of secular variation of the geomagnetic field has been important in attempting correlations in sediments that span the last 10,000 years (Chapter 13). Magnetic stratigraphy is expanding further through the use of nondirectional magnetic properties of sediments, such as magnetic susceptibility and a host of other magnetic parameters sensitive to magnetic mineral concentration and grain size (Chapter 14). It is now becoming apparent that geomagnetic intensity variations, as recorded in sediments, may well be the next important magnetostratigraphic tool, promising to provide high-resolution global correlation within polarity chrons. The authors thank the following people, who made important suggestions for improving the manuscript: B. Clement, D. Hodell, K. ~ u a n E. i Irving, D. V. Kent, E. Lindsay, B. MacFadden, M. McElhinny, S. K. Runcorn, and J. S. Stoner. Parts of this book were written while one of us (J.E.T.C.) was on sabbatical in Kyoto (Japan) and Gif-sur-Yvette (France). M. Torii, C. Laj, and C. Kissel made these sabbaticals possible. Finally, we express special appreciation to Marjorie Opdyke, who patiently and skillfully produced the original figures for this book with loving attention to detail. The authors of Magnetic Stratigraphy received their doctoral training from the University of Newcastle-upon-Tyne, where Professor S. K. Runcorn inspired their life-long interest in paleomagnetism. We are saddened by his murder in December of 1995 and dedicate this volume to Dr. Runcorn as an expression of our gratitude to him as our friend and mentor. Neil D. Opdyke James E. T. Channel1

Table Key

Throughout this book, compilations of important magnetostratigraphic studies are presented as blacktwhite (normal/reverse) interpretive polarity columns. These are accompanied by tables that include the following information for each study. Rock unit Age range Region h

NSE NSI NSA M D

A designation of the rock unit, usually as a formation name, locality, or core/hole/site number Age range of the stratigraphic section(s) Country or region of study Site latitude (positive :north, negative :south) Site longitude (positive :east, negative :west) Number of sections Number of sites or site interval (m). P (includes passthrough magnetometer measurements) Number of samples per site Section thickness (m) Associated absolute age control: K (K-Ar), A ( 4 0 ~ r / 3gAr),F (fission track), R (RbISr), S (strontium isotopic ratios), 0 (FI80), As (astrochronology) Rock magnetic measurements/observations: K (susceptibility), J (Js/T), I (IRM), V (VRM), A (ARM), H (hysteresis measurements), TdI (thermal demagnetization of IRM), T (electron microscopy), 0 (optical microscopy), X (X-ray) Demagnetization procedure: A (alternating field), T (thermal), C (acidlchemical) How was the magnetization direction determined? B (blanket demagnetization), Z (Zijderveld/orthogonal

xiv

A

NMZ NCh %R RT F.C.T. Q

projections), V (vector analysis), P ("principal component" or 3D least squares analysis) Data presentation: F (Fisher statistics presented), D (declination plotted), I (inclination plotted), V (VGP latitudes plotted) Number of magnetozones (polarity zones) Number of chrons Percentage of reverse polarity Reversal test: R + [positive in the A, B, C rating of McFadden and McElhinny (1990) or according to McElhinny, (1964)], R- (negative) F (fold test), C (conglomerate test), Co (baked contact test); indicated as positive or negative Reliability index (see Chapter 5.7)

It should be noted that the quality of the data is sometimes not reflected in the reliability index. For example, some marine sediments are such good magnetic recorders that it is possible to get excellent quality magnetostratigraphic data with only basic techniques such as blanket demagnetization with alternating fields, as can be seen in some whole core data from the Ocean Drilling Program.

Introduction and History

1.1 Introduction Paleomagnetism has had profound effects on the development of Earth sciences in the last 25 years. In the early days, paleomagnetic studies of the different continental blocks contributed to the rejuvenation of the continental drift hypothesis and to the formation of the theory of plate tectonics. Paleomagnetism has led to a new type of stratigraphy based on the aperiodic reversal of polarity of the geomagnetic field, which is now known as magnetic polarity stratigraphy. Magnetic polarity stratigraphy is the ordering of sedimentary or igneous rock strata into intervals characterized by the direction of magnetization of the rocks, being either in the direction of the present Earth's field (normal polarity) or 180" from the present field (reverse polarity). This new stratigraphy greatly influenced the subject of plate tectonics by providing the chronology for interpretation of oceanic magnetic anomalies (Vine and Matthews, 1963). In the last 25 years, the geomagnetic polarity time scale (GPTS) has become central to the calibration of geologic time. The bridge between biozonations and absolute ages and the interpolation between absolute ages are best accomplished, particularly for Cenozoic and Late Mesozoic time, through the GPTS. In most Late Jurassic-Quaternary time scales, magnetic anomaly profiles from ocean basins with more or less constant spreading rates provide the template for the GPTS. Magnetic polarity stratigraphy on land or in deep sea cores provides the link between the GPTS and biozonationslbioevents and hence geologic stage boundaries. Radiometric absolute ages are correlated either directly to the GPTS in magnetostratigraphic section or indirectly through

2

1 Introduction and History

biozonations, and absolute ages are then interpolated using the GPTS (oceanic magnetic anomaly) template. In view of the paramount importance of the calibration of geologic time to understanding the rates of geologic processes, the contribution of magnetic polarity stratigraphy to the Earth sciences becomes self-evident. The dipole nature of the main geomagnetic field means that polarity reversals are globally synchronous, with the process of reversal taking 1O3-lo4 yr. Magnetic polarity stratigraphy can therefore provide global stratigraphic time lines with this level of time resolution. Three other techniques in magnetic stratigraphy, not involving the record of geomagnetic polarity reversals, have become increasingly important. Rock magnetic stratigraphy refers to the use of nondirectional magnetic properties (such as magnetic susceptibility and laboratory-induced remanence intensities) as a means of stratigraphic correlation. Paleointensity magnetic stratigraphy, the use of the record of geomagnetic paleointensity, and secular variation magnetic stratigraphy, the use of secular directional changes of the geomagnetic field, have been used as a means of stratigraphic correlation in Quaternary sediments. Individual texts exist for paleomagnetism in general, such as those of Irving (1964), McElhinny (1973), Tarling (1983), Butler (1992), and Van der Voo (1993); however, none is available for the rapidly growing discipline of magnetic stratigraphy.

This work was followed by that of Matuyama (1929) on volcanic rocks from Japan and north China. This study included many examples of lavas that yielded directions that were reverse with respect to the present geomagnetic field (Fig. 1.1). Matuyama correctly attributed the reverse directions of magnetization to areversal of the geomagnetic field. Matuyama separated his samples into two groups: Group I, which were all Pleistocene in age, gave NRM directions which grouped around the present direction of the geomagnetic field. The group I1 directions were antipodal to those of group 1 and to the present directions of the geomagnetic field, and most of these bvas were older than Pleistocene in age. This was the first hint that the polarity of the geomagnetic field might be age dependent. These early studies were followed by those of Chevallier (1925), Mercanton (1926), and Koenigsberger (1938). The modern era of studies of reversals of the geomagnetic field began with those of Hospers (1951,1953-1954) in Iceland and Roche (1950,1951, 1956) in the Massif Central of France, which elaborated on the early work of Brunhes. In both of these studies, the relative age and position of the rocks were fixed stratigraphically. As in Matuyama's studies, the results indicated that rocks and sediments designated as Upper Pleistocene and Quaternary in age possessed normal directions of magnetization, whereas

C

1.2 Early Developments Directions of natural remanent magnetization (NRM) reverse with respect to the present ambient magnetic field of the Earth were known to early pioneers in paleomagnetic research. Brunhes (1906) reported directions of magnetization in Pliocene lavas from France that yielded north-seeking magnetization directions directed to the south and up, rather than to the north and down. He attributed this behavior to a local anomaly of the geomagnetic field. Brunhes demonstrated that the baked contacts of igneous rocks are magnetized with the same polarity as the igneous rock. This was the first use of what is now referred to as the baked contact test, which can be used to determine the relative age of magnetizations in the vicinity of an igneous contact. When lava flows are extruded, or dikes are intruded into a host rock, the temperature of the rock or sediment into which the magma is intruded is raised above the blocking temperature of the magnetic minerals. The magnetization is, therefore, reset in the direction of the field prevailing at the time of baking. Brunhes carried out this test for both normally and reversely magnetized igneous contacts.

-z

Y MANCHURIA

ygum 1.1 Directions of natural rernanent magnetization in basalts from China and Japan (after Matuyarna, 1929).

4

1 Introduction and History

reverse directions of magnetization appeared in rocks of early Pleistocene or Pliocene age. Einarsson and Sigurgeirsson (1955) utilized a magnetic compass to detect reversals in Icelandic lavas and began to map the distribution of magnetic polarity zones in the rock sequences. They found about equal thicknesses of normal and reverse polarity strata, which implied that the magnetic field had little polarity bias in the Late Cenozoic. A study by Opdyke and Runcorn (1956) on lava flows from the San Francisco Peaks of northern Arizona showed that all young lavas studied were normally magnetized, whereas the stratigraphically older series of flows contained reversely magnetized lavas. Based on these studies, it was assumed that the last reversal of the field took place close to the Plio-Pleistocene boundary. Rutten and Wensink (1960) and Wensink (1964) built on Hosper's studies in Iceland and demonstrated that the youngest lavas were normally magnetized and that older underlying lavas were reverse, and these in turn were underlain by a normal sequence. They subdivided the Plio-Pleistocene lavas into three magnetozones N1-RI-N2.They correlated N2 to the Astien and R1 to the Villafranchian and deduced that the earliest glaciations in Iceland were of Pliocene age. Magnetometers capable of measuring the magnetization of igneous rocks had been available since before World War 11. Johnson et al. (1948) developed a rock generator magnetometer which could measure natural remanence in sedimentary rocks. Blackett (1956) developed astatic magnetometers whjch were a further improvement in sensitivity. Paleomagnetic study of sedimentary rocks began with the classic study of Creer et al. (1954) which documented 16 zones of alternating polarity in a 3000-m section of the Torridonian sandstone (Scotland). These rocks passed the fold test, indicating a prefolding magnetization, implying that reversals were a long-term feature of the geomagnetic field. These authors also reported reversals from rocks of Devonian and Triassic age. Simultaneous with the developments outlined above, which seemed to favor reversal of the dipole geomagnetic field as an explanation for the observations, a discovery was made in Japan which cast doubt on the field reversal hypothesis. Nagata (1952), Nagata et al. (1957), and Uyeda (1958) demonstrated that samples from the Haruna dacite, a hyperthermic dacite pumice from the flanks of an extinct volcano in the Kwa district of Japan, was self-reversing. The striking magnetic property of this rock is that, when heated to a temperature above 210°C and cooled in a weak magnetic field, the acquired thermal remanent magnetization (TRM) is in opposition to the applied field. Since self-reversal in rocks could be demonstrated, it became important to ascertain whether all reverse directions could be explained in this way, or whether both the self-reversal phenomenon and reversal of the main dipole field were taking place.

ce for Field Reversal

.3

5

Evidence for Field Reversal

mebaked contact test first employed by Brunhes played an important role is settling the self-reversayfield reversal controversy. If field reversal is the norm, one would expect baked contacts to be magnetized in the same polarity as the baking igneous rock. If self-reversal is the norm, one would expect the baked contacts to have different polarity from the igneous rock in a significant proportion of the cases, since it is unlikely that both igneous and country rock would have the same magnetic mineralogy. Wilson (1962) conducted a survey of such studies and found that in 97% of the contacts studied, the polarity of the baked contact was the same as that of the igneous body. These observations supported the hypothesis of geomagnetic field reversal. By the late 1950s, several lines of evidence supported the hypothesis that the geomagnetic field reverse polarity. (1)All rocks of Late Pleistocene and recent age, both igneous and sedimentary, that had been studied possessed normal polarity magnetizations. (2) Reverse polarity magnetizations were observed in rocks of Lower Pleistocene or older age and in sedimentary and igneous rocks involving different magnetization processes. (3) Transitional (intermediate) directions were observed in Iceland in both reverse-to-normal and normal-to-reverse transitions, as well as in the Torridonian of Scotland. (4) Self-reversal had been observed in the laboratory in only one rock type (Haruna dacite) although many attempts had been made to replicate the experiment in other rock types. (5) As described above, in 97% of the contact tests, the baked contact and the baking rock possessed the same polarity. By the late 1950%many paleomagnetists believed in geomagnetic field reversal although geophysicists in general were still skeptical. In the early 1960s, the stage was set for the convincing demonstration of geomagnetic field reversal. A series of interdisciplinary studies were carried out in which KlAr dating and the measurement of magnetization polarity were carried out on the same lavas (Cox et al., 1963; McDougall and Tarling, 1963b). These studies established that rocks of the same age had the same polarity and led to the establishment of the first radiometritally dated polarity time scale. As information accumulated throughout the 1 9 6 0 ~the ~ magnetic polarity sequence was progressively modified (Fig. 1.2), particularly with the discovery of short magnetic polarity intervals (Gromm6 and Hay, 1963). The long intervals of constant polarity of the geomagnetic field were designated by Doell and Dalrymple (1966) as magnetic epochs and named after the pioneers in the study of the geomagnetism (Brunhes, Matuyama, Gauss, and Gilbert); the shorter intervals (events)

.

6

1 Introduction and History

and oceanic islands and were not in stratigraphic superposition, representing a major departure from standard stratigraphic procedures. The anomalous situation arose in which, unlike classical stratigraphy, no type sections for the magnetic epochs (polarity chrons) are available, yet type localities for magnetic events (polarity subchrons) are known. The developments which led to the first dated polarity time scale were soon followed by studies of magnetic stratigraphy in sediments and investigations of oceanic magnetic anomalies. Khramov (1958) in the Soviet Union reported reverse and normal magnetizations in Plio-Pleistocene nonmarine sedimentary sequences from the Caucasus. These studies were followed by the discovery of reversals in marine sediments (Harrison and Funnel, 1964; Linkova, 1965). Opdyke et al. (1966) observed the samesequence of polarity

Cox et al., 1963 I

I

I

-

McDougall & Tarling, 1963,1964

'=

I

I

-

I

I

-

I

Grornrne & Hay, 1963 I

I

I

-<

I

I

I

'2

Cox et al., 1963b I I

4 1

3

7

1.3 Evidence for Field Reversal

Evernden et al., 1964

--<

I

1

I

I

I

A (D

I

Cox et al., 1964b, 1965

I

I

I

I

I

I

I

1

I

I

I

Vema core 21 -148

' ' ' 2

GPTS 1966

Eltanin track 19

Doell et al., 1966

A 1

I

I

I

I

t

I

Doell & Dalryrnple, 1966

I

I

-'2

Chamalaun McDougall, and 1966

X

McDougall and Wensink, 1966

X

I

I

--x-

I I

I

Cox et al., 1966

X,

I

I

I

I

I I

I

1

McDougall et al., 1966

--

I

-x

I

I

I

,

I

I

I

t

u

I

N

N

m

N

!

I

McDougall and Charnalaun, 1966 I

I

,

,

,

,I

2

2

2

m

e

o

,

I

a m

T~me(Ma) Figure 1.2 Evolution of the geomagnetic polarity time scale of the last 2.6 My (after Watkins, 1972).

were named after the locality of discovery such as Jaramillo Greek in New Mexico and Olduvai Gorge in Tanzania. It should be noted that the time scales were produced from lavas that were scattered over several continents

Figure 1.3 Comparison of oceanic magnetic anomalies from s h ~ ptrack Eltanin 19 (Pitman and Heirtzler, 1966) with magnetic polarity stratigraphy in core V21-148 (Opdyke, 1968) and the magnetic polanty time scale of Doell and Dalrymple (1966).

I

1 Introduction and History

reversals in marine sediments that had been compiled by Cox et al. (1963) from dispersed volcanic outcrops. The same reversal sequence was then observed in oceanic magnetic anomalies by Vine and Wilson (1965) and Pitman and Heirtzler (1966). The fact that the same reversal sequence appeared to have been recorded by three different recording media (Fig. 1.3) closed the argument in favor of field reversal and relegated self-reversal to a subsidiary process. The resolution of this argument in favor of field reversal tipped the scales in favor of the hotly debated hypothesis of plate tectonics and sea floor spreading.

The Earth's Magnetic Field

2.1 Introduction The magnetic field of the Earth has fascinated human beings for well over 2 millennia. The Chinese invented the magnetic compass in the second century B.C. (Needham, 1962) and knowledge of the magnetic compass reached Western Europe over a thousand years later in the twelfth century A.D. The first truly scientific paper on geomagnetism was written in 1262 by Petrus Peregrinus and entitled "Epistola de Magnete" (Smith, 1970). In a series of experiments on lodestone (magnetite) spheres. Peregrinus defined the concept of polarity and defined the dipolar nature of a magnet stating that like poles repel and unlike poles attract; however, the treatise was not published until 1558. The work by Peregrinus undoubtedly influenced the later work of William Gilbert, who published the book "De Magnete" in 1600. Gilbert studied the variation of the angle of inclination (or magnetic dip) over lodestone cut into the shape of a sphere, perhaps one of the greatest model experiments ever made in Earth science. His conclusion was that "the Earth itself is a great magnet." From the 14th century to the present, compasses were widely used as a navigation aid on both land and sea and were carried by explorers throughout the world. The angular distance between the polar star, which does not move in the heavens, and the direction of the north-seeking end of the compass needle was often faithfully recorded. This magnetic deviation from true north is called the magnetic declination (D). In midlatitudes, from 40°N to 40's. the declination of the present geomagnetic field can vary up to 40' from true north. In polar regions, near the magnetic poles, declination anomalies of up to 180" are observed (Fig. 2.1).

10

2 The Earth's Magnetic Field

I80°W Figure 2.1 Declination for the International Geomagnetic Reference F~eld(IGRF) 1990 (after Baldwln and Langel. 1993).

If a magnetized needle is suspended on a fiber and allowed to swing freely, the north-seeking end of the needle will not only point north, it will also pgint down in the northern hemisphere and up in the southern hemisphere. This property of the earth's field, its inclination, was discwered by George Hartmann in 1544 (Smith, 1968). If a dip needle which measures the angle of magnetic inclination is carried from high northern latitudes to high southern latitudes, it will be seen that the inclination of the field varies systematically from vertical down at the north magnetic pole, to the horizontal at some point at low latitude (magnetic equator), and vertical once more (but with the north-seeking end of the needle pointing vertically upward) at the south magnetic pole (Fig. 2.2). It should be noted that the north and south magnetic poles are not 180" apart and that the magnetic equator is not equidistant from the two magnetic poles. If the geomagnetic field is analogous to that of the magnetized spheres studied by William Gilbert, the intensity of the field, as well as the inclination and declination, would vary systematically over the Earth. After Gauss invented the deflection magnetometer, this was found to be the case, the magnetic field strength at the magnetic poles being almost twice that at the magnetic equator. The field at any point on the Earth's surface is a vector (F) which possesses a component in the horizontal plane called the horizontal component (H) which makes an angle (D) with the geographical meridian (Fig.

2.2 Inclination for the International Geomagnettc Reference F~eld(IGRF) 1990 (after in and Langel, 1993)

declination (D ) is an angle from north measured eastward ranging to 360'. The inclination (I) is the angle made by the magnetic or with the horizontal. By convention, it is positive if the north-seeking or points below the horizontal or negative if it points above. The ontal component (H) (Fig. 2.3) has two components, one to the north and one to the east Y. The following equations relate the various quan-

F sin I

H = F cos I

Z

X=HcosD

Y=HsinD

=

Tan I

=

ZlH

TanD=Y/X

F 2 = H2 + Z2 = X2 + Y2 + Z2,

(2.1)

(2.2)

(2.3) e dipole nature of the geomagnetic field, originally suggested by , was first put to a mathematical analysis by Gauss (1838), who measurements of declination, inclination, and intensity from 84 widely ced locations to derive the first 24 coefficients in the spherical harmonic Pansion of the geomagnetic field. The results of this analysis indicated at, within the uncertainties of the data, the geomagnetic field had no rces external to the Earth and that it had a dominant dipole component. ce 1835,when Gauss first determined the intensity of the dipole moment, e calculations have been repeated many times, and this value has been aeasing steadily over the last century and a half. During this century,

12

2 The Earth's Magnetic Field

13

2.1 Introduction

N

+ North (Geographical)

Geomagnetic north p

-Best

fitting dipole

South magnetic

Geographic pole $ Downwards Figure 2.3 The direction and intensity of the total field vector (F) resolved into decllnat~on from geographical north (D) and incl~nat~on from horizontal (I). Equations (2.1), (2.2), and (2.3) relate the various quantities.

.

the decay rate has increased, and during the last 30 years has reached a rate of 5.8%/century (see Barton, 1989). If the trend were to continue, the main dipole field would disappear by the year 4000 A.D., a possible but unlikely event. Modern spherical harmonic analyses confirm Gauss' findings and lead to the conclusion that: (1)the geomagnetic field is almost entirely of internal origin, (2) -90% of the field observed on the surface can be explained by a dipole inclined to the Earth" axis of rotation by 11.5" (Fig. 2.4), (3) the Am2, (4) the axis of the magnitude of the dipole moment is 7.8 X centered dipole (geomagnetic pole) emerges in the northern hemisphere between Greenland and Ellesmere Island at 79.0°N, 70.9"W. The great circle midway between the geomagnetic poles is called the geomagnetic equator (Fig. 2.4). Magnetic observations of the elements of the geomagnetic field were begun as early as the 16th century in London and the 17th century in Paris. It was realized early on that the declination and inclination of the

Figure 2.4 Graphical presentation of the magnetic, geomagnctic. and geograph~cpoles and equators (after McElhlnny, 1973).

geomagnetic field change with time (Fig. 2.5). This relatively rapid change of the magnetic field with time is called secular variation. The declination of the field at London has been changing at a rate of about 14" per century from O" in 1650 to 24"W in 1800 and is now 5"W. The change in the magnetic elements is not constant over the Earth's surface and, for example, changes less rapidly over the Pacific Ocean than in, say, South America. Global secular variation is often displayed as "isoporic" charts, which are contour maps of equal annual change of a particular element of the field, such as the vertical component. From isoporic charts constructed from observations dating back about 200 years, isoporic centers have drifted westward at rates of up to about 0.28"tyr over this time interval. This observation is a manifestation of the westward drift of dipole, quadrupole, and octupole components of the field. Drift rates of these components can be calculated directly from the spherical harmonic coefficients and their changes with time, Changes in the rate of westward drift and the growth and decay of isoporic features occur on a decadal time scale. The westward drift is generally attributed to differential angular velocity of Earth's lithosphere and outer core, where the nondipole components of the field are thought

.

2 The Earth's Magnetic Field

15

2.2 The Dipole Hypothesis

of hours, it is theoretically possible to detect ancient field behavior on this time scale from paleomagnetic data.

2.2 The Dipole Hypothesis

I

I

The fact that the geomagnetic field can be modeled as a geocentric dipole is of fundamental importance in paleomagnetic studies. If we assume that the time-averaged dipole is aligned with the axis of rotation of the Earth and situated at the center of the Earth (an axial geocentric dipole), the and geographic axes will coincide, and the geomagnetic equator will coincide with the geographic equator. The geomagnetic inclination (I) will vary systematically with latitude (A) according to the formula:

Decllnat~ons(degrees) Figure 2.5 Change of decl~nationand inclination at London and Paris from observatory records (after Gaiber-Puertas, 1953).

Tan (I) = 2 Tan (A). If M is the moment of the geocentric dipole and the radius of the Earth is r, the horizontal (H) and vertical (Z) components at any latitude (A)

to originate. It is important to note that although westward drift is dominant, not all secular variation features drift westward; some are more or less stationary and some are moving slowly eastward. The geomagnetic field varies over a wide range of time scales (Table 2.1). The longer term behaviors (4-7, Table 2.1) are produced in the Earth's interior, an$ the short-term behaviors (1-3, Table 2.1) are atmospheric or ionospheric in origin. Secular variations of the field, first observed in observatory records, can be monitored further back in time using paleomagnetic data from marine and lake sediments with high accumulation rates. The longer term variations (5-7, Table 2.1) have also been documented by paleomagnetic study of rocks and sediments. Since some lavas and archaeological artifacts can become permanently magnetized in a matter

can b e derived from the following geometric relationship (Fig. 2.6), where I is the inclination from the horizontal. Using the formulae used by Gauss for his deflection magnetometer, H = (M cos A) /$ Z = (2M sin A) lr'

Table 2.1 Scales of Geomagnetic Variabiliv

Geomagnetic behavior

Duration

1. Pulsations or short-term fluctuation 2. Daily magnetlc variations 3. Magnetlc storms 4. Secular variations 5. Magnetic excursions 6. Reversal transition 7. Interval between reversals

minutes hours hours to days lo2-10' yr 10"1O4 yr 10-1-lo4 yr lo5-lo6 yr

Figure 2.6 Field lines at the Earth's surface for the axial geocentric dlpole (after McElhinny, 1973).

16

2 The Earth's Magnetic Field

Tan I = ZIH if D

=

=

2 Tan (A),

2.3 Models of Field Reversal

17.

(2.5)

0°, colatitude ( 8 ) is given by Tan 1 = 2 cos 0 (0 < 8 < 180").

(2.6)

The relationship makes it possible to determine the paleolatitude of a site from mean paleomagnetic directions at any point on the Earth's surface. In the discussion given above, the declination ( D ) was presumed to be zero, although in most paleomagnetic investigations this is not so. The paleomagnetic pole represents the position on the Earth's surface where the dipole axis (oriented to give a field in agreement with the mean paleomagnetic direction) cuts the surface of the globe. Every paleomagnetic investigation, if successful, will yield a declination (D)which represents the angle between the paleomagnetic vector and the present geographic meridian. The declination defines a great circle that passes through the sampling site and the paleomagnetic pole. The mean paleomagnetic inclination gives the distance along the great circle (or paleomeridian) to the paleomagnetic pole, according to the dipole formula given above (2.5). The pole position can be calculated in the present geographical coordinate system using the following relationship. If Dm, I, are known at a sampling site with latitude A and longitude @ then the position of the paleomagnetic pole P (A1,@')can be calculated sin A '

=

@' = @ @' = 4,

sin Acos &I+ cos Asin OcosD,

+p

(whencos 8 > sin hsin A')

+ (180 - P )

(when cos 8 < sin A sin A'),

(2.7) , (2.8) (2.9)

where sin p = sin 0 sin Dlcos A'. By convention, latitudes are positive for the northern hemisphere and negative for the southern hemisphere; longitudes are measured eastward from the Greenwich meridian and lie between 0 and 360". Other symbols are defined in Fig. 2.7. In paleomagnetic research, two types of paleomagnetic pole positions are calculated. A virtual geomagnetic pole (VGP) is a pole calculated from magnetic data which represent a very short period of time, such as the observatory record of D and I of the present field, or a spot paleomagnetic field reading from a single lava flow. A paleomagnetic pole, on the other hand, represents the field averaged over a considerable length of time (on the order of 50 ky) such that short-term (secular) variations are averaged out. A paleomagnetic pole position is essentially the mean vector sum of many virtual geomagnetic pole positions. In the middle 195Os, it was known that the mean Plio-Pleistocene paleomagnetic poles more or less coincide with the pole of rotation of the Earth. A compendium of poles by Tarling (1983) yields an average

Figure 2.7 Calculation of a paleomagnetic pole from mean dlrcctions of magnet~zation(Dm, Im) at sampling site (S) generates a paleomagnetic pole (P) (after McElhlnny, 1973).

paleomagnetic pole indistinguishable from the present pole of rotation (Fig. 2.8). These observations support one of the basic hypotheses in paleomagnetism, the dipole hypothesis, which states that the Earth's dipole field averages to an axial geocentric dipole over periods of time longer than the averaging time for secular variation (10-50 ky). The dipole hypothesis may be tested using widely spaced sampling sites for rocks of the same age, or by testing the relationship of inclination to site latitude in marine cores. Opdyke and Henry (1969) and Schneider and Kent (1990a) demonstrated that Pleistocene paleomagnetic inclinations in widely spaced marine cores are consistent with the dipole hypothesis (Fig. 2.9). Merrill et nl. (1979) and Schneider and Kent (1988, 1990a) have documented asymmetry between normal and reverse intervals over the last few million years, which can be accounted for by increasing the quadrupole term for reverse polarity intervals relative to normal polarity intervals. This asymmetry is difficult to explain due to the symmetry of the induction equations (Gubbins, 1994). Outer core boundary conditions (in the mantle or inner core) can be invoked to break the symmetry of the dynamo equations (see Clement and Stixrude, 1995).

2.3 Models of Field Reversal Although the geomagnetic field may be molded as if the Earth were a permanent magnet, it is clear that this is not the case as thermal disordering (Curie) temperatures of permanently magnetized materials are exceeded

18

2.3 Models of Field Reversal

2 The Earth's Magnetic Field

-

Figure 2.8 Paleomagnetic poles for igneous rocks younger than 20 Ma (after Tarling, 1983).

I

S

I

I

I

I

I

-60" -30" 0" 30" 60" N

Latitude Figure2.9 Inclination versus latitude in Pleistocene marine sediments. Line is expected inclination derived from the dipole formula (after Schneider and Kent, 1990a).

l9.

by burial greater than 10-20 km. The recognition of polarity reversals of the dipole field also provided a powerful argument against permanent magnetization. In 1947,Blackett proposed that planetary dipole fields are an intrinsic property of rotating bodies and subsequently designed experiments that refuted the claim (Blackett, 1952). Beginning with the work of Elsasser (1946) and Bullard (1949), it is now known that the geomagnetic field arises from convective and Coriolis motions in the fluid nickel-iron outer core. According to dynamo theory, outer core flow can amplify a small instantaneous axial field to observed dipole values, although the details of the process remain poorly constrained. No detailed theory predicting the form of geomagnetic secular variation or the mechanism of geomagnetic reversals is yet available. Gubbins (1994) and Jacobs (1994) have summarized recent developments on this topic. Uncertainties in the details of the geomagnetic field generation process lead to little consensus on the process of geomagnetic polarity reversal. The models of Cox (1968,1969), Parker (1969), and Levy (1972) involved interaction between the main dipole field and the nondipole field. According to Parker (1969) and Levy (1972), changes in distribution of cyclonic eddies, which account for the nondipole field, can produce regions of reverse torodial field which effect reversal of the poloidal (dipole) field. According to COX(1968, 1969). the critical factor is the relative field strength of the high-frequency oscillations of the nondipole component relative to the lower frequency (cyclic) oscillations of the dipole component. A highamplitude nondipole field coincident with a low in the main dipole field could, in this model, produce an instantaneous reverse total field, which is then reinforced by the dynamo process. According to Laj et al. (1979), the supposed coupling of the nondipole and dipole fields in the Cox model is unlikely in view of the different time constants of the two field components, and the mechanism (if feasible) would be expected to generate very frequent reversals. HiIlhouse and Cox (1976) supposed that the dipole component of the field dies in one direction and grows in the opposite direction during a polarity reversal and that the nondipole components do not change over times sufficient to span several reversals. The model predicts that sequential reversals, of opposite and same sense, recorded at a sampling site should yield similar transitional VGP paths. This has not been found to be the case and hence this model is no longer favored. The models of Hoffman (1977) and Fuller et al. (1979) are based on the idea that the reversal process floods through the core from a localized zone. These models can be tested as they predict the VGP paths during a polarity transition for either quadrupole or octupole transition field geometries. The VGP paths are predicted to lie either along the site longitude

20

2 l l t e Earth's Magoet~cField

(near-sided) or 180" away (far-sided) dependingon the senseof the reversal, the henlisphere of observation, and where in the core the reversal process is initiated. These models are generally no longer favored due to the lack of evidence for near-sided or far-sided VGP paths. Olson (1983) attributed reversal to a change in sign of the helicity (the correlation between turbulent vclt>cilyand vorticity). triggered by a change in the balance between heat loss at the core-mantle boundary and solidihcation at the inner-outer core boundary. In this model, the duration of a polarity chron will depend on the strength of the dipole held and the process of reversal would he rapid (-7500 yr). more or lcss consistent with palcomagnetic observations. The solar dipole fluid reverses polarity approximately every 11 years (Eddy. 1976)and the observations of the solar reversal process provide clues to the reversal process on Earth (see Gubbins. 1994). The high frequency of solar licld reversal allows the process to bc observed directly. In addition. the flux distrihution a1 the solar surface can he monitored. in contrast to the Earth, where the flux distribution at the core-mantle boundary must bc approximated hy downward continualion. Relatively low temperature sunspots are regions of intmse magnetic field, considcrcd as sites of expulsion of toroidal flux. The locations r~fsunspot pairs move from the solar polcs loward Lhe equator, leading up to a rcvcrsal of the solar dipole ficld. Bloxham and Gubbins (1985) and Gubbins (1987a,h) interpret flux concentrations in maps of the radial component of the geomagnclic field as due to eddies at the core-nlantle boundary, which they liken to sunspots. Two .'core spots" have been identified beneath South America and southern Nrica. The process at thc Earth's core-mantle boundary leading to the revcrsal o l the geomagnetic dipole may be similar to the solar reversal process. with flux spots beneath South America and southern Africa being analogous to sunspots. These corc spots produce local ficlds opposed to the Earth's main dipole field. Bloxham and Gubbins (1985) have suggcstcd that these areas of high magnetic flux account for thc high rate of secular varislion in the Atlantic region. The corc spots are apparently strengthening as the main dipole ficld decays. They are apparently moving to the south and could possibly lead to a revc~salo l the main dipole field. As the core spots in thcsc regions become larger, they might he obsenred at the surface as large regional deviations of the ficld (magnetic excursions).

2.4 Polarity Transition Records and VGP Paths In the past decade, an mcreas~ngamount of h~gh-qual~ty palcomagnehc data have been acqu~rcdfrom polarity tranwionc in high depos~tionrate

2.4 Polarity Transition Records and VOP Paths

21

sedimentary sequences (c.g., Clemenl, 1991; Tric el a/., 1991a; Laj el ul., 1991) and in volcanic rocks (c.g., PrCvot et a/., 1985; Hoffman. 1991, 1992). The records provide important insights into the configuration nf the ficld during a polarity transition; however, the interpretations are controversial. Some data indicate that transition fields have a largc dipolar component. Forexample, the records of the Cobb Mountain subchron from the western Pacific and North Allanlic have similar VCiP (virtual geomagnetic pole) paths, implying dipolar transitional fields (Clcmenl. 1992). The VC;P paths of somc spatially distributed Matuyama-Brunhes polarity transitions coincide (Clement, 1991) whereas others do mot (Valet et ul., 1989). Another controversy has involved the apparentclusteringof polarity transition VCiPs along two longitudinally constrained paths, one at -80"W longitude over North and South America and the other at -100"E over eastern Asia and Australia (Fig. 2.10). Some studies emphasize the statistical significance of the prcfcrrcd VGP paths (e.g., Clement, 1991: Laj er nl., 1991,1992). Valet et 01. (1992) have disputed both thc claim that transition fields are dipolar and the statistical clustering of transitional VGPs to lrlngitudin~lpnlhs. Quidcllcur and Valet (1994) considered that the apparent clustering of VGP paths is biased by uncvcn sitc distribution. The highest concentration of sampling sites is ins out hen^ Europe, -90" away from the VGP longitudinal path across the Americas. Egbert (1992) showed that. except in the case where the transition field is entirely dipolar. the VGPs are likely to cluster -90" in longitude away from the sampling sitc. According to Conslable (1992), thc preferred VGP path across the Americas isapparent even after the sites locatcd -90" away from the path are removed from the record, It is important to realize that prcfcrrcd longitudinal VGP paths do not ncccssarily mean that the geomagnetic field is dipolar during polarity transitions (Gubbins and Coe. IYY3). Prevot and Camps (1993) cxamincd a largc population of transitional records from volcanic rocks younger than middle Miocene and concluded that the transitional VGPs have no statistical bias in their distribution. On thc other hand. more selective use of data from volcanic rocks leads to a different conclusion. HolLman (1991, 1992) found that transitional VGPs from distributed sites of volcanic rocks youngcr than I O Ma fall into two clusters in the southern hemisphere, one over Australia and the other over South America (Fig. 2.10). I h c clusters arc locatcd close to the Asian and American longitudinal VGP paths resolved from sedimentary transition records. Thcsc dala lead lo the conclusion lhal transilirmal field stales are long-lived (i.e.. characteristic of sequential reversals) and have a strong dipolar componcnt, which is incIined to the Earth's rotation axis. Bloxhaln and Gubbins (1987) have downward continued the ficld sincc 1700 AD, to the core-mantle boundary, and it appears that magnetic flux

22

2 The Earth's MagneticF~eld

ngurr2.10 (a) Cluatercd VGPsco~iceotraredover Southeast Asiaa~rdSouth AmcricalAntarcrica from polarity lranailion records from lava flr,u,s. (h) klcans of VGPs f r r m lava Bows supnimposed on shaded prcferrcd lon~itudinalVGPpaths from sedimentary revcrsal lransilion rccords [after Hoffman. 1992).

a t the core-manllc boundary is also concentrated approximately along the east Asian and Amcrican longituditral paths. Gubhins and KcHy (1993) have analyzed p a l e o m a ~ c t i cdata for the last 2.5 My from both lavas and sediments and suggested that thc magnetic field at thc core's surface has remained similar to that of today over this period of time. l'hc persistence of the preferred paths lhrough time and thc fact that paths pass through regions oI high modern flux activity at the core-mantle boundary suggest that these rcgions are pinned in their present position and have been for some time. The long tirnc scales involved indicate that the position of these rcgions is controlled hy mantle (as opposed to outer cure) processes. Laj el al. (1991) and Hoffman (1992) have pointed out that radial Hux centers

associated with rapid secular change in today's ficld arc closcly associated with regions of fast-P-wave propagation in the lower mantle (Dziewonski and Woodhouse, 1987). indicating the presence of rclativcly cold mantlc; however, this inlerpretation of mantle tomography in terms of temperature variations is not universally accepted. The preferred IongiLudinal VGP paths lie either side of the Pacific Ocean, which is a region characterized by low-amplitude secular variation. At prescnt, there are no foci of rapid secular change in the Pacific region and none have been known yince mcasurcmcnts o l the magetic lidd in the region hegan several hundred years ago. Runcorn (1992) suggested that this low rate of secular change may be caused by a well-developed high conducting D mshell under the Pacific. According to Runcorn (1992), as the dipole begins L o reverse polarity. the magnclic lield induces currents in the Pacific D" shell, causing a torquc that rotates the outer corc rclative to the mantle until the reversing dipole lies along the American or East Asian longitudinal paths. Clemcnt and Stixrude (1995) havc invokcd magnctic susccptibility anisotropy of the inner core to account for several problematic features of the paleomagnetic field. The Earth's inner core exhibits anisotropy in seismic velocity which can hc cxplaincd in tcrms of the prcferrcd oricntation oI hexagonal close-packed (hcp) iron. Room temperature analogs of hcp iron have a strong magnetic susceptibility anisotropy, which could produce an inner corc ficld inclined to the rotation axis. This may help to explain not only the well-known far-sided effect (Wilson, 1912) but also the preferred VGP paths during polarity transitions. When the outer core field is weak during a polarily reversal, the persisting inncr core licld may bias the total field to produce the observed VGP paths and clusters observed in transition rccords. During polariq reversal, as the outer core field s o w s in the opposite direction, it may lake a few thousand years to diffusc through thc inner core and reset the inner core field. Polarity reversals in the paleomagnetic record appear to be accompanied by gcomagnctic intcnsity lows (c.g., Opdyke at., 1966). although in some sedimentary records, the apparent lows may be an artifact of the remancncc acquisition process. Paleointensily studics by Valct and Mcynadiet (1993) imply reducing geomagnetic field intensity during PlioPlcislocene polarity chrons, with intensity recovery immediately post reversal. The present axial dipolc moment is 7.8 ti 1#%m2. If it is assumcd that the calibration of the relative paleointensity to absolute values is correct, then thc dipolc momcnt during rcvcrials falls lo about 1 ti lo2' Am2. or about a 10% of its present value. Valet and Meynadier (1993) pointed out that the low*?in apparcnt paleointensity within the Brunhes Chron correspond to the age of documented field excursions (Ch. 14).

I

I

I 1,

1

'I

!

1

II

j

I

,i il

24

2

The Earth's Magnetlc Field

2.5 Statistical Strucfure of the Geomagnetic Polarity Pattern As the GPTS is refined and our record of polarity reversals hccomes more complete. estimates of the statistical structure of the polarity sequencc become more realistic. From present to 80 Ma, since the Cretaceous long normal superchron. the polarity sequencc approximates to a gamma process where the gamma [unction ( k ) equals -1.72 (Phillips, 1977) or, in an updaled estimate, -1.5 (McFadden, 1989). The Poisson process is thc special case of a gamma proccss where k = 1,corresponding to a random proccss. The observation that tho gamma function is greater than 1 means that either the process has "mcmvry," which reduces the likelihood of areversal immediately post reversal, o r that Ihe record of reversals lor the last SO My is incomplete with many short-duralionpolarity chrons missing from the record. 'The oceanic magnetic anomaly rccord is characterized by numerous "tiny wigglcs" referred to as cryptochrons in the. time scale (Candc and Kent, 1992b). Inclusion of cryptvchrons in the rccord can. depending on their duration, reduce the valuc ol k to values very close to I, indicative of a Poisson distribution (see Lowric and Kent, 1983). Although the issue of tiny wiggles in the oceanic magnetic anomaly records is not scttled. it is now popular to intcrpret them as due to palcointensity variations rather than short polarity chrons. If this is correct. it implics that polarity reversal of the geomagnetic field is not a Poisson process but that the probability of a reversa drops to zcro immediately after a revcrsal and then rises to a steady value. The rate of reversal has varied with time and thc variation in rate is quite smooth, with a steady decrcase [rum 165 Ma to 120 Ma and a steady increase from 80 Ma to the present. The change in revcrsal rate results in a change in the potential resolution o l magnetic polarity stratigraphy as a correlation tool. From 120 Ma to 80 Ma, during the Cretaccous normal superchron, thc rcvcwal process was not funclioning. which may imply that the geomagnetic field is more stablein the normal polarity state. McFadden and Merrill (1984) analyzed the relative stability of normal and rcverse polarity states and concluded that there is no evidcnce for a stability hias bctween the two polarity states. The existence of a superchron of reverse polarily in the Late Carboniferous-Pcrmian of comparable duration to the Cretaceous normal superchron tends to support the notion that the polarity of superchrons is not dictated by stability bias. There isevidencc, however, for a difference in Lhe relalive proportions or zonal harmonics which make up the time-averagcd field For the two polarity states. On the basis of data for the last 5 My: Mcmll andMcElhinny (1977) concluded that the reverse pcllaf ly lield has a larger quadrupole and octupole contcnt than the normal

2.5 Sla~isticalStructure of the Geomagnetic Polarity Pattcrn

25

polarity ficld. Schncider and Kent (1988) reached a similar conclusion based on the analysis of deep-sea core inclinalion data. This apparent asymmetry for normal and reverse polarity states may bc attributed to an axis of magnctic susceptibility anisotropy in thc inner core inclined to the rotation axis (Clement and Stiwrude, 1995).

27

3.1 Basic Principle

retain a rcmanent magnetization. Many clay minerals and common ironbearing minerals such as siderite, ilmcnitc, biotite, and pyrite are paramag-

1

4

1

Magnetization Processes and Magnetic Properties of Sediments

1 11

1

II

I I

3.1 Basic Principle All matter responds to an applied magnetic ficld, due to tbe cIfect of the field on electron motions in atoms. although the response is very weak for most materiqls. The ions of some transitional metals, notahly Fez-. Fe'-. and Mn2', carry an intrinsic spin magnctic moment and are rcferred to as puramagnetic ions. Materials which contain these cations exhihit an enhanced response to an external field. The weakest response is called diunragndisnz. Diamagnetic materials do nor contain paramagnetic ions in appreciable concentratio~~s and when placed in a magnetic field ( H I , a weak magnetization ( M i ) is induced, by the effcct of thc magnetic field on orbiting electrons, which is antiparallel to lhe applied field ( H ) .Thc susceptihility (k) (wherc Mi = kH) is therefore small and negative. No (remanent) magnetization (M,) remains after the magnetizing field is removed. Many of the most common minerals (such as quartz. halite, kaolinite, and calcitc) are diamagnetic, and rucks which are composed largely of these minerals will have negative susceptibility. An enhanced response to a magnclic field is referred to as para~nugr~etism. Paramagnetic materials contain paramagnetic ions. The susceptibili~y is large (relative to diamagnetic susceptibility) and positivr: and due to the alignment of the intrinsic magnetic moments in the applied field. The susceptihility is proportional to thc reciprocal of ahsolutc temperature. When the applied ficld is removed, thc weak slignmcnl of spin momcnts is randomized by lhermal vibrations and paramagnetic materials do not 26

A few minerals, some iron oxides; oxyhydroxides, and sulfides, exhibit a third typc of magnetic behavior called ferro~nagnefirnr (sensu lato). In these materials, the paramagnetic cations are closely juxfaposcd in the crystallaktice such lhat spinmoments are couplcd either by direct exchange or bv. superexchange through an intermediate anion. Ferromagnetic suscep. tihility values vary widely but are high (rcl.dtive to paramagnetic susceptibility values) and positive. Unlike diamagnctic and paramagnetic susmptibility, ferromagnetic susccplihility is strongly depcndcnl on the strenglh of the applied field (H). At low H, the induced magnetization (M,) is proportional to H, and therefore the initial sz~scrptihility(k) is constant, and the magnetization is lost alter removal of the applicd field. However, at higher values of H, part of the magnetization is retained by the material after the removal of the applied field. This hchavior can be summarized by use of a hystere-vis l c ~ o pwhich , is a plot of magnetization ( M ) against applicd field ( H ) (Fig. 3.1). At lnw H. M increases proportionally (hence, k is constant for low H) and the magnetization proccss is reversible. A t highcr H, the magnetization is not . proportional to H, the curve is not reversible, and a small proportion of thc magnetization, referred to as the remurzcnr magrcetizatlon (M,), is retained by the material after ficld removal. Eventually. as

800

-600

-400

200

0

ZOO

400

600

800

H Appl~edField (mT)

Figure3.1 Hysteresis loop rrrnn apink pelagiclimestone. illustrating lhs saturationmagnetization (Mr), saturation remunence (Mrs). and cocrcive force (Hc). The loop is conctricrcd (u'aspwaislrd) due to prcscou: of low-corrciuily rnagctitc and high-coercivity hemaLilc.

28

3 Wdgnctization Processes and Magnaic Properties ui Sediments

H increases further, the suturuti(~nmngfletizalion (M,) is reached. Removal of the applied field leaves the sample with a suturation remadent magneriza-

lion ( M , ) . II the applied field is now increased in thc opposite sense ( - H ) . the total magrictizaliun of the sample will fall to zero at a value of H refcrrcd to a s the coercive,fi?rcr(H,) for that sample. The back field which leaves the sample with zero relnanent magnetization is referred to as the remanent coercivily or the coerciviry of rema~tcnci!(H,,). The hysteresis loop can be completed by taking (be sample to its saturation magnetization in the back field and imposing an increasing positive H until the saturatron is achieved again. 'The hysteresis properties of ferromagnetic materials are destroyed at a characteristic temperature for the mineral relerred to as thc Curir or Ngel tempernlure, above which the behavior is paramagnetic. l'he term ferromagnetism (se~lsoluto) is used t o refer to materials which show hystcrcsis behavior. There are various types o l ferromagnetism which differ in magnetic properties due to different arrangcmcnls of spin moments within the crystal lattice. In the case of ferromagneri~m (fensu stricto), the adjacent spin moments are parallel and of unicorm magnitude (Fig. 3.1).This behavior is rcstrictcdto nalive iron. nickel, andcobalt where paramagnetic ions are sufficiently closely juxtaposed for direct exchange interactions. Minerals (on Earth) do not exhibit this typc or behavior. In ferrimagnetiswr adjacent spin momenis are antiparallel but a net moment results due lo their unequal magnitude. In anlirerrt~magnetisni,adjacent spins are antiparallcl and of uniform magnitude and hence no net moment results. If, h'ou~ever,the antiparallelism of ihc spin moments is impcrfccl due to canting, a weak net moment results. Ferrimagnctism-and canted antiferromagnetism arc the behaviors exhibited by miuerals which are capa-

ble of carrying magnetic remanence, the magnetic mint.ml,~.'l'hc ferrimagneticminerals generally havehighsusceptibilities,highmagnetization inlcn,ities, and relatively low cnercivities, whereas the converse is true for antiferromagnetic minerals.

3.2 Magnetic Minerals a. Magnetite (Fe304) and t h e

Titanomagnetites (xFe2TiO4.11

- x]Fe304)

Magnetite is one of the cnd members of the ulvospincl-magnetite solid solution scrics (Fig. 3 . 3 ) and is the most common magnetic mineral on Earth, Magnetite is a cubic mineral wilh inverse spinel structure. The disordering (Curie) temperaturc varies more or less lincarly with Ti content from 1 5 0 ° C for ulvospinel to 580°C for magnetite (Fig. 3.4~1).Hence. ulvospinel isparamagneticat room temperature. Ti(anomaqetitesareferrimagnetic at room temperature lor values of .x bclow about 0.8. Complctc

ALIGNMENT OFSPIN MOMENTS

Fen6magneusm Nst Moment

Ferr8nlagnetlsm

Antlfermmagnetism

Canted

Y,,

Anl#ferr~magnet~em

1

Figure 3.2 Arrdngrmcnt ot sp~nmomcnts asso~latrd~ t t hdlllerml clnsscs of maenenc kha\ lor

4

Wust~te FeO

Magnetite

Fez04

d

Hemat~te Maghemlte Fez03

Figure3.3 Ternary diagram indicating the important iron oxide m a ~ l c t i minerals. c thc titanomagnetite and ritannhcmatite colid strlulion series. ilnd the lililnomilghrmi~sr~xidatirlntrend.

30

3 Magnetization Processes and Magnetic Properties of Sediments

. :,

Cornposihon. x

Figure 3.4 Saturation magncrization (I,) and Curie tcmperaturr (T,) as a function of

(a) titannmagnctite composition and (b) titanofiemalile compositic,n (affer Nagata. 1961; Butler. 1992).

solid solutlon between ulvosplnel and magnetite occun only above 6DOnC. and ex5olution lamellae are common in igneous rocks. The exsnlut~onis usually not between ulvospincl and magnetite, but between llmenite and magnct~tedue to the hlgh-temperature (druteric) oxidation of ulvosplnel to ilmen~te,which usually accompanies exsolution (Haggerty, 1976a.b). As the ilmenlte lamellae develop, the interven~ngtitanomagnetlte compos~tion

becomes progressively depleted in titanium. Further high-temperature oxidation results in the replacement of ilmenite and magnetitc by aggregates of hemalile. mtile, and pswdobrookitc. High deuteric oxidation states are most common in acidic plutonic rocks which cool slowly and have high water content. A wide variety of deuteric oxidation states can be observed within a single lava How and the controls are not clearly defined. In general. unexsolvcd titanomagnetites are most coinmon in basic igneous rocks which have coolcd rapidly. such as newly erupted submarine basalts. The exsolution process involves a change in grain size, grain shape. and composition of the titanomagnetite. The changes will generally involve a decrease in x, a decrease in grain size, and an elongation of the grain shape, and will therefore Lend to increase the coercivity and blocking tcmpcraturc o[ the grain. ThereIorc, the exsolution of magnetite and ilmcnitc by deuteric oxidation may serve to increase the stability of the rcmancncc acquired during the process. The ilmenite separating the exsolved titanomagnelite rods is paramagnetic at room temperature and behaves as a nonmagnetic matrix for the ferrimagnetic titanomagnetite grains. Exsolution lirmellae are a characteristic of titanomagnetite grains affected by dcutcric oxidation. Low-temperature oxidation of titanomagnetites by weathering processes andlor oxidizing diagenetic conditions results in maghcmatization without exsolution, and subsequent formation of hematite. Titanomagnetites are a common constituent of igneous and mctamorphic rocks and are therefore aprominent detrital component in scdimcnts. Many sediments also contain low-titanium magnetite grains. which arc uncommon in igneous rocks and are generally believed to be hiogcnic. Many organisms, including bacteria and inolluscs_produce magnetite either by extracellular precipitation or as an integral part of their mctaholism (e.g., Blakemore. 1975; Kirschvink and Lowenstarn, 1979; Blakemorc et ul., 1985; Frankel, 1987; Lovley et ul., 1987: Lovley, 1990; Bazylinski cl a!., 1988; Chang and Kirschvink, 198% Sparks et nL., 1990). Electron microscopy of magnetic extracts of lakc sediments (Snuwball, 1994) and a wide range of marine sediments (Vali el ul., 1987: McNcill, 1990: Yamazaki er a/.. 1991) has shown that magnetite is often found in a restrictcd single domain (SD) . . grain size range, similar to the grain size of magnetite produced by living bacteria (Fig. 3.5). 'lhc grain size and low-titanium composition of this magnetite phase is such that it is an important carrier of itable magnetic remanence (see Moskowitz er nl., 1988, 1904).

b. Hematite (m-Fe,03) and Titanohematite (xFeTi03.[l

- x]Fe,O,)

Mineral compositions intermediate between ilmcnitc and hematite arc referred to as the titanohcmatilcs. Hematites have rhombohedra1 symmetry and corundum structurc. As for the titanomagnetites, thedisordering (Neel)

3.2 Magnetic Minerals

Marine Sedimentary Magnetite

0.2

0.4

temperature varies with x, from -218°C for ilmenite to 675"Cfor hematite (Fig. 3.4b). Compositions corresponding to values of x greater than about 0.8 are paramagnetic at room temperature. Other compositions are ferrimagnetic for values of x in the 0.5-0.8 range and antiferromagnetic for 0 < x < 0.5. Complete solid solution occurs only above about 1000°C and exsolution of the two end members is common except w h a e x < 0.1 or x > 0.9. Titanohematites are associated with some metamorphic and plutonic rocks but almost pure hematite is much more abundant in metamorphic, sedimentary, and igneous rocks. Hematite is a very important as a remanence carrier in both clastic and chemical sediments. The hematite may occur in sediments as dctrital spccularitc or as an authigcnic pigment which grows during diagenesis. Hematites are antiferromagnetic, the weak net moment being due to slight canting of the spin moments (Fig. 3.2). The saturation magnetization is much lower than that ol' magnetite, and the remancnt coercivity is significantly higher.

I

1 0.0

0.6

0,8

33

1.0

Axial Ratio

c. Maghemite (y-Fe20d and Titanornaghemite (xFeTi03.[1- xJFe20,]

Modern Bacterial Magnetite Langth

Leyth

( ~ m )

(A)

\*

Maghemite has the spinel structure of magnetite and the chemical composition of hematite. Maghemite and titanomaghemite are metastable phases, with maghemite and titanomaghemile inverting to hcmatite and magnetite, respectively, above 250°C The disordering (Curie) temperature, is therefore, difficult to determine but is believed to be about 640°C.The magnetic properties arc similar to those of magnetite, with slightly lower saturation magnetization and comparable remanence coercivity. Maghemite is common as a low-temperature oxidation product of magnetile in bolh igneous and sedimentary rocks.

d. Goethite (a-FeOOH)

I

1 0.0

0.2

0.4

0.6

0.8

1.0

Ax~alRatio Figure 3.5 Cornparison of rile and shapc of milrinc sedimmlilry ma:nctitr and modem bacterial magnetite inm clcctron microscilpy (after Vsli r2i a/.. 1987) SP: supcq~arilmilgnrtic: SD: single domain; TD: two domain: MD: multidnm;~in~

Goethite is the most common of the iron oxyhydroxides and is antiferromagnetic with a disordering (Neel) temperature of about 120°C (Hcdley, 1971). Goethite has weak saturation magnetization and very high coescivity (Hedley, 1971; Rochette and Fillion, 1989). It dehydrates to hemalitc at ahoul 300°C. Gocthitc is an important product in low-temperature oxidation, particularly of iron sulfides, and is a constituent of manganese nodules and of the Pacilic redclay facics, where il may precipitate directly from seawater. Lepidocrosite (y-FeOOH) is less common than goethite, has a Nee1 temperature of -196°C. and is therefore not an imporlant rernancncc carrier. hut it dehydrates to maghemite and may therefore indirectly wntribnte to magnetic remanence

34

e. Pyrrhotite (FeS,,

7 Magnetlratwn Proceswl and Magnct~cRopertlcs of Srd~rnmts

where x = 0-0.14)

Fe7S8. the most magnetic pyrrhotitc, is monoclinic and ferrimagnetic, with a Curie temperature oI about 325°C. FeaSla is less important magnetically, being ferrimagnetic in a very restricted tempcrature range (-100-200°C). The saturation magnetization of Fe,S, is about ten times that of hematite and about hall that of magnetite and maghcmite (Clark, 1984). Remanence coercivity is generally greater than that of magnetite and maghemite (Dckkers, 1988, 1989). Pyrrhotite occurs in basic igneous rocks (Soffcl. 1977. 1981) and can grow during sediment diagenesis in certain reducing environments (Kligfield and Channell, 1981; Freeman. 1986). The most common iron sulhde, pyritc (FeS2). is paramagnetic.

f. Greigite (Fe&

3.3 Mqnetitation Processes

35

e temperalurc at which the relaxation lime becomcs large relative t o oratory experimental timc is referred to as the blocking temperature. or a rock or scdiment sample, the grain population will have a blocking tenzptirature spectrum reflecting the compositional and grain size range of the magnetic minerals. For weak magnetic fields, the TRM intensity is pproximately proportional Lo the applied field, although lower cooling tes will enhance the TRM intensity. It is important to note that natural ocking temperature of a grain during igneous cooling will be lower than e laboratory-determined blocking temperature. The diIfcrcnces are ater at lower igneous cooling rates and become smaller when the blocktemperature is close to the Curie temperature (Dodson and McClelland-

b. Chemical Remanent Magnetization (CRM)

Greigite is Ierrimagnetic and carries a stable magnclization in some lake sediments (Snowball and Thompson, 1990; Snowball, 1991) and in rapidly deposited clastic marine sediments (Tric e t a / , , 1991a; Roberts and Turner, 1993; Reynolds et ul.. 15194). Coercivities arc similar to those of magnetite and therefore generally somewhat less than for pyrrhotite. As for pyrrhotite. the Curie temperature is about 320°C and thermomagnetic curves are characterized by an increase in magnetic moment (and susceptibility) as the iron sulfide oxidizes to magnetite.

3.3 Magnetization Processes a. Thennal Remanent Magnetization (TRM) The remancnt magnetication acquired by a grain during cooling through its blocking temperantre is referred to as thermal remanent magnetization. The blocking temperature of a grain is lhe temperature above which the magnetic ordecing in the g a i n is randomized by thermal energy during laboratory heating. The magnetic ordering of the grain will eventually become randomized at any temperature: the time that it is likely to take is referred to as the relaxation time (T,), which is a function of coercive force (H,). grain volume ( v ) , spontaneous magnetization (J,),and temperature (T). T, = To exp (vH, JJ2kT)

where k is Boltzmann's constant and To is the frequency factor (-1W9 s). Thc exponential form of this equation means that, for a particular grain, the relaxation timc increases logarithmically with decreasing temperature.

Chemical remanenl magnctizatinn is produced by an increase in grain . . volume at temperatures below the blocking temperalurc of the grain. The , process is analogous to TRM acquisition hut in the case of CRM the relaxalion time Isee Eq. (3.1)l is increased by change in grain volume ( v ) rather than ambient temperature (T). The logarithmicincrease in relaxation time with grain volumc [cads to the concept of blockirtg volume, which is the critical grain volume at which the relaxation time heoomcs geologically significant. CRM is a very important acquisition process in sediments because magnetic minerals com~nonlygrow authigenically during diagcncsis. Magnetite, hematite, maghemite, goethite, and pyrrhotitc can all grow anthigcnically in certain diagenclic conditions. Alteration during weathering can produce a CRM carried by goethite andlor hematite in a wide variety of rock types.

c. Detrital Remanent Magnetization (DRM)

L.

can align themselves with the ambientfield during deposition. The alignment occurs largely within the upper few centimeters of soft scdiment rather than in the water column (Irving and Major, 1964). Magnetite, titanomagnetite. hematite, and maghemite arc common carriers of DRM in sedimcnts. The lock-in depthui DRM has been estimated hy laboratory experiments (Kent: 1973: Lovlie, 1974; Tauxe and Kent, 1984), by comparing the position of rcvcrsals withoxygen isotope datain cores withvarying sedimentation rates (de Menocal et ul., 1990), and by observation of thedamping of secular variation records (Hyodo, 1984: L,und and Keigwin, 1994). The estimates often Lie in

36

1 Magnetuatlon Processes and Magnellc Properties of Sedments

the 10-2O cm depth range below the sedimentiwater interface The lock-ln depth and w~dthof the lock-ln Lone will depend on the gram slze spectrum. sed~mentationrate, and bioturbat~ondepth.

d. Viscous Remanent Magnetization (VRM) VRM is an important source of secondary magnetization in rocks and sediments. Grains with relatively low relaxation time will tend to hccomc remagnetized in the ambient weak field. The intensity of the VRM has been found both theoretically and cxperimenlally to he proportional to the logarithm of time. and the proportionality constant (rclerred to as the viscosity coefficient) increases with temperature (Dunlop. 1983). Grains with volumes close to the blocking volume (i.e.; close to thc threshold between single domain and superparamagnetic behavior) are likely lo be the most imporldnt carriers of VRM, although multidomain grains can also carry appreciable VRM. For grains with high coercivity, such as fine-grained hematite, the VRM acquired during the Brunhes chron can have high coercivity and blocking temperature up to about 350°C. VRM components of the NRM will generally be aligned with the present Earth's ficld.

Magnetic Pmperlies of Msrinc Sedimen

. Self-Reversal Mechanisms If-revenal is a phenomenon where the remanent magnetization of a le is antiparallel to ihc direction of the ambient magnetic lield at the f remanence acquisition. I1 is now clear that self-rcversal is rare and nomenon is restricted to a icw well-doeumenled cascs, particularly ohematilcs. Self-reversal can occur as a result of magnctostatic raction between adjacent magnetic phascs vith different blocking tematures and usually results from negative exchange coupling between a ongly magnetic. low Curia point phase and a weaker high Curie point ase with moment aligned with the ambient magnetizing field. The phases y be distinct mineral compositions or compositional gradients within ins. If the lower blocking temperature phase has greater volume and spontaneous magnetization than the higher hlocking temperature phase. the net remanent magnetization can be antiparallcl to the ambient field. The best documented cascs of self-reversal are in a dacile containing tilano' hematile in the 0.45 < x < 0.60 compositional range (Uyeda, 1958) and in some oxidircd titanomagnetite composilions (Schult, 197h; Heller and Petersen, 1982). Pyroclastics from the 1985 eruption of Ncvado del Ruiz : (Columbia) are reversely magnrtized, and the self-reversal process has hccn replicated in laboratory-induced acquisition of TRM (Heller er nl., 1986).

e. Isothermal Remanent Magnetization (IRM) This is a high-field magnetization process and therefore docs not generally occur in nature except in the case of lighlning strikes. As the properties of the IRM are useful for determining the magnetic mineralogy and grain size, IRM is produced in the laboratory using a pulse magnclirer, electromagnet. or supcrconducting magnet. The IRM is the remanencc measured after removal of thc sample from the magnetizing field. The intensity of the IRM will increase as the strength of the magnetizing ficld is increased until thesaturationIRM (S1RM)is acquired. Themagnetizingfieldrequired to attain the SIRM is a function of the remanent coercivity (H,) of sample grains.

f. Anhysteretic Remanent Magnetization (ARM) ARM is an artificial remanencc produced in the laboratory by placing the sample in an alternating high field (lypically 100 mT) with a low d.c. biasing ficld (typically 50-100 pT) and allowing the alternating field to ramp down smoothly to zero. ARM is very useful because it has similar characteristics lo TRM, and the characteristics oI A R M can be used to determine grain size and composition of magnetic minerals present in the sample. The anhysteretic susceptibility (k,,,) is the ARM divided by the biasing field.

3.4 Magnetic Properties of Marine Sediments Themagnetic properties ofmarine scdimcnts depend not only on the nature of primary magnetic minerals contributed Crom dctrital or hiogenic sources. or by precipilalion from seawater; but also on the diagcnctic conditions which determine the alteration of primary magnetic phases as well as the authigenic growth of ~econdarymagnetic minerals. The important remancnce-carrying minerals in marinil sediments arc magnetite /Fer04). titanomagnetite, hematitc (a-Fe,O,), maghemite (y-Fe20;), gcothite (aFeOOH). and iron sullidcs such as pyrrhotite (FeS, , ) and greigitc (Fe3Sa). Of the magnetic minerals listcd above, magnetite, titanomagnetite. grcigitc, andgocthite can bepriwmryremawncc carricrsin marine sediments. Hematite, pyrrhotite, and greigite occur as authigcnic secondary minerals formed during early diagcnesis, and goethite can be a product of low-temperature oxidation (weathering), particularly of iron sulfides. 'The magnetic properties of sedimenls Irom the four marine environments lor which there is most information are revieu,cd below.

a. Shallow Water Carbonates The major~tyof shallow water platform limc
38

3 Ma&netnarjon Processes dnd Magnetic Propcnnrq of Sed~mml\

where early magnetizations have been recorded in lhis facics. For example, Silurian rccLslope limestones from Indiana yield magnetite ~nag~letizations which predate compaction-related tilting (McCabe r!r nl.. 1983). Wclldefincd Plio-Plei~toccnemagnetostratigrapllies have been acquired from shallowwatcr limestones and dolomites recovered by drilling in the Bahama Bank (McNeill er nl.. 1988) and from the Mururoa Atoll (Aissaoui et ~ 1 . . 1990). Single domain magnetite of probable biogcnic origin is thought to be the carrier of remancnce in the partially dolomitircd carbonates from the Bahamas, and the dolomitizatioi~process does not appear to remagnctire the carbonates (McNeill. 1990). A similar cnnclusio~iwas drawn froni study of Carboniferous limestones in Wyoming (Bcske-Diehl and Shive. 1978). In the Mediterranean area, thc "suuthern Tclhyan" Mesoznicplatform limestonrs gc~lerallycarry very weak magetizations, which are either too weak for precise measurement or are secondary and due to the weathering of pyrite to hematite and/or goethitc. "Helvetic" limestones from the European Mesozoic contincnlal margin carry a receut VRM with no resolvable primary magnetization components (Kligficld and Channcll, 1981). Thc more marly shallow water Jurassic limestones frnm southern Germally carry a low-coercivity magnetization allributed to dctrital magnctite: however, the purerlimestone facics (bafflestones) carry a secondary magnetization due to both goethite and hemalite produced by late oxidation of pyritc (Heiler, 1978). A magnetite magncti~alionin marly shallow watcr Late Cretaceous limestones from the Munster Basin predates lalest Cretaceous folding and therefore appears 10 be primary in origin (Hellcr and Channell. 1979). In North America, Paleozoic neritic and lagoonal limestone exposed in New Jersey and New York such as the Ordovician Trenton limestone (McElhinny and Opdykc. 1973: McCabe er ul., 1983), the Siluro-Devonian Helderherg Series (Scotese et ul., 1Y82)+and the overlying Onodaga limestone (Kent. 1979) carry a synfolding or postfolding remanent magnetizaLion with a dircclion implying remagnetization duringLate CarbonilerousEarly Permian time (Scotese el al., 1982: McCabe ct GI., 1983). These limestones conlain spheroidalibolryoidal magnetite which grows after pyrite during diagenesis (Suket nl., 1990. 1993), and the parlicular diagenctic conditions which bring about this transformation have been associsted with the migralion of hydrocarbons during Alleghenian thrusting (McCabe er ul., 1983). Comparison of the coercivity of TRM and ARM suggests that the magnetite is mainly in the MD size range (McElhiuny and Opdyke. 1973: Kent, 1979). Curie point determinations and energy-dispersive X-ray analysis indicate that the carrier is a low-Ti magnctite (Scotese ct nl.. 1982; McCabe et al., 1983). Remagnetized Paleozoic limestones from North

3.4 Magncric Propcrtics of Marine Scdimente

:"

39

America and Britain have characteristically "wasp-waisted" hysteresis loops (Fig. 3.6) (Jackson, 1990: McCabe and Channell, 1994). The origin of such wasp-waisted loops can be attributed to mixtures of grains with contrasting ct>crcivities. In North American remagnctizcd limestones, the wasp-waisted characteristic of the loops is observed not only in the bulk rock but also in the "nonmagnetic" residut: allcr magnilic extraction of the large (> few micron) sized sphcroidal/botryoidal magnetite (Sun and Jackson, 1994). The large magnetite grains do not. therefore. appear to contribute to the characteristic wasp-waislrd loop shape, u'hich could be produced by niixing u l SD magnetite with a volumetrically dominant SP magnetite phase (Jackson el a/., 1993: hlcCabe aud Channell. 1994: Channell and McCabc, 1994).

b. Pelagic Limestones and Calcareous/Siliceous Ooze Deep-sea (pelagic) limestones, and thc unlithificd analogues (calcareous ooze) recovered i'rom accan drilling. are the group of marine sediments that have bccn most thoroughly studied from a magnetic viewpoint.. The interest in this sediment type stcms From thcir efficiency as recorders of the ancient geomagnelic ficld and theiruseful~lessas a means of establishing the corrrlalion of polarity reversals to microlussil biozonalions. 'l'hcrc arc many cxamples of the use of this type nf sediment in magnetostratigraphic studies (e.g., Opdykel 1972; Opdyke rt a/.. 19741 Alvarez er nL, 1977: Channell et a/., 1970; I.owric el a/.. 1982; Tauxe r t a/., 1983~). Many pelagic limesto~lesand calca~consiiiliceousoozes have rather simple magnetic properties, exhibiting a single magnetization compunenl carried by lert-imagncticmagr~etife or titanomagne~ite,typically with hystcresis ratios lying in the pseudo-single domitin (PSD) grain size range (Fig. 3.7). These sediments acquire thcir magnetization (DRM) by mechanical rotation of magnctitc grains into line with the ambient geomagnetic field in the bioturbated upper few centimeters oC soft sediment. As discussed above (53.3c), thc depth (bclow the sedimentiwater interlace) and size of the "lock-in zone" in which the grains lose lhcir mohility will depend on grain size, magnctic mineralogy, scdimcntatiofl rate, and the extent ol' bioturhation. Estimates of this depth for pelagic sediments are generally a few tens of ccntimcters, which corrcspond to a few times 10";ears. at typical pelagic sedimentation rates. Sediments which carry a magnetite DRM arc oltcn cxccllcnt recorders of the direction of the geomagnetic ficld at a time close to the time of deposition. Magnetitein pelagic scdimcnts was thought to be entirely detrital (e.g., Lovlic. 1974): however, Curie temperatures of magnetic separates trom deep-sea sediments often indicatc the presence of almost pure magnetite

-300

a06

41

4 Magnetic Properties of Marine Sediments

3 Magncrization Pr~xesscsand Magnetic properties f, sedimrnts

-2W

.ZM)

-100

0

MD 100

200

300

H: Applies Fisld (mT)

Figure3.6 Hvstereris loops from Lim~stonesremagnetizateddurin~~at~ ~ ~ Pcrmiaa time: top. C~rhoniferou% limrrfones from Britain bottom. ordovickn rrom New York State (after .McCabe and Channell. 1994).

~ ~j~~~~~~~~

h

~

j

HcrlHc

f ~ ~ ~ ~ ~ ~ . ~ ~ ~ ~ Figure 3.7 (a) Typical hysteresis loop for white ~ e l a g i hc e s t o n e (MaiolicaFomarion. Italy). (b) HysLeresis racios for pelagic limestones nl the Mlolica Formation, singlc domain (SD), pseudo-single domain. and rnullidomain [MLI) magnctitc fields according lo Day a ul. (1977) (data &cr Channrll nnd McCabe, 1994).

3 Magnetization Procerscr and Magnrric Properties of Sediments

(Fe,O,). As poilited out by IIcnshaw and Mcrrill(l9XO). this is inconsistenL with a lotally dctrital sourcc as detrital source rocks (such as mid-ocean ridge basalts and igneous rocks cxposed on land) contain titanomagnetitc, rather than purc magnetilc. The relatively reccnt discovery that scveral marine organisms. including baclcria and molluscs. precipitate ncarly pure magnetite either extraccllularly or as an integal part of their ~~ietaholism (Blakemore. 1975: Kirschvink andLowcnstam, 1 979;Moskowitz et a1.. 1988. 1994) may providc the alrsurcr Lo this dilcmma. Ma-petite crystals of probably bacterial origin have been observed in pclagic sedimcnts using trilnsmission elcclron microscopy (Peterscn ct a/., 1986: Vali rl irl., 1987: Yamazaki er nl., 1991). 'l'hrse typicallpoctahedral o r parallclcpiped-shaped magnetite grains are usually 0.05 to 0.1 pm across and within the single domain grain size range, zvhich is optimal for retention or magnetic rcmanence (Fig. 3.5). Magnetite of badtcrial origin may fonn at a few centimeters below lhe sediment-water interlace at the transition from iron-oxidi~ingto ironreducing conditions (Karlin r
-

43. (Fig. 3.7), d o not appear to contain maghemite or hematite, hut often contain optically visible pyrite. which readily wcathcrs lo goetliite and finep i n e d hematite, Thc mineral goethite can also play a more primary role in the magnctiralion of pelagic sedimcnts. It is a constituent of manganese and can prccipitale directly from seawater and. logelher with ironbearing clay minerals may be a precursor to pigmentary hematite (in thc red limestones) and to pyrite (in the white Iimestoncs). Neither maghemati=ation of primary magnetitc nor authigenic growth of pigmentary hematite appears to occur in the white pela,oic limcstonrs. The rcmanence carried by pigmentary hematite in tlie red limestones appears. in some cases, to be due to growth O r hematite in lhe uppermost few mctcrs of soft sediment (Channel1 rf al., 1981a). The dominant control o n limestone color and on magneticproperties isprobably organic content, thc burial of organicmatter being largely controlled by scdimenlation rate.

c Pacific "Red Clay" Facies Not aU deep-sea sediments are efficient recorders of the gcomag~xeticfield at the time of deposition. The lackof a prunary magnetizatioi~may hc due lo the magnetic mineraIogy or grain size of the detritus, to adverse diagenetic conditions which result in the alteration ol' primary magnetite. ur to the growth of authigenic magnclic mineral phases. l'he so-called red clay facics occurs over a large part of the midlatitude Pacific (Davies and Gorsline. 1976). I t is dcvoid of calcareous and dliccous microfossils and accumulated below the CCD at ralcs of about 25 cmiMy. In this facies, the primary magnctiration generally degrades at a few meters dcpth helow thesedin~entwater interlace. The boundary hctwcen the primary and stahlc magnetic records often occurs in the later part of the Ciauss chron and coincides closely with tlie latc Pliocene onset of northcm hemisphere glaciation (Opdykc and Foster, 1970: Kcnt and Lowrie. 197.C;Prince c,t at.. 1980). A number of different explanations havc hccn offered to account for the degradation of the primary magnelkation. Kent and Lnwric (1974) considered that oxidation of primary niagnctitc to maghe~niteaccounts for the observed increasc in VKM downmre. The SP-SD grain size threshold for maghemile is greater than that for m a p e t i t , and therefore the oxidation process might hc cxpccled to increase the VRM contribution to remanence. In this scenario. the increased sedimentation rates in the upper part nf the corc cliaractcri~edby stable prin~aryremancncc retard tlie oxidation (due to increased burial of organic marter). Johnson el ul. (1975) alsu atirihutcd the instabilily of the magnetization to the diagenetic growth of maghemite downcore, but considcrcd thc secondary magnetization to be a CRM rathcr than a V R M . Henshaw and Merrill (1980) have suggested that secondary

44

3.5 Magnetic Properties of Terrestrial Sediments

3 Magnetlzatlon P r w a a r , and Magnolac Rapertas al Sedtments

45

acquire VRM. In the Late Cretaceous Gurnigel Flysch, thc VRM d in the Brunhes Cluon domindtcs the NRM such that the primary ization cannot be resolved (Channell et nl., 1979). In turbiditic seences, it is generally good practice to sample thc linest upper parts of ividual flows; however, even in the finer silt llorizons. M D titanomagnedominates the magnelite fraction. Turbiditic deposits are attractive magnetostratigraphic study as their high sedirnentalion rates give high tential magnetostratigraphic resolulion. The high sedimcntation rates may, however. result in a high rate of burial of organic nlattcr, sulfate reduction to sulfide. and the alteration of primary detrilal titanomagnctite to iron sulfide minerals.

ferromangancse oxides and oxyhydroxidcs [such as the jacobsile (MnFezO;) solid solution series] are important authigenic magnetic phases and can carry a CRM which masks the primary DRM. Yamazaki and Katsura (1990) used a suspension method to determine the magnetic moment dislribution and ferrimagnetic grain size in redeposited red clay. The mean grain diameter decreases below I m depth in the sediment. and this is consistent with changes in the frequency dependence of susceptibility, which indicates an increasc in SP grains below this depth. These data imply that the increased importance of VKM downcore may he due to reduced grain size of eolian detrital magnetite in preglacial time (due to less vigorous atmospheric circulation), rather than to downcore maghematization.

.. , .~ *, . ~

d. Hemipelagic Sediments and Turbiditic- 'LFlysch''Deposits

, ,

In areas of high sedimentation rate andlor iestricted bottom-water circulation, the burial of organic matter can result in reducing diagznctic conditions and the formation ol' iron sulfides. Paramagnclic pyrite is by far the most common iron sulfide in scdimcnts, and it can grow by reduclion of primary magnetite. The metastable iron sulfidcs mackinawite /FeS) and greigite (Fe3S,) are, unlike pyrite: capable of carrying magnetic remanencc, as arc some compositions of the more stable pyrrhotitc (FcS,,,). Pyrrhotite coexists with m a e e t i t e and pyrite in young siliciclastic sedimcnts from the Sea df Japan (Kobayashi and Nomura, 1972) and in Mesozoic "helvclic" limestones (Kligfield and Channell. 1981). Greigite in siliciclastic marine sediments from reducing depositionaltdiagenetic euvironmcnts has been recognized in northern Italy (Tric el aL, 1991a). New Zealand (Roberts and Turner, 1993): and thc Norlh Slope of Alaska (Reynolds et nL, 1994). Karlin and Levi (1985) have documenled the progressive dissolution of primary magnctitc, due to reducing diagenetic conditions, in hemipclagic silts and clays from the Orcgon coast and in laminated diatc?mdceous oozes from the Gulf o l California. Sedimentation rates were estimated to be 121 and 135 cmlky, respectively. A decreaseinNRM, SLRM. and ARM intensity occurs in the topmost meter as dissolution reduces the population of linegrained magnetite. Increases in coercivity and rclatiue stability of ARM relative to IRM, from about 1 m depth in the cores to the hasc at 4 m. are interpreted in tcrms of a reduction in mean cffcctive magnetite grain size as dissolution proceeds. A similar scquence of events has bcen ohserved at O D P Site 653, in a high sedimentation rate core from the Tyrrhcnian Sea (Channell and Hawthornc. 1990). Turbiditic "flysch" deposits can carry a primary magnetization; huwever, the principal rcmancncc carrier is usually multidomain (MD) titanomagnctile, hence the remanent cocrcivity is low and the sedimcnts tend to

.

.

.

3.5 Magnetic Properties of Terrestrial Sediments

~

,

,

.

a . Lake Sediments

In early studies of lake sediments, detrital hematite was considered to be the major carrier of rcmancnce on iht. basis of the apparent observation .of the Morin transition at -10°C (Creer er ol., 1972). Later studies showed

.

%thatthe reduction in nlagnetization intensity at this temperature was due t o the rcoricntation of grains by growth oricc cryst~lsnnd that the dominant carrier is fine-grained single domaitl magnetite (Stober and 'Thompson, 1977; Turner and Thompson, 1979). The grainsize of the magnetite ranges from multidomain to superparamagnelic in some Norlh American lake sediments (King ct ul., 1982), detrital sources and bacterial activity being the controlling factors. Bacterial magnetite has also been documented in ,'European lakes (e.g.. Snowball, 1994). Sediments acquire a dctrital remanent magnetization (DRM) by mechanical rotation of the magnetite grains in response to the ambient geomagn e t i c field. Although the D R M may be acquired within a fcw days in redeposited lake sediment (Barton et nl., 1980). observations of magnetic coercivity o l lakc sediment from closti to ihc scdimentlwater interlace gave an estimate of lock-in time for a moderately stable DRM of about 100 years (Stobcr and Thompson, 1977). Lock-in times of a few hundred years may significantly distort the secular variation record particularly if thc lockin proccss occurs over an appreciahle depth range in the sediment. If the lock-in or moment fixing function can be estimated (by expcrimenl) fur a particultar sediment, the secular variation record can be deconvolved by numerical methods (Hyodo. 1984; Lund and Kcigwin, 1994). Comparison of the magnetic properties of natural and redeposited lake sediments has shown that the grains arc probably not fixed in place by dewatering of the

46

3 Magiletiznurm

Magnetic Properties of TcrresL~ialSediments

YroScsses and Magnetic Proprrtirs of Scdimcnts

sediment to a critical level, but possihly by gcls that form in the pores of the wet sediment and are particularly prevalent in organic-rich sedimcnt (Stober and Thompson. '1979). Thc process of sediment drying oftcn disrupts the fidelity or the record; however, this may he due to alteration of remanence-carrying iron sulfides. Greigite is an important carrier olmagnclic rernanencc in lake scdiments (Giovanoli. 1979; Snotvhall and Thompson, 1988: Sliowhall, 1991). Grcigite has coerciuity similar to that ofmagnetile, however it can he recognized by thermomagnetic analysis and Mosshauer spectroscopy. Although greigite is an authigcnic mineral in lakr and marine scdiments. it grows during early diagenesis and can carry a high fidelity record o l the genm~glleticticld (e.g., l'ric er a/., 1991b). ,

~

. .,,

b. Loess Deposits

- , re 3.8 Hystcrcsir loops from Chincsc iocss and i~lterveniogpaleorol (aRrr Hellcr and Evans, 1995).

,' ,

The first magnetic stratigraphies of Chinese loess deposits resultcd in drastic revision of the duration ol' Chinesc loess dcposition, from -1.2 My to -2.5 My (Heller and Liu, 1982. 1984). In addition. magnetic susceptil~ility records definc lithologic loesslpalchsul cycles which ct~rrelatewith glacialinterglacial marine dlHOrecord (Kukla ct a/., 1988: Liu 't nl., 1985). At lirst, it was thought that. the susccptibility llucfuations were controlled by variations in thc concentration of windblown detrital miignetitc. due to increased loess deposition during glacials superimposed on a more constant hackgroui~ddeposition of windblown magnetite (Kuklit et oL, 1988). It is now generally accepted that the enhanced susccptibility in the palcusols (during integlacials) is a result OL authigcnic "ma.getic enhancement" as a result of elevated rain~illllternpcrature.The magnetic enhancement is generally attrihutcd to bacterial production of line-grained SP and SD magnetite during pcdogenesis (Zhuu el ul., 1990: IIcller ef al., 1991: Mahcr and Thompson, 1991,1992: Banerjce ef a/.. 1993; Evans and Hellcr, 1994). although some authors ctmsider maghcmite to bc i~nportnnt(Eyrc and Shaw, 1994). T ~ loess E sedimcnt has higher coercivitics than intervening paleosols (Fig. 3.8) duc lo the greater contribution of liernalite in thc loess ( H d e r and Liu. 1984: Mahcr and Thompson. 1992).

c. Continental Red Beds Continental nonmarine clastic "red hcds" often have stable magnetization carried by two distinct phases of hematite, one detrital and one authigenic (Cullinson, 1965.1974: Purucker ct al., 1980; Tauxe et A/., 19X0).The magnetization of coarser grained (often drab-colored) sandstone units is oftcn dominated by "spccular" hematite, of prcsumed detrital origin, whereas finer grained units (oftcn more reddened) rend to have CRMs dominated

by fine-grained authigenic "pigmentary" hematite. Continental red beds have been important in mapetostratigraphiestudies of the Paleozoic (DiVenere and Opdyke, 1990, 1991a.b), Mesozoic (e.g., Helsley and Stciner, ., 1974), and Cenozoic (Tauxc and Opdykc, 1982); however, the age of the hematite (and its magnetization) has been the subject or considerable debate (Purucker et al,, 1980; Walker er al,, 1981; E. E. Larson er al., 1982; Steiner, 1983; Lovlie cr aL, 1984). Redeposition experiments Of naturally disintegrated hematite-bearing sediments indicate that the hematite grains can carry a detrital remanence (DRM) closely aligned to the ambient field ( T a u e and Kent, 1984; Lovlie and Torsvik, 1984). Characteristic magnetization components in rip up clasts from Triassic and Carbonifcrous red beds are randomly directcd, indicating an carly origin for the magnetization (Molina-Garza el al., 1991; Magnus and Opdykc, 1991). Probably the h o s t convincing evidence for the penecontemporaneous acquisition of redbed remanence is the cxtensive studies of the red beds of the Siwaliksystem I of Pakistan by Johnson e t a / . (1982. 1985). Opdyke et al. (1982). Tauxe and Opdyke (1982). and many others. The Siwalik sediments contain bright red hematite-hearing units which range in age from Middle Miocenc (15 Ma) to Late Pliocene (2.5 Ma). The hemalite magnetizations predate Pliocene-Pleistocene folding. and the Miocene sedimcnts contain no evidence of detrital magnetite. Tauxe et al. (1980) carried out a conglomerate test and demonstrated an early magnetization in red siltstone clasts formed in an overbank environment and subsequently incorporated in the gray sandstone of a river channel. Upon thermal demagnetization, the sedimentary clasts were randomly, but stably magnetized, showing that the CRM directions were acquired early in the history of the sediment. Further evidence that the characteristic magnetization of these beds was acquired

48

3 Md&net?unonProcrscs and Mdgneuc Properties of Sed~rnents

early in their history is the fact that the magnetic stratigraphy from these sediments can he correlatcdto the GPTS derived from the occanic magnetic anomaly record. Independent fission track dates on tuffs in Lhe sedinlentary sequence conlirm the correlation. Tauxe etal. (1 980) suggcst that thc CRM of thcsc sediment5 was acquired within 20 ky or less or deposition. This will not always he the case in reddencd sediments. where authigenic pigmentary hematite can significantly postdate deposition.

Laboratory Techniques

4.1 Introduction Sampling in indurated rocks is usually conducted using a hand-held drill. A sample 5 to 10 cm in length is cored and left standing in thc hole. The sample is then urientcd using a slotted tuhe which fits over the core. Orientation devices usually have aplare on which a sun ormagnetic compass is mounted and aligned with the slotted tube. The compass is lcveled and the azimuth and dip ol the samplc are recorded. A mark is made on the sample along the slot, and the downward direction is marked on the samplc. In volcanic areas, it is usually necessary to take sun compass readings, as local magnetic anomalies may distort the Earth's magnetic field. Most modern orientation devices are equipped for sun compass readings. The direction of a sundial shadow is recorded. along with the precise time of observation. The direction and dip of the bedding at the site are recorded to enable the bedding plane to he returned to the horizontal. Oriented hand-sized samples are olien takcn from outcrop and in this case the dip and strike of a flat surface are marked and recorded. The samples are then returned to the laboratory, where they arc fashioned into a standard size. either cores with diameter 2.5 cm and length 2.3 cm or, alternatively, cubes of 8-15 cm3. The direction and intensity of the natural remnant magnetization (NRM) are measured using a magnelometer. Most laboratories in North America, Europe, and Japan are equipped with cryogenic magnelomelcrs in which the pick-up coils and SQUID sensors opcrate at liquid helium temperature (Goree and Fuller. 1976: Weeks et a]., 1993). Other types of magnetometers such as the rock generator type (Johnson ct al., 1948) or

b .

50

4 Laboratory Techn~ques

fluxgate spinner magnetometers (Foster, 1966) are still used in some laboratories (see Collinson, 1983). The Intensity of the sample magnetic field is measured in three orthogonal directions. This allows the edlculation of the total moment of the sample as well as the declination and incllnatron o i the magnetization relat~veto the orlentation llnc on thc core or cube. The magnetization direction of the sample is then rotated Into eeoeraohir w " r - coordinate5 and rotatcd again, about thc: strike of the bedding, to account for bedding tilt. The NRM of rock and sediment samples is often thc resultant of more than onc magnetization component. Individual components may be acquired at different stages in the history of the sample. Components acquired during deposition of the sediment or cooling of the igneous body are refcrred to as "primary." Those associated with later geological events such as sediment diagenesis, deformation. uplift. and weathering are referred to as "secondary." The objective of this chapter is t o providc a practical guide to methods of determining (1) the componcnt content of a magnetization and (2) magnetic minerals which carry the magnetization components. For detailed discussions of mineral magnetic propertics and of instrumentation, see Collinson (1983). O'Reilly (1984). and Dunlop (1990). Iilformation concerning modes of occurrence of magnetic minerals in marine sediments has increascd considerably in the last few years (Ch. 3) such that an insight into the magnetic mineralogy is an important step in determining the relative age of resolved magnetization components. In magnetostratigraphic studies, il is necessary to establish which, if any, of the resolved magnclization componcnts are primary and can be associated with the depositional or cooling ageof the sedimentary origneous rucks. Estimating the age of magnetization components is a critical stcp in palcomagnetic and magnetostratigraphic studies. Determination of the mineralogy of magnetic carriers dues not generally provide unequivocal evidence for (but is often a guide to) the relative age o l magnclization components. Ficld tests (the tilt test and conglomerate test) (3 5.3) are additional important means of constraining the age of magnetization components. Note that the normal andrcverse directionsrecorded by aparticular component do not necessarily indicate that the component is primary: on the other hand. thc absence of antiparallel reverse directionsmay be an indication that the resolved eomponents are secondary, or influenced by sccundary components.

4.2 Resolving Magnetization Components

4.2 Resolvmg Magnetlzallon Components

,~

. .

' '

'

51

to as the remanent coercivity, or the coercivity of remanence (9 3.1). The mmanence direction of individual grainswill be altered by different external field strengths depending on the "coercivity" of the individual grain, which is controlled by composition and grain size. A natural sample will have a "coercivity spectrum" reflecting the range of magnetic grain sizes and present in the. sample. This dependence of coercivity on com+tion and grain size is utilized in alternating field demagnetization to resolve the magnetization components carried by different grain populations. If the NRM of the sample is a composite of individual magnetization components which have distinct coercivity spectra, the alternating field method will be suitable for resolving the individual components. The normal procedure is to progressively demagnetize the sample by a stepwise increase in peak alternating field-the peak field is smoothly ramped down to zero-in field-bee space in order to avoid acquisition of ARM (Fig. 4.1), and the remainingremanence measured after each step. In a natural sample, grain remanences with coercivity lower than the peak altemating field will be reoriented as the peak field is ramped down, and their contributions to the remaining remanence wiU be randomized (Fig. 4.1). The net effect is to "demagnetize" those grains. As the peak demagnetizing field is progressively raised, a larger proportion of the total magnetic grain population becomes demagnetized and ceases to contribute to the remanence of the sample. There are two principal types of alternating field demagnetizers in current use. For both types, a Mu metal shielded solenoid generates an alternating field, with a frequency of a few hundred hertz, which is smoothly ramped down ro zero from a designated peak alternating field. The sample is either placed in this field in three steps, such that the alternating field acts along the three orthogonal axes of the sample, or rotated in a specially designed tumbler about three (or in some cases two) orthogonal axes simultaneously, within the solenoid. The rotation of the sample within the coil system has the advantage of reducing the tendency for ARM acquisition; however, the observation that spurious remanent magnetization can be acquired along the axis about which the sample is rotated (Wilson and Lomax, 1972) has led to wide use of demagnetizers in which the sample is not rotated and is demagnetized in three individual steps along the orthogonal axes of the sample. The availability of highly efficient Mu metal shields has tended to negate the advantage of rotational methods in reducing the problem of ARM acquisition. For a full description of the design of alternating field demagnetizers, see Collinson (1983).

a. Alternating Field Demagnetization

b. Thermal Demagnetization

The strength of the external magnetizing field whlch is required to reduce the remancnt magnetlzanon ol a population of grains to zero ir referrcd

The time required for magnetic remanence of a gram to decay In zero external field (the relaxation time) is a functlon of ambient temperature,

4 Laboratory Techrnqurq

JL

Peak alternating field Caerclvfty of gram A

2 Resolving Magnetization Components @ . ,!

53

MAGNETITE

Blocking temperature 1%)

HEMATITE

Figun4.1 Decay af alternating field strength during AF dcmagnetkalion. As the experiment isconducted in Mu metal shields in the abencc of the Earth's ticld, the net magnetic rnomcnt of a large population of grains with cocrcivitics less than lhc pcak field is zcro (after Vao der V w . 1993).

grain size, and composition (Fig. 4.2). The blocking temperature (or unblocking temperalure) is the temperature at which the relaxation time of the grain falls within the timc range of laboratory experiments (-101 lo4 s). As the function is exponenlial, the exact length o l time of ihe experiment does not, within reason, critically affect the observcd blocking temperature. Two populations of magnetic grains with dilferent origin, and therefore different composition andlor grain sizc, might bc expected to have distinct blocking temperature spectra. This principle is utilized in thcrmal demagnetization to separate magnetization components carried by different magnclic grain populations. Normal practice is to heal the sample in progressive temperature increments (until the magnetization falls below the magnetometer noise level), allowing the sample to cool in field-free

Blocking temperature (DCI

Figun4.2 Blockin~ublockingcontnurs jrlining time-lrmperaturc points for populations of SD magnetite and hematite (after Pullaiah el ul., 1975) illustraling thal high trmprralures in the laboratory timc scab (say 1 houi) can unblock a thermo\~iscousmagnetizalinn acquired at lower trrnperaturcs on gcologic timc scales (sag 10 My).

space after each heating increment. The remancncc of all grams with blocking lemperatures less than the maxlmum temperature will become randomized as the samplc cool5 m field-free space. The magnelizat~onmeasured after each heat~ngstepcan be used to determine the direct~onsof magnett7atton componcnts with distinct blocktng temperature 5pectra

54

4 Laboratory Techniques

As for alternating field demagnetizers, the advent of highly efficient Mumetal magnetic shlcld~nghasgreatlyimproved and simplied the shielding of thermal demagnetization equ~pment.Most thermal demagnetizers are built inside a long (-2 m) Mu metal sh~eldand a cylindrical furnace capablc of holding the samples lies adjacent to a cylindrical blown-air cooling chamber. The sample boat is pushed from the furnace into the cooling chamber inside the Mu metal sheldlng, and samples can be cooled whlle another set of samples IS being heated. The furnace must be noninductively wound and the ambient field in the coolmg chamber should not exceed a few nT, or a partla1 TRM may b e acquired by the samples.

c Chemical Demagnetization Leaching with hydrochloric acid can provide a means of selective destruction of magnetic grains. The method is applicable t o siliceous rocks, particularly red sandstones, which contain a secondary magnetization due to authigenic growth of fine-grained, pigmentary hematite. The dissolution rate of coarse-grained hematite (specularite) of detrital origin is lower than that of the finer-grained, pigmentary hematite of diagenetic origin, due to the higher ratio of surface arca to grain volume for finer grains. Leaching in 3N to 10N hydrochloric acid, and remanence measurement at certain time intervals during the leaching, provides a means of resolving components with distinct "leaching spectra." Slots are cut in the sample to maximize the s u r f a d area of the sample in contact with the acid.

d. Orthogonal Projections The most widely used method for analyzing changes in magnetization direction and intensity during demagnetization (using any of the above methods) is the orthogonal projection (Wilson, 19hl), also referred to as the Zijdervelddiagram (Zijderveld, 1967).This is asimple but extremely useful means of displaying the change in magnitudc and direction of the magnetization vector during demagnetization. The end points of the magnetization vector are plotted both on the horizontal plane, where the axcs are N, S , E, and W, and on the vertical plane where the axcs are U p and Down (Fig. 4.3). Straight-line segments on these plots indicate that the magnetization vector being removed in this interval has constant direction. This direction can be computed and defined as a component of the magnetization. It is important to note that straight-line segments on orthogonal projections may not always represent single components. They may also represent the composites of two or more components with identical demagnetization (coercivity, blocking. or leaching) spectra. Invariably, in a particular projection, there

, ,

.

,, ~

O l ? l l 5 L I S 4

Peak anernrnlngfield IH) 01ternperalllre

(T)

+

8'

8,'

~

I

Figure 4.3 Equal atca and orthognnill projection of demagnetization data for (a) nonoverlapping, (b) partially overlapping, and (c) overlapping coerdvity or blocking ternpcrarurc spccrra (after Dunlop, 1979).

may be several straight-line segments indicating the presence of multiple componcnts. Curvature of the llnes at the junctlon between two straightline segments lndlcates overlap in demagneti7atlon spectra (Fig. 4.3). The magneli~ationcomponents cannot be resolved if the overlap in demdgnetl-

56

4 Laboratory Techniquev

wtion spectra is too great. If the blocking temperature spectra overlap, this does not necessarily mean that the coercivity spectra will overlap, and the choicc of demagnetization technique will be critical to whether the component content of the magnetization can be resolved. The points on the orthogonal projection which define. straight-line scgments are commonly picked by eye. Thc componcnt direction associated with a particular straight-line segment can then be computed by a "lcastsquares fitting" technique (e.g., Kirschvink, 1980). This method gives a maximum angular deviation (MAD) value which provides an estimate of the goodness of fit to the straight line to the points on the projection. MAD values greater than -15" usually indicate that the component h poorly defined.

e. Hoffman-Day Method In the special case in which three magnclization components with ovcrlapping demagnctization spectra are present, the Hoffman-Day method (Hoffman and Day, 1978) may help to rcsolve the component directions. The mcthod involves plotting subtracted vectors on an equal area stereographic projection, at each stage in the progressive demagnetization process (Fig. 4.4). In a demagnetization interval characterized by two overlapping dcmagnetization spectra, the subtracted vectors will plot on a great circle between the two component directions. If three overlapping demagnetization spectra arc present (Fig. 4.4). the intersection of the two great circles can givc an estimate of the direction associated with the component (B) with intermediate blocking tempcrature, even though this component direction is not directly resolved by the demagnetization process. Orthogonal projection in this case would not resolve the intermediate component.

f. Remagnetization Circles If two magnetization components with partially overlapping demagnetization spectra are present in the samplc, the resullant vector will tracc a great circle in stereographic projection as the overlapping interval is demagnetized. If one o l the two components has dispersed directions, the great circles from individual samples will converge to give the direction of the undispersed component (Fig. 4.5). The method is applicable in the special case where one of the components is dispersed. If both components have uniform directions, thcre will be no appreciable convergence of the great circles, and the method cannot b e used on its own. The dispersion of one of the components can occur in a number of situations. For example, thc

,

Figure4.4 Uiffcrcncc%,cctorpaths cclrrespc>ndingto (a) nonorerlapping. (b) partially overlapping, and (c) ovcrlappinp corrrivity or blocking temperature spectra [aitcr Hoffman and Day, 1978).

primary component in conglomerate clasts will be dispersed relative to the secondary component, which postdatcs conglomerate deposition. In this case the secondary componcnt could he resolved by remagnetization circlc analysis. The mcthod can also be applied in circumstances in which one of the two components postdates and the other predates the folding. In theory, both components can be resolved by applying thc mcthod both before and after tilt correction. Remagnetization circlc analysis is most often used in conjunction with individual component directions resolvedfrom orthogonal projections. If curvature of lines on orthogonal projections precludes the resolution of individual components for some but not all samples from a site, remagnetization circle analysis of these diiection trcnds can augment and improve the estimate of the site mean direction dcrivcd from the orthogonal projections, using the procedure o l McFadden and McEIhinny (1988).

58

4 Laboratory Techniques

'

Fqure 4.5 Remagnrli~atiancircles obtained hy AF demagnetization of a two-component magnetization in which ilna component is dispersed. Dashcd scgments are grrat circle extrapolations of thc data (after Halls. 1976).

P, 80 = ( d 2 s i n h ( ~ ) ) e " sin ~~~ 0 '68. (4.2) The parameter K determines the distribution of the points and is called the "precision parameter." If K = 0 then the points are distributed evenly over the sphere and the directions are random. It the values of K are large, the points are clustered tightly about the true mean direction. If a group of points representing directions of magnetization are distributed on a suhere, the best estimateof the mean direction of magnetization (the cente; of these points) can be found by the vector addGion of all directions. The direction of magnetization of a rock is given in terms of the declination D mcasurcd over 3606 clockwise from true north and the inclination I measured positive (negative) downward (upward) from the horizontal. The unit vector representing this direction can be determined by its three direction cosines: 1 = cos D COF I north component east component

m

down (up) component

n = sin I.

sin D cos I

=

The resultant direction of Nsuch direct~on?may be determined by summlng the direction cosines and is given by:

Y

=

C m,

IIR

,=I

The techniques outlined above are designed to isolate individual components of magnetization. An individual component is often combined with others from the same horizon or from a particular sampling site. Thesc vectors are rcpresented as unit vcctors and often displayed as points on an cqual area (stereographic) projection, representing points on a sphcre. Fisher (1953) developed a method for analyzing the distribution of points on a sphere. Hc predicted that thcse points would be distributed with the probability density P given by

z h

Z = 1IR

n,.

,=I

The vector sum will have a length R which is equal to or lc$s than the number of observation and is gwen by:

The declination and inclination of the vector mean direction R are given by:

zm , / z N

Tan D R = where $ I S the angle between the dircct~onof magnetization of a sample and the direction (rl. = 0) at which the density of the observations is a maxunum. If thc density of the points is axially symmetncal about the true mean direction. with the points being distributed through 360°, then the probability of finding a point between @andB SO is:

+

h.

n,

,=I

1=l

i\'

Sin IR

=

llR

n,. ,=I

The best estimate k of the precision parameter

K

for k > 3 is:

4.4 Practical Guide to the Identi6catim of Magmetic Minerals

k

=

(N

- 1)I(N - R).

(4.7)

The angle from the true mean within which 95% of the directions will lie is given approximately by:

hs= 1401(fi) degrees

(4.8)

This value represents the angle from the mean beyond which only 1120 o l the directions will lie. Fisher (1953) has shown that the true mean direction of a group of (N) observations will lie within a circular cone about the resultant vector R with a semiangle u for a certain probability lcvcl(1 - P ) when k > 3. P i s usually takcn as 0.05 and if a i s small, the circle of 95% confidence is given by: ass = 1 4 0 / ( 2 m ) .

(4.9)

Thevalue of uysis oftcn uscd to determine whether twosetsof directions differ from oncanother. Ittheay5valuesoftwo mean directionsof magnetization do not overlap themean directions of magnetization areusually believed tobe significantly different at thc 95% confidence level. A more rigorous test of whether two populations are different was proposcd by Watson (1956a). Fisher statistics are almost universally used in paleomagnetic analysis; however, the Fisher distribution is not ideally suited to paleomagnctic data for two reasons. Firstly, paleumagnetic directions are not always symrnetrically distributed about the mean direction: elliptical distributed data sets are commonly observed. Secondly. paleomagnetic directions are often bimodal comprising both normal and reverse polarity directions. The Bingham distribution (Bingham, 1974; Onstott, 1980) is bettcr suited to paleomagnctic data as noncircular (elliptical) and bimodal data sets are allowed. In this treatment, the "orientation matrix" (computed as the normalized matrix of sums of squares and products) is similar to thc moment of inertia matrix with unit mass assigned to each data point on the sphere. The largest (principal) eigenvector provides a very satisfactory estimate of the mean direction for thc population: however, the estimation of confidence limits about the mean is more problematic, resulting in the low popularity of this method in paleomagnetic analyscs. Tauxe ct al. (1991) advocated the use of "bootstrap" methods to calculate the confidence limits of paleomagnctic mean direction. The advantage of the method is that less stringent assumptions are made concerning the nature of the distribution (the probability density function). The disadvantage is thal it is not suitable for data sets smaller than "a few dozen." The method is therefore inapplicable in cases where less than -30 magnetization components are resolved (samples collected) at a paleomagnetic sampling site. The bootstrap m e l h d involves random sampling of the data set (population: n) to produce a large number (-nZ) of pseudo-data sets each with

/

61

the same population (n) as the original data set. Obviously. some data points may occur more than once in a given pseudo-data set. The parameter of interest, such as the mean direction, is calculated for each pseudo-data set, and the variability is used to calculate the confidence limits. For each pseudo-data sct, the normal and reverse modes are separated about thc normal to the principal cigcnvcctar. The mean direction for each mode in the pseudo-data set is estimated as the sum of unit vectors, as in the Fisher procedure [Eq. (4.3)]. The mean of the bootstrapped means, the principal eigenvector of the orientation matrix derived from the bootstrap means, and the mcan of the original data set should b e very nearly the same. The confidence circle (ays)associated with themcan is thcn estimated from the bootstrap means assuming that they have a Kent (1982) distribution, which is the elliptical analogue of the Fisher distribution.

4.4 Practical Guide to the Identification of Magnetic Minerals Because magnetic mineral phases tend to have characteristic modes of occurrence, the identification of magnetic temanence carriers is a useful clue to the age of magnetization components. The following is designed to provide a guide to the identification of common rcmanence-carrying magnetic minerals. The focus is on convenient methods which are useful in sediments wherc the magnetic mineral grains are well dispersed and in low concentration. Most of the methods utilizc instmmcnts which are commonly available in paleomagnetic laboratories. In weakly magnetized sediments, the characteristics of artificially imposed remanences (such as ARM and IRM) arc very useful for determining the magnetic mineralogy of the sample. However, in order to determine the carriers of NRM components, the characteristics of the NRM itsellmust bc determined as fully as possible. a. NRM Analysis

As deycnhed above, progresswe demagnet~cdtionis the standard procedure for determining the component content of the NRM. Thermal and alternaling field demagnetization characteristics ind~catethe blocking temperature and coerciv~tyspectra, respectively, of the NRM componentq. Blocking temperature is a function of compos~t~on and grain s i x of the magnctic carrier, For SD grains of constant composition, the blocking temperature will be proportional to grain s i ~ e[Eq. (3.1)]. Due to the large gram slze of hematite at the SDIMD threshold (-2 mm), thc SD range for hematite covers most of the remanence-carrymg gram sizes encountered in aediments. Thcrcforc, in the case of hematite, blocking temperature is a useful guide to grain sue. For titanomagnctites (xFeiTiOd.(1-x)Fe,04) the block-

.

63 ing temperature is a function of composition and grain size. ASx decreased Curie point increases almost Linearly (Fig. 3.4), and as blocking temperature cannot exceed the Curie point, the available blocking temperature range will increase as x decreases. For SD grains of uniform shape and composition, the blocking temperature will be a function of grain shape as well as grain volume. In the MD range, the rapid decrease in coercivity with increasing grain size might be expected to give rise to an accompanying decrease in blocking temperature. The blocking temperature range in itself is not generally diagnostic of magnetic mineralogy, other than the fact that the maximum blocking temperature cannot exceed the Curie or Nkel point of the mineral composition. In the case of minerals with relatively low Curie or Nee1 points, such as geothite (-12O0C), the blocking temperature range is rather diagnostic because it is so restricted. Mineralogical changes as a result of heating during thermal demagnetization can obscure the blocking temperature spectrum. For example, maghemite alters to hematite at about 3WC, iron sulfides will oxidize in air to form magnetite at about 300°C and hematite tends to convert to magnetite at temperatures above 600°C. These phase changcs, if identified as such, can be useful as a means of identification of magnetic minerals. The coercivities of remanence of hematite and geothite are typically about two and three orders of magnitude greater, respectively, than that of SD magnetite. Coercivity of remanence in naturally occurring hematite and titanomagnetite is largely a function of grain volume in the case of hematitewandgrain shape and volume in the case of titanomagnetites. For magnetites, coercive force (H,) increases with x and with decreasing grain size in a wide MD grain size range (0.1-1000 pm) where H, is proportional to d-'', with crushed grains kdving higher coercivity than hydrothermally grown or naturally occurring crystals (Heider et al., 1987) (Fig. 4.6). Below about 0.05 pm, the coercivities will decrease sharply close to the SP-SD grain size threshold (see Maher, 1988). Although demagnetization of remanence carriers from AF or thermal demagnetization data is rarcly unequivocal, combinations of these data from the same samples can be fairly diagnostic. For example, in Fig, 4.7, three components, A, B, and C, are apparenl in the orthogonal projection of thermal demagnetization data. The blocking temperature spectra of components B and C overlap. In the A F demagnetization diagram for a specimen from the same sample, only components A and C a r e apparently resolved. Component B has low blocking temperature (2UO-300°C) and high coercivity (>SO mT).In view of the discussion above, fine-grained hematite is a possible carrier for this component. This is consistent with the red pigment of the limestone sample and suggests that this component is a secondary component of diagenetic origin. Component C has a maximum

. I

MAGNETITE

,

Crushed grains

Grain diameter (microns)

gura 4.6 Coercive force as a function of grain diamctcr for magnetite (after Heider er aL, 1987). The apparent higher coereivities of crushcd grains arc attributed to their stressed sratr and adhcring liner fragmenis.

N. Up Thermal dernagnetlzatlon

..

HamVsn A A

a

I

ConponentA Component 0 ComponentC

Figure 4.7 Orthogonal projection nf a threr-component magnetization (from a pink pelagic limestone) illustrating conuastinp response in thermal (tcmpcraturcs in ' C ) and alternating 6eld (peak ficlds in mT) demagnetization (after Channrll et ol.. 1982a).

64

4 Laboratory Teciuuques

ff

blocking tempmature of over 600°C and is therefore carried, at I<& in part, by relatively coarse-grained hematite. The coerciviry spectrum of component C suggests that a low-coercivity mineral (magnetite) contributes to this component. Component A has low blocking temperature and low coercivity, and the in situ direction is close to that of the present field; therefore the component is a recent (probably viscous) overprint. It is important to note that the more primary magnetization component is not necessarily the final component revealed by the demagnetization procedure. In the example mentioned above, the secondary component C has the highest coercivity of the components present in the sample.

4 4 Practical Gulde to the Identlhcatlon of Magnetlc M ~ n r r a l r

Js

Hematite at low

65

Magnetite an0 unsaturated

b. J,/T Curves The change in saturation magnetization with temperature is an important indicator of magnetic mineralogy, becausc it is independent of p a i n size and shapc. The tempcrature at which the saturation magnetization disappears is the Curie or Niel tempcrature, which is the temperature at which thermal energy precludes spin moment alignment. This temperature is characteristic of the magnctic mineralogy. Most Curie balance systems are nor sufficiently sensitive to determine Curie or Necl points from weakly magnetized sediments without extraction and concentration of magnetic mincrals. Magnetic extraction procedures are inevitahly grain size and compositionally selective, and therccore the JJTcurves may not bereprcsentative of themagnetic mineral content of the sample. Significant improvements in Curie balance sensitivity have, however. been achieved by sinusoidally cycling the applied field and making continuous: drift corrections through Fourier analysis of the output signal (Mullender e t a / . , 1993). Horizontal balance systems are generally preferrcd to vertical systems, as they reduce weight loss problems arising from dehydration of clay minerals. For tilanomagnetites and titanohematites, Curie and Niel temperatures indicate the composition ( x ) (Fig. 3.4). and in fidvorable circumstances Curie temperatures corresponding to several magnetic phases can be observed (Fig. 4.8). Alteration of magnetic minerals during heating may obscure the Curie or Nee1 point of the originaI mineral and produces irreversible curves. This irreversihility can be useful for detecting the presence of maghemite, which converts to hcmatite at about 300°C (Fig. 4.8), and for determining the presence of iron sulfides, which alter to magnetite when heated in air. Ideally, irreversible curves should be rerun in vacuo or in an inert gas atmosphere.

c. IRM Acquisition and Thermal Demagnetization of 3-Axis IRM The shapc of the curve of acquidtion of IRM wlth increasing magnet~zing field is a measure of the coercivity spectrum of magnetic mlnenls present

.

~

''.

.&,

,

,,.:

Figure4.8 Typical saturation magnrti~ilti<m-trmp~ratu~e (Js'T) cunzcsfor vari~usminzrological compositions (after Piper. 1987).

in the sample (Dunlop, 1972). Thc common magnctic phases present in sediments. SD magnetite, MD magnetite, fine SD hematite? coarse SD hematitc, and gocthite have rather characteristic and distinct coercivity spectra and IRM acquisition curves (Fig. 4.9). The intensity of thc SlKM and the magnetizing field at which the SIRM is acquiredvary withmineial-ogy and grain size. Thermal dcmagnclization o l a composite 3-axis IRM (Lowrie. 1990) is verv useful for determining magnetic mineralogy. An IRM is applied sequentially along three orthogonal axes of the sample using decreasing magnetizing fields. typically 1 'I'0.4 , T, and 0.12 T'. Thcrmal demagnetization of the three orthogonal "soft," "medium," and "hard" IRMs can sive insights into thc magnctic mineralogy (Fig. 4.10). >

~

~~~~

d. Hysteresis Loops The advcnt nf thc altcrnating gradient forcemagnetometer (p,MAG) allows hysteresis loops to be generated for small (-50 mg) samples, even for weakly magnetized sedimcnts. Saturation magnetization ( M , ) ; saturation remanence (M,,), and coercive force (H,) can hc detemincd rrom the loops, and back field experiments on the pMAG can yield values of remanent coercivity (IT,,) (Fig. 4.11).For samples conlainingmagnetite only, hpsterc-

sis Parameters determined in this fashion can be plotted on a hysteresis ratio plot to estimate the domain state of the magnetite population (Day ef al., 1977) (Fig. 4.12). e. AF Demagnetization of IRM and ARM

., '

bility to AF demagnetization of thermal remanent magnetization (TRM) In SD magnetite increases with decreasing magnetizing field, whereas thc converse is true for MD grains. This observation is not strictly a domain state delimiter: however, it is a uscful grain size parameter (Dunlop, 1986). In the modified version of this test (H. P. Johnson et 01. 1975), it is assumed that the stability of strong field TRM will be similar to that of SIRM and . that the low-field TRM will he similar to that of ARM. The .... stahilitv , of -- -- - ~ - ~ odified test compares lhe stability of AF demagnetization of SIRM and RM, is very simple to apply, and avoids the need to heat thc sample, which may produce mineralogical change. According to the H. P, Johnson et al. (1975) test. if the ARM is more stable to AF demagnetization than ~

Amodification of the Lowrie-Fuller (1971) test introduced by H. P. Johnson eta!. (1975). provides a useful means of determining the relative importance of SD and MD grains in magnetite-bearing rocks and sediments. The original Lowric-Fuller test is based on the observation and theory that the

4 hactlcal F u d e to the Identmcatmn of Magnet~cM~nerals

Detrital level containing MD magnetite

* E

0.6

' l ' l ' l ' l F l r

2

z*

r

,

,

,

!

,

I

T

I

-SD 0.5

-

1

PSD

s

?

!

69

0.4

0

1

-

t

-

.7

0.2

-

r

0.1 i

z -0.6

-0.4

-0.2

0

0.2

0.1

0.6

O

Magnetizing Field (T)

1

1

2

3

4

5

M"

6

1 7

HrcHc

4.12 1Iystzresia ratios plotted o n a Day PI nl. (1977) diagram, illustrating a mixing line containing PSD and hlD magnetite (after Stoncr er a / , IYY5a).

Background sediment containing PSD magnetite

t ' ~ ' ~ ' ~ ' ~ " ' I " ' ~ ' ~ ' ~ ' ~ ' i malizedmoment axis) of the IRM acquisition and IRM demagnetization Ik,

Magnetizing Field (T) Figure 4.1 1 Hysteresis loops ior Uppcr Pleistocene detrilal laycrs from thc Labrador Sea utmtaining hll) magnetite and inlcrvening background sedinlenl containing PSD rnagnctitc (after

Slrmrr rt a/.. 1YY5a).

SIRM. the magnetite is considered to be predominantly in the SD grain size range (Fig. 4.13). Cisowski (1981) has shown that strongly interacting SD grains respond to the test, but that noninteracfing or weakly interacting SD grains d o not give the predicted response. Cisowski (1981) proposed a test for the degree of interaction based on the point of intersectiod (on the

P. Low-Temperature Treatment of IRM ansitionsin the magneticproperticsofmagnctite,pyrrhntite,and hcmatite ur at low temperatures and they provide a potential means of magnetic neral identification. Magnetite exhibits a crystallographic phase transition m cubic to monoclinic at 110-120 K. Associated with this transition. the tropy constant goes through zero as the easy axis of magnetization gesfrom ['IM)] to [I 1 I] (Nagataetul,, 1964). Thc so-callcdVcrwcy transin in magnetite can be recorded by a decrease in intensity of an IRM imat low tempcraturc as it is allowed to warm through the transition rature or alternatively as an IRM imposed at room temperature is owed to cool through this temperalure (Fig. 4.14). Thc transition has n clearly observed in the NRM iron ore samples (Fuller and Kobayashi, 7) and the SIRM of Eurilpean Mesozoic limcstoncs (e.g., Hcller, 1978; uritsch andTurner, 1975; ChanneU and McCahe, 1994) and deep-sea sedits (e.g., Sloner et ul,, 1996. Fig. 4.14). The transition should be manifest relatively sharp change in magnetization intensity and should not he used with thesteady decrease in low-temperature IRM intensity due to ocking of superparamagnetic grains. It was considered that the Venvcy sition v a s restricted to M D grains (Fuller and Kobayashi, 1967); how-

70

4 Laboratory Techniques i

-

IRM dsmagnct#rat,an

--O-

ARM drmagnehzation

a ~ o l ~ cl~mestone a (Italy)

0

0

40

60

80

100

Applied Field (mT)

Figure 4.13 ' l l c point of intersection of LRM acquisition and IRM ltcmagnetizntion fbr a sample of white pelagic limcsfonc is close to 0.5. consistent with noninteraclint. SD magnetitc. ~ c c b r d i nLo~ ~isnwhki(1981). nonintcracting or weakly interacting SD pbs do nit give the predicted response to the Jcrhnson et ol (1975) tcsr hased on the relative AF slability ol ARM and IRM.

ever, it is now clear that the transition occur7 in near-stoichiometricmagnetite crystals for all skes above 37 nm, below which the transition is maskcd by superparamagnetic unblocking (Ozdemir et al., 1993). Small deviations from stoichiometry in magnetite, including oxidation to maghcmite, have the effect of suppressing the Verwey transition (Ozdemir et al., 1993). The Morin transition in hematite at - 10°C duc to a change in crystallographic orientation of the. spin axis, is more elusive. I t has been observed in iron ore samples (Fuller and Kobayashi, 1967) but not, as far as we know, in sediments. Apparent observation in lake sediments (Creer et al., 1972) has been reinterpreted as duc to disruption of grain orientation as a result of ice formation in unconsolidatedsediment (Stober and Thompson, 1979). The pyrrhotite transition at 34 K has been recognized in igneous and metamorphic rocks (e.g., Dekkers eta!., 1989).

g. Continuous Thermal Demagnetization of IRM Continuous measurement of remanence during heating has several advantages over stepwisc thcrmal demagnetization. There is less likelihood of mineralogical changes, and no PTRM or VRM will be acquired by the

' A 1" " , 8

0

20

8

8

,

40

,

'

60

80

Temperature

100

120

140

160

(OK)

Figure 4.14 Veswey transitions iil magnetitz-bearing Upper Plcirlr~ccnesodirncnts from the Labrador Sea. Coarser grained detrital layers with higher proporlion ol MD magnetite show larger remanencc drop than PSDdorninatcd background sadiments. IRM acquired in 5MI mT field at 10 K (anrr Stoner rr a/., 1995a).

samples. Cryogenic magnetometers havc no1 been used on a routine basis r continuous thermal demagnetization due to problems of constructiun a heating system that docs not interfere with the SQUID sensors. The most commonly used magnetometer for continuous lhemal demagnetization is a modificationof the Digico magnetometer (e.g.. Heiniger andHcller, 1976). The loss of sensitivity of thc larger fluxgate ring, which provides room for the furnace, relative to the room temperature ring (-factor oi 10) is such that, for most sediments, the demagnetization experiments must be carried out on the stronger IRM rather than on thc NRM. Continuous demagnetization of superimposed IRM components can be usefuI for determining the maximum blocking temperatures of two IRM components.

Initial susceptibility is often routinely measured after each measurement step during thermal demagnetization in order to monitor changes in magnetic mineralogy during heating. Thc maghemite-to-hematite transition at about 300°C produces a decrease in susceptibility which is a useful means of identifying the presence of this mincral. The susceptibility of maghemite is 2 to 3 ordersof magnitude greater than that of hematite. This susceptibility

72

4 Laboratory Techniques

decrcase will bc masked if iron sulfides or titanohematite are present, because hcating in air oxidizes the iron sulfidcv and titanohcmatitcs to magnetite, which has high susceptibility, comparable to that of maghemite. The formation of magnetite during the thermal demagnetization procedure can result in the development of spurious magnetizations. Note that the most common iron sulfide, pyrite, is paramagnetic and cannot therefore carry remanence.

idue ultrasonically in a 4% solution of sodium hexametaphosphate to lsperse the clays. (vi) Extract the magnetic fraction using a high-gradient magnetic separation technique (Schulze and Dixon. 1979). or alternatively, pass the solution (several times, if necessary! slowly past a small rare

?%. ,;?@,

*. ,#~ ' ah

,

*.

An additional method for monitoring the change in mineralogy during heating involves the change of remanence coercivity (H,). The remanence coercivity of maghemite (and magnetite) is 1 to 2 orders of magnitudc less than that of hematite, The maghemite-to-hematite transition a t about 300°C will be accompanied by a change in remanence coercivity. The measurements of H,, can be carried out after each hcating step with an alternating force gradient magnetomcter (@MAG)but the measurement procedure is considerably slower with more standard paleamagnetic equipment. For example, after each hcating step, the sample is given 1 T IRM, and a back field IRM (antiparallel to the 1 T IRM) is then imposed in steps of 0.1 T. The back field corresponding to zero IRM corresponds to the rcmanence coercivity (H,,). The increase in remanence coercivity in the 300-40O0C temperatwe range has been cited as evidence for the presence of maghemitc in red peIagic limestones (Channel1 et al., LY82a).

j. Separation Techniques Magnetic mineral separation techniques are invariably selective, and not fully representative of the grain size and composition of magnetic minerals present in the sample; however, magnetic separation may be necessary for SEMITEM. X-ray, Mossbaucr, or chcmical analyses. The importance of fine SD grains as remanence carriers emphasizes the necessity of making the separation tcchnique as sensitive as possible to the fine grains. As grain shape has become an important criterion for distinguishing detrital and biogenic magnetite, separation procedures to prepare representative extracts for SEM and TEM obscrvation havc become more important. A recommendcd procedurc which has been successful for extracting magnetite (including the SD fraction) is as follows: (i) Crush the sample (ifnecessary) in a jaw crusher with ceramic jaws. (ii) Use a mortar and pestle to produce a powder. (iii) Dissolve carbonate with I N acetic acid buffered with sodium acetate to a p H of 5, changing the reagent eve? day until reaction ceases (sevcral weeks). (iv) Rinse the residue with distilled water. (v) Agitatc the

5.2 Pol'adtyZone and Polarity Chmn Nornedelaturc'

75

vestigated. For instance, the North American APWP crosses the equator Lower Paleozoic time, introducing some ambiguities as to which is the orth polc and which is the south pole, particularly in regions where there as been largescale local tectonic rotation. In practice, the polarity can be etermincd by following the motion of the pole sequentially back in time from the present. The terminology of magnetic polarity stratigraphy has been codiiied by e IUGS subcommissionon the Magnetic Polarity Timescalc (Anonymous, 979). This body introduced the following uscful definitions.

Fundamentals of Magnetic Stratigraphy

1. Magnetic stratigraphy or magnctostratigraphy-the

element ofstraals with the magnetic characteristics of rock units. tigraphic classification-the organization of rock strata on variations in magnetic.character. 3. Magnetostratigraphic units (magnetozoncs)-bodies of rock strata magnctic characteristics which allow them to be differentited from adjacent strata. tigraphic polarity classification-the organization of units bascd o n changes in the orientation of remanent strata, related to changcs in the polarity of the Earth's

5.1 Principles and Definitions The purposc of magnetic stratigraphy is to mganize rock strata into identifiablc units based on stratigraphic intervals with similar magnetic characteristics. Shorl-period behavior of the magnetic field (secular variation) can be useful for correlation, particularly in Holocene lacustrine sediments and arcl~aeomagneticstudies (Ch. 13). At the other end of the geologic time scale, palEomagnetic pole positions have been used as a means of correlating Precambrian rocks (Irving and Park, 1972). Relative geomagnetic paleointensities as well as a wide range of nondirectional magnetic properties can be used for correlation (Ch. 14). The magnetic characteristic most often used, however, is the polarity of magnetic remanence. The polarity is said to be "normal'" (north-seeking magnetization givcs a northcrn hemisphere pole, as today) or '"revcrse" (north-seeking magnetization gives a southern hemisphere pole). When correlated to the geomagnetic polarity time scale (GPTS), the polarity record observed in stratigraphic sequcnces allows them to be placed in a chronostratigraphic framework. The most imporlant geomagnetic property for stratigraphic purposes is aperiodic polarity reversal of the gcomagnetic dipole ficld. The direction of magnetization of a rock is by definition its north-seeking magnetization, and as explained above, the magnetization can usually bc designated either normal or reverse polarity. A problem may arise in Early Paleozoic or older rocks, or in displaced terrains, when it is unclear in which hcrnisphere the rocks acquired their magnclization. An ambiguity may be introduced as to the true polarity of the rock. In this case, an apparent polar wander path (APWP) musl be established for the individual tectonic unit being 74

5. Magnetostratigraphic polarity reversal horizons are surfaces or vcry

thin transition intervals in the succession of rock sirata, marked by changes in magnetic polarity. In practice, the term magnetostratigraphic polarity reversal horizon may be allowed tostand for a finite transition interval. Less only, the polarity change takes place through a substantial interval of , and the term magnetostratigraphic polarity transition zone should tostratigraphic polarity reversal horizons and magnetostratiaphic polarity transition zones may be referred to simply as polarity versa1 horizons and polarity transition zones, respectively,if in the c o n t ~ x t the reference is to changes in magnetic polarity. Polarity ersal horizons or polarity transition zones provide the boundaries for arity stratigraphic units. 6. Magnetostratigraphic polarity units (or zones) are bodies of rock strata, in original sequence, characterized by their magnetic polarity, which allows them to be differentiated from adjacent strata. .!*,$*.

..%. .". * :;

5.2 Polarity Zone and Polarify Chron Nomenclature

, ~,

The basic magnetostratigraphic unit is the magnetostratigraphic polarity zone (i accordance with the nomenclature recommended in the Interna-

.

77.

tional Straligraphic Guide, pp. 8-14, see Anonymous, 1979). A magnetostratigraphic polarity zone may be referred to simply as a magnetozone, if in the context it is clear [hat the reference is to magnetic pvlarity. Polarity zoncs are bounded above and below by polarity reversal horizons or polarity transition zones (sce definition 5, above). Polarity zones may be of different rank. Magnetostratigraphic polarity zones may (1) consist of strata with a single polarity throughout or (2) be composed of an intricate alternation of normal and reverse units or (3) be domiw~ntlyeither normal or reverse magnetozones, but with minor subdivisions of the opposite polarity. Thus a zone of dominantly normal polarity may include lower rank units of reverse polarity. The words normal and reverse, where it is possible to indicate the polarity unambiguously. may be included as a part of thc term name as an aid lo clarifying its significance. Howcver, as mcntioned above, in premiddle Paleozoic rock sequcnces. it is somctimcs difficult to determine what is normal and what is reverse. Also, in rare circumstances, excursions of the paleofield may provide polarity units that cannot be tlassified as normal or rcversc. but may be designated as intcrmediate in direction. The end ol a successful magnctostratigraphic survey is the idcntification of a series of normally and reversely magnclized polarity zones of varyingJength arranged in stratigraphic ordcr, At this stage, it is useful to give each magnetozone (polarity zone) a numberllettcr and geographic designation, such as, from base of the section: San Pcdro N1, R1, N2, R2, etc., or Gubbio A-, B+, C-, D+, erc. At Gubbio (Alvarez et aL, 1977; Napoleone et ab, 1983). polarity zones were designated in ascending alphabetical order, normal zones distinguished by "+" and reverse zones by "-". The thinner polarity zones are given lower hierarchy and are designated as subzones. At Gubbio, subzones within D+ are labeled D l + , D2-, D3+. The labeling of polarity zones (magnetozones) in individual sections (Fig. 5.1) is useful as it is agnostic and separates the observations from the correlation to the geomagnetic polarity time scale. It is a mistake to confuse polarity zone labels with oceanic rnagnctic anomaly numbcrs; for example, a polarity zone should not be referred to as "anomaly 34." Oceanic magnetic anomalies and polarity zones are distinct phenomena, and they should not be confused. They may, however, have a common cause and be time correlative.

MAGNETIC POLARITY ZONES

197

reverse

Figurn 5.1 Magnetic stratigraphy of the u p p a Cretaceous Scaglis Rossa F m at Gubbio The VGP latitude 6 presented for each sample. Thc polarity z o n e are lahc'cd alphilhetically (after Alvarcz er al., 19771.

80

81

5 Fundamentals of Magnetic Stratigraphy

x = (X,

and. X , X,?

X,,,

=

-

x c o ~ l x c-yXCU)

stratigraphic position of event of younger chron boundary of older chron boundary,

= location = location

that 3An.lr(0.253, for example, is alocation within the subchron. 25% the base of the subchron. labeled polarity chrons according to the correlative M-sequence oceanic magnetic anomaly. In contrast to Cenozoic marine magnetic anomalies, most but not all of thc Mesozoic (M-sequence) oceanic magnetic anomalies correlate to reverse polarity chrons. Intervening nortnal polarity chrom are given the number of the next (oldcr) reverse chron with "n" appended. As for the Cenozoic, we recommend use of thc prefix "C" when the label refers to a polarity chron (time interval) rather than an oceanic magnetic

i Figure 5.3 Magnctic polarity time scale and polarily chron nomenclature frrr Iatc Miocene tu ~ ~ c c (after n t Caodr and Kent. 1YYZa).

i ! 1

I

(3An) and a preceeding rcvcrse interval (3.4~).The upper normalircverse interval is designated as 3An.l. The rcverseinterval would then be 3 A n . l ~ . In this way a complete designation of the timc scale can be achieved. The location of a geological occurrence in time within such a system may be prccisely placed using the chron label with suffix "x", where

The most complete record of the reversal pattern of the geomagnetic field since 160 Ma is pwserved in the crust o l the ocean and detected by seabornelairborne magnetometer surveys. 'These magnetic anomalies are interpreted to b e due to reversals of the geomagnetic field recorded in the rocks of the ocean floor. The sea floor in a major ocean is quasistratigraphic in that it gets younger toward tbe spreading axis: however, it is not in superposition in thc usual sense. The reversal record as preserved in the sca floor has become the template (a sort of type section) fur geomagnetic reversal history o l the last 160 My and is the basis for the geomagnetic polarity time scale (GPTS). The designation of typc sections. in the classical stratigraphic sense, for the geomagnetic reversal pattern since 160 Ma is unnecessary. On the other hand, type sections for magnctobiochronology. the correlation of the biologic rccord to thc magnetic polarity sequence, might be desirable. For examplc, a potential type sctlion for the Paleogenc might be the Contessa section in Italy (Lowrie ct a/., 1982). The type section concept for polarity history is the only way in which the polarity record can be buill up for time intervals not recorded by present-day sea floor. For periods prior to 160 Ma. for which no oceanic magnetic anomalies are availablc. the establishment of polarity sequences must bc carried out using classical stratigraphic principles involving type seclions. An example of a viable magnetostratigraphic typc section is that. recording the Kiaman Superchron (Interval) in Ausrralia, whcrc the superchron is thought to he tied directly to the mck record (Irving and Parry,

82

3 Fundamentals of Magnetic Strat~graphy

5.3 Field Tests for Timing of Rernanence Acquisition lindouhtedly, one of thc most important techn~calchanengcs in any paleomagnetlc research is to determ~ncthe age of magnetization components It 1s of utmost importance to determine wh~ch.if any, of the magnetizat~on components was acquired near the time of deposition of the sediments or coohng of the igneous rock. In mannc sediments, experiments have shown that the detrital remanencc (DRM) is acquired at some position below thc sediment-water interface, where the water content falls to a level at wh~chthe magnetic grams can no longer rotate (Irving and Major, 1964; Kent, 1973; Verosub, 1977). For pelagic sediments. a reasonable cstimate of the depth at wh~ch the DRM becomes locked is 10-20 cm below the sediment-water intcrface (3.3~).The age of the sediment magnetization w l l not, therefore, coincidc with the age of sediment deposition but will lag behind by intervals dictated by the lock-in depth. In marine sediments retrieved from the bottom of the ocean, the stability of the record can be e9tablished by laboratory demagnetization experiments, by the Internal coherence of the reversal pattern, and by a revenal test. A rcversal tcst is based on the fact that normal and reverse intervals of the core should yield mean directions 180" apart (Fig. 5.4). This is an important test since any magnetnation component acquired later will tend to affefect the reverse and normal directions differently. It is usual to test whether the circles of confidence at the 95% level overlap lor the reversc and normal mean duechons. The results of this test depend on sample size and dispersion of thc data, and if one of the sets of duect~onshas a large

Mean of reverse sitesflipped through 180"

x Mean of normal sltes

+ Dlpole fleld at sampllng slte Figure 5.4 \ l e m dtrc~llonsof ~n<~rx!lall! 2nd rcscrrcl) rn.sgnrl~/cJsllc\ fmm Curllr Ranch. Ssn P ~ d r \j 11.0. (41i731)a) ullh xc'rlhpp!nz circle, ut 45 ' I :amhdr'ncz Rcvcr,r' lacan dirc.11a~i_s Pupped through 180" (after Johnron et 0 1 , 1975).

83

5.3 Field Tests for Timing of Rernanence Acquisition

then no significant difference will be found. McFadden and McElhinny (1990) have proposed a simple version of the reversal test in which the mean directions of the reversed and normal data sets and assare utilized. ~t is assumed that both sets have a common precision. In the case where both sets of data have N greater than5, the following equation can he used calculate the critical angle y, with p = 0.05, whcre N = N, + Nz and and Rz are the vector sums for the two polarities. cos y,

=

1

(N -

- R I - Rd(Ri RIRZ

+ R*)

[(:)A

-

(5.1)

the anglc y, between the two mean directions is greater than the angle then the hypothesis of a common mean direction may be rejected at the percent confidence limit and thc reversal test fails. The "positive" revcrsts are classified on the basis of y,, as A iT y, 5 So, as I3 if 5" < y, s 10", a s C if 10" 5 y, 5 20". Very few data are classified as A since the requirements are so stringent. The data presented in Fig. 5.4 are classified as 6,since the anglc y, is greater than 10' and less than 20". If y, exceeds 20" the revcrsal test is considered to be negative. Sedimentary rocks often becomc involved in tectonic processes and are. then exposed through erosion at the Earth's suiface. This makes possible er tests for stability such as the bedding tilt tcst suggested by Graham 49). Graham pointed out that if thc direction 6f remanent magnetization asstahle from the timeof rockformation, the direction of rcmanence determined from tilted rock strata would be dispersed but would coincide after correction for hcdding tilt (Fig. 5.5). It is now well known that directions of magnetization can be acquired before, during, and after the rocks have been deformed. It is therefore necessary to dctcrmine the stage in the tilting process at which a particular magnetization component reaches maximum concentration.Thisis usually doneby incrementally untiltingthebeds andcalculating the Fisher (1953) concentration parameter (k) at each step. The stage s process at which thc Fisher "precision" or concentration parameter )is maximized is usually taken to represent the attitude of the hcds at the ime that the magnetization component was acquircd. In Fig. 5.5. it can be seen that the highest values of k for component A occur when the beds arein eirpresent position (0% untilling). Therefore, component A was acquired ter folding of the rocks. The B component of magnetization reaches its eatest concentration when the beds are unfolded by -50%; thiscomponent therefore acquired during the deformation process and is a synfoldiig netization. Component C reaches its highest concentration after 100% ding. The C component was therefore acquired before Iolding of the beds and is the only component which can be used for magnetostratigraphic

5.4 Firld Sampling for Magnelic Polarity stratigraphy

ta have a Fisherian distribution. A bootstrap method is employed to

Figure 5.5 Schcmatic p~esscntation~ ~ l f oand l d amglomeratr test. n r e c magnetization mmponcats are prercnt in thew aedimenls. Thc A component is pmstfolding and u,ar acquircd with thc scdiments in thcir present position. Component B is synlt?lding and war acquired when rhe rocks nerr riltcd by 50% of their prtscnt attiludc. Component C ir prcfolding and the directiuns group hrsr after resturing the huds to horizunral. The magneti~ationdirectioor in syndeposiliond rip-up clasts show randomly directed C cornponenuof magnrtiration, in~licatihg that the C component was acquired v e v carly in the history of the sediment. I; is thc Fishor (1953) concentration paramrtcr.

purposes. The improvement in grouping may be tested statistically (McElhinny, 1964). In this test, the precision of the mean direction before tilt correction ( k , )is compared to the precision after tilt correction (kz). The ratio kz/k,can be compared with difference ratio tablcs with equal degrees of freedom (2N - 1).This test has been widely applied hy palcomagnctists for many years and is still inuse. McFadden and Jones (1981) pointed out that the McElhinny (1964) tilt test was invalid because the two estimates of k are related through the bedding attitude. The McElhinny (1964) fold test is often said to be conscrvative; however, it is clear that in some cases it can be misleading, particularly when the magnetization bcing tested was acquired during folding (e.g., Kent and Opdykc. 1985). McFadden (1990) pruposed a fold test based on correlations between the distribution of magnctization directions and bedding attitudes. Recently hvo new tilt tests have been proposcd that have considerable merit Watson andEnkin (1993) argued that the problem could be solved by parameter estimation. This method legitimizes the use of Fisher's precision parameter ( k ) . Their method involves parametric resampling of the original

. :

'n these clasts have magnetic properties similar lo those of thc parent sediment and possess a stable component which is randomly directed. then the investigator has demonstrated lhnt the component in question was acquired early in the history of the rock (Fig. 5.7).The randomness of the clast magnetization directions can be tested using the Watson (1956b) test for randomness. Another important test of early acquisition of magnetization (Graham. 1949) is the "slump" test. Syndcposition slumps are often present in rock xquenccs; however, application this test is uncommon. Tf magnetization was acquired during deposition, remanence directions from slumps should be different from the unslumped parent rock. Alvarez and Lowrie (1984) have demonstrated lhal Upper Cretaceous pelagic limestones in Italy contain large slun~psthat yield directions which deviate widely from those in the unslumped parent ruck, indicating that these sediments were magnetized early in their history.

of

5.4 Field Sampling for Magnetic Polarity Stratigraphy ; The geomagnetic lield reverses in an aperiodic lashion, with intervals or

g,&

- 1 li

r

i.

constant polarity having durations ranging from 20 ky to 50 My. An ideal sampling scheme would be one that is designed to monitor the geomagnetic field every -5000 years. This is often impractical or impossible due to the

asu!s uo?lnlosgg sap!u[ AM [assan d a 0 aql preoqe s!sL~eur!010saugnox lo3 pgsn uaaq seq '8upeds u13 02- ~e sluamarnseam luapuadapq asnpold un3 q3:q~ ' l a i a ~ ~ o l a u 9 eqZno~q1-ssed m v 'par!nba~ ale saxdares aleredas pue a$args!p 'sase:, asaql n~ ,uogeqau8emap leuIraq1 op ol l o 'sluamnlls -u! asayl u! arqenehe ale ueyl splay 8u!leuralle ~ a q f i %u!sn q no!lez!lau8r?m -ap an!ssa18old op 01 alqal!sap ualjo s! I! ley1 s! luamamseam snonuquos jo y3eqmlrrp a u o .fiuqdmesqns lnoql!n\ 'sa1o3 aIoqM l o ss~os-+leq 1!1ds arnseam ul pasn aq ues sralamolau8em s!ua%ohrs q9norql-swd -saqns s!lsqd , u ~ a - ~ q l ! ~Zu!ldmes qseq-01-yaeq suonupuos a q n b a ~ plnoM uo!lnlosar qsns 'sale1 uo~ie~uaumpas :,@eladiv ,uo!leinp s!ql jo slenlalu! aldmes ol ldma~ep1noqs amaqss Bur[dures Aue 'Allcap1 ,(suo~qsqnsuo!unaa pun u ! e i u n n ~q q o 3 '.>a) prosal Kl!le[od aql u! l u ~ s a l dale uo!$eiup u! Kq 02 se uoqs se suo~qsqns LI!IE[Q~.a1qe!1vn STplag syau4emoa8 aql jo sleuansl uaamlsq su1!1 aqL '(286~"lo luamal3 ',4'a) pasn uaaq a m q luam!pas jo s a q s urn-£ u!ql 'suo!l!sue~~Al!lelod 30 Spnls ayl u~ 'alpueq pue luauo oi l{n:,y~~p amm an! sa1dm.e~[[ems ' J ~ A ~ M:pa~ns O ~ -Ram aq plnw saldwes lallerus y m m ley1 ysns s! slalamo~an;imulapow $0 h!~!l!suas aqL .,ms 01-L lnoqa s! amnloh aldmes 'Sllensn 'Lq 01- ueql i a l e a ~ 8uo!lernp q l w ~uorq:,iOue[od an[osal ppoM K ~ j m 01jo sale1 uo!lq -oam!pas s!5elad ~ u e l s u o 3 ~Ll!suap o+ le f i u ~ ~ d u.laM r e ~a n satdmes aqlj! luam!pas am o p ccramr?looa 3!lse[d Zu!ssald Kq l o 'dlp ale sluam!pas aql J! saqns %u!~esLq slenralm us-01 ie saldmes anomar 01 as!l>eld uorumo:, s! I! 'salos keluam!pss au!iem Su!ldmes UI -palduns 8u!aq qsor jo ad&

alu~aur<,l4lrt,~ aql jo ootrrsodap aql Brrrleplsod l u a ~ o d w o 8 s -e pue 3 1 ~ 1 ~ U 1 0 [ 4a~qml ~ UOII o -!sodap Bu!lapa~di u ; l u o 6 w o ~v awrpn! ~ slseplvnpr.\!pu!woq ~ u o ~ u ~ ~ p u o ~ 1 o z !-u!rre~ lau8e~ -r~sm3dmol~~luam!pas yunq3 q ~ n e ~ s n o i a p u o q ~ e ~ 1 1aleratuolauos 01sa1 srllsod L'S anB!j

s

S

'(1661) emzPoW P1re SolnlemElS WOJJ e l e a (q) .(c961) aq.ipdo plre qFno9 w o l ~e l e a (e) .Xu!p[opn 30 sa4ue1 %lj~ 3 0 sdnod u! s@!n 0001 mu13 e w e m aqljo uo!mqpxrp aql smoqs lunr8ols!q axr. ' s e p ~eu?!~oaqqo Bu!ldmesal ~ulawerpdyosa~dmexao ~ l oBu!q!lon~o , aa~8ap sql $0 uo!laun$ e se (7) lalanremd un!leJluasuoJ ( ~ ~ 6 IaqsrJ 1) a q l p anpb a= 9.5

fiu~~l!lun luaaad ,.

. .

~ '

.

mr

P

88

5 Fundamentals of Magnetic Stratigraphy

1983. Weeks et ul. (1993) describe a modified 2G-755R magnetometer in which the SQUID sensing coils are arranged for high resolution measurement of magnetization from continuous "u-channels" trackcd through the sensing region. The width of the response function of the pick-up coils is about 3.5 cm. Deconvolution ofthesignalinthe timedomain (Constable and Parker. 1991) can give comparable resolution to that achieved by back-toback discrete sampling. Thc u-channel (see Tauxe et al., 1983h; Nagy and Valet, 1993), typically 1.5 m in length with a cross sectional area of about 4 cm'. allows continuous sampling of the core with less sediment distortion than for discrcte back-to-back sampling. Thc rate of sedimcnt accumulation is generally variablc through time. Sadler (1981) emphasized this variability and termed thc process inherently "unsteady," characterized by large changes in rates of accumulation over short time periods. In some environments, measured rates of sediment accumulation during floods may be -10' cmlhour. whereas accumulation rates integrated over 10"ears would give much lower values, implying Iargc changes in rates of sedimentation, abundant hiatuses. and unconformitics. The ideal condition lor a continuous record of geomagnetic polarity would be a continuous rain of sedimentary particles accumulating uninterrupted through time. Such ideal conditions are rare, but in certain cases apparently continuous sediment accumulation has been recorded for marine sediments over significanl periods of time (5-10 My) and for lacustrinc sediments over intervals of -lo4 years. Variable rates of sedimentation caused by tectonics, or changes in climate or paleoceanographic circulalion, are well known. When conditions are favorable, complete and continuous reversal sequences have been recorded in marine cores (Opdyke et a!., 1966; Fostcr and Opdyke, 1970: Tauxe eta)., 1983~)and in outcrop (Lowrie and Alvarez, 1977b; Channel1 and Erba, 1992). Some marine sediments prcserve such excellent records of the geomagnetic field that even shortperiod beha\'ior of the field, such as the process oC tield reversal, can be studied (Opdyke et al., 1973; Clement and Kent. 1984).Lacustrine sediments can also be excellenl recorders of the geomagnetic field and they have often been used tostudy secular variation and magnetic excursions (Ch. 13). Unlike deep-sea sedimcnts, terrestrial sediments are rarely, if ever. homogeneous with respccl to lithology or sedimentation rates. Terrestrial sections are usually selected for sampling for particular reasons. such as the presence of important vertebratc launas, radiometrically dated horizons, or climatically indicative lcvcls in loess, tills, or pollen-rich deposits. In the Siwaliks of Pakistan and India. the sedimentary rocks comprise mudstones, sandstones, and gravels deposited in the subsiding foreland basin in the front of the rising Himalayan mountain chain (Opdyke eful., 1979: Johnson el al., 1982,1985; Tauxe, 1980). The sediments were deposited by mcandering river systems across the floodplain and consist of fining upward se-

5.4 F~eldSampling for Magnet~cpolarity Stratigraphy

r

89

quences of river channel sands and overbank silts and muds on which soils developed. Such a sequence must contain many disconformities. An analysis of the variability of sedimentation in this facies indicates that time gaps of thousands of years are present (Friend er al., 1989: McRae, 1990a.b). Nevertheless, Johnson et al. (1985) have shown that over periods of the order of 10Stolo6years the magneticpolarity record is amazingly complete. Sampling in the Siwaliks foredeep sediments was confined to the finegrained mudstoncs and siltstones, with sandstone intervals not samplcd. The sites were. therefore, distributed at random intcrvals determined by Lithology and outcrop availability, with stratigraphic spacing of sites of 5 ,:; to 10 m. Studies on sedimcnts in Argentina deposited by unchanneliized >A,. ..,,: flows on sandflats and distal alluvial fans in an intermountain basin indicate i.. <~,. that the record is fairly complete on time scalcs grcater than l(P years, although in this case the stratigraphic spacing of sites was -25 m (Beer, 1990).Most investigations of intermontane fluvial systems have shown them to be Caithful recorders of the geomagnetic field lor features with duration greater than -105years, although obviously with lower rates of sedimentation shorter sampling intervals are required. Several studies in Pakistan and Argentina have dealt with the variability and reproducibility of polarity zones along strike (Behrensmeyer and Tauxc, 1982; Johnson et al., 1988; Beer, 1990). Johnson et al. (1988) have shown that individual polarily zones (Fig. 5.8) may vary along strike by up to a factor oC two in this type of sedimentary environment. It is a great advantage in magnetostratigraphic studies if sections of appreciable thickness (102-103 m) are available for sampling. Thick sections increase thc probability that more than one polarity zone is present in the section and increases the chances that the reversal pattern can be correlated to the GPTS. The following technique for sampling unindurated sediments was developed by Johnson et al. (1975) for study of Plio-Pleistocene nonmarine sediments of the San Pcdro Valley of Arizona. During the initial field work a suitable sampling interval is chosen. Sitcs are preferentially sclccted in finer grained sediments (silt size or finer) at stratigraphic spacing of 2-20 m. Sections are often sited in arroyos where crosion has exposed frcsh unweathered material well away from local peaks or promontories which might he the sites of lightning strikes. A1 individual sites, it is often ncccssary to excavate a pit to obtain unwcathered material, using a pickmattock, spade, or some other suitable implemenl. Aftcr unweathered material is exposed, a horizontal or vertical face isfashioned on a tist-sized sample, using a hand rasp. The strike and dip of the surface arc measured and marked with tine pen or pencil. Three or more independently oriented samples are generally chosen at each site, so that the polarity can be determined unambiguously and rudimentary site statistics can be deter:

90

5 Fundamentala of Magnetic Stratrgraphy

i DISTANCE ALONG STRIKE (km) 0

2

4

7

8

10

12

14

5.6 Cornelatton of Polanty Zones to the G P ~ S

91

P

5.5 Presentation of Magnetostratigraphic Data Magnetostratigraphic data can be presented as plots of declination and inclination of the primary magnetization components as a function of stratigraphic position in the section. Magnetization directions from individual stratigraphic horizons are often represented as virtual geomagnetic poles (VGPs) computed using the dipole formula (Ch. 2). VGPs from individual sampleshurizons arc compared with the overall paleomagnctic pole computed from the mean magnetization direction for the entire section. Variation in VGP in the section is then expressed as a VGP latitude which represents the latitude of individual V G P in the coordinates of the overall palcomagnetic pole. VGP latitudes close to +90° represent normal polarity, and VGP latitudes close to -90" represent reverse polarity (Fig. 5.1). VGP latitudes are plotted as a function of stratigraphic position of the samples1 horizons. Thc polarity interpretation is usually prcscnted as black (normal) and white (reverse) bar diagrams.

5.6 Correlation of Polarity Zones to the GPTS Figure 5.8 Variation rrf the thickness of polarity zones along 15 km or strike lu Miuccne mvlasse sediments in Pakistan. An individual polarity zone mil? vary in thickness by a lactar of 2. 'Time scdc is from Msnkinen and Ualqmple (19791 (after Johnson el d..1988).

mined. After the sample is returned to the laboratory it is fashioned into a cuhc (approximately 2.5 cm on a side) using a hand saw or grinding wheel. Although thc technique described above is laborious, once the cubes have hecn fashioned they are easy to handle and measure. Unindurated sediments are often sampled by pressing the samples into plastic boxes or cylindexs in the field. The technique is rapid and is convenient for sample transportation and handling. If thermal demagnetization is necessary, the samples must be dried and/or induratcd. removed from the plastic containers. and wrapped in aluminum foil for heating and measurement. Lava flows or indurated sediments in outcrop can be drilled using a portable gasoline-powcred drill. Pelagic limestones in land section, such as those at Gubbio (Italy). are usually sampled by drilling individual oriented cores (2.5 cm diameter, 5-10 ~min length) spaced at 20-50 cm stratigraphic intervals. In lava sequences, three or more cores are usually taken from each lava flow. The most intensive study of this type has been performed by Watkins and Walker (1977) on the Icelandic lavas.

The invesligator is ohen confronted with a seriesoi normal and reverse polarity zones, the pattern of which may correlate with more than one segment of the GPTS. Usually. it is initially assumed that the sedimentalion rate, or rate or extrusion of the lavas, is more or less conslanl. and trial correlations are made on the basis of the fit of the observed polarity pattern to the GPTS. In practice, correlations to the GPTS are aided by radiometric ages or biostratigraphic events in the seclion, which have bccn correlatcd to the GPTS clscwhcrc. For cxample, in the Haritalyangar section in India (Fig. 5.9): the correIation of polarity zones to the GPTS is aided by magnetostratigraphic data from Pakistan where some of the same Miocene vertebrate fossils are correlated to radiomelric agcs (on volcanic ashes) and hcncc to the GP'I'S. Thc fossil assemblage at Haritalyangar is equivalent to that in Pakistan which is correlated to the chron 3A to chron 4A interval, and on this basis the reversal scqucncc can bc unequivocally correlated to thc GPTS on the basis of pattern fit (Fig. 5.9). The resulting regression is significant at the 99% confidence Level. allowing sediments in India to be securely correlated to thc GPTS and lo similar scdimen~sin Pakistan. It also indicates that the rate of sedimentation is reasonably constant. The Haritalyangar section is ideal in that it is long and continuously cxposcd. Short sections which contain few polarity zones are much morc

92

5 Fundamentals of Magnetic Strat~graphy

i

E

f

f

0

'

~

" 8"

" 7"

rates are fairly constant such that the polarity zone pattern reflects the polarity chron pattern in the G n S .

Van der Voo (1WO) has recently suggested a reliability index for paleomagnetic studics. The indexis based on seven reliability critcria, and is designed for paleomagnetic studies that contribute to construction of apparent polar wander paths. Wc propose a reliability index for magnetostratigraphic studies bascd on ten criteria. Some of the criteria are the same as Lhose proposed by Van der Voo (1990): however, somc arc unique to magnetostratigraphy. The criteria and their justification are outlined below:

U L

1

93

5.7 Quality Criteria for Magnetostratigraphic Data

Average sedlmentatlon rate = 30 cmky

11

Oudluy Crltrne lor Magnetr~~tr~~lgraph~c Datn

L ,

HARITALYANGAR SECTION

57

6

5

Time (Ma) Figure 5.9 Magnctostratigraphy of the Haritalyangar scction in India and correlaliun to the GWS of Mankinen and D a l v p l r - 11979). The high linear regebaion value ( r = O.YY8) indicates that the ctlrrrlarion to the GPTS is feasihle and that the scdimcntation ralc was more or ICES constant (alter G. D. Johnson cc "1, 1983).

difficult to correlate. In some cases, prominent subchrons may be missing because of hiatuses. Spurious apparent poIarity subchrons can be an artifact of a hard sewndary magnetizationnot (completely) removed by the demagnetization techniques. Cross-correlation techniques can be used to test the correlation of polarity zones to the GPTS (e.g., Lowrie and Alvarez, 1984; Langcreis er aL, 1984).The polarity zone pattern in stratigraphicsection can be compared to the GPTS and the correlation wefficient computed for each successive match. Thc minimum value of the correlation coefficient which is significant at the 95% confidence level can be used to screen possiblc correlations. The optimal correlation can then be chosen based on stratigraphic age constrainls. It is important to notc that this technique is most useful in sections which record multiple polarity chrons, and where sedimentation

1. Straligrdphlc agc known to the level of the stage, and assoc~ated paleontology presented adequately The placemenl ol loss11 occurrences and ranges w ~ t hTespect to the magnetic polarity determmat~onsiq often critical to interpretation. 2. Sampling localities placed in a measured stratigraphic section. Unless thc section is accurately measured, correlation of the polarity zone record to the GPTS will bc compromised. 3. Complete thermal or alternating field dema$netization performed and analysis of magnetization components carried out using orthogonal projections (Wilson. 1961: Zijderveld, 1967). All modern studics employ demagneti/ation, and vector analysis allows the identification of magnetization components present in thc samples. Blanket demagnetization at an individual temperature or alternating field valuc is no longer considered adequate. 4. Directions determined from line-fitting least squares analysis (Kirschvink, 19x0). This type of analysis has been widely applied in recent years. 5. Numerical data published completely. Black-and-white bar diagrams are not adequate. Data are best presented as VGP latitudelstratigraphic distance plots andlor as dcclinationlstratigraphic distance and inclination1 stratigraphic distance plots. The statistical parameters (e. g., Fisher, 1953) should bc fully documcnted. 6. Magnetic mineralogy determined. It is often useful to know the magnetic mineralogy,, which may constrain the timing of rrmanence acquisition. 7. Field tests for the age of thc magnetization are presented (tilt tests? conglomcratc tcsts). If possible, these should always bc attcmptcd since they constrain the timing of remanence acquisition.

5 Fundamentals of Magnetic Stratigraphy

, 8. Reversals arc antipodal. The Geld should *verse through 180". If this revcrsal test fails. the investigalor probably has not fully removed the secondary magnetization components. 9. Radiometric ages are available in the stiatigraphic sections. Radiometric dates, especially "Ar/-'9Ar or U-Fb ages from volcanic ashes or hentonites, are very useful for correlating polarity zones lo the GPTS. 10. Multiple sections. Magnetic stratiflaphy is more convincing if multiplc sections yield similar or overlapping polarity zone patterns.

Quite clearly, few studies will be able to meet all these criteria since, for example. radiometric dating is oflen dot possible, and field tests are impossihlc in flat-lying beds or deep-sea cures. However, ratings of a t least 5 (out of 10) should be achieved by modern magnetostratigraphic studies.

The Pliocene-Pleistocene Polarity Record

6.1 Early Development of the Plio-Pleistocene GPTS The Plio-Pleistocene gcomagnelic polarity time scale (OPTS) developed From the early studies of Cox el 01. (1963~)nnd McDougall and Tarling (1963a) which coupled K-Ar age deternlinations and magnetic remanence data in volcanic rocks from olil'ornia and the Hawaiian Islands. By lhe end of the 1960s. most of the Plio-Plcistoct!ne polarity chrons had been idcntilied in radiometrically dated volcanic rocks (Fig. 1.2). A s more KAr age determinations became available the chronogram technique lor optimizing the age of polarity chron boundaries was introduced (Cox and Dalrymplc, 1967). The chronogram is a plot of standard deviation or error function of available agcs close to a polarity chron boundary against trial agcs for the boundary (see Harlandet dl., 1990, pp. 106-109). Theminimum error function corresponds to the best estimate of the agc of thc polarity chron boundary. Subchrons within the Plio-Pleistocene werc initially idcnlified by their associated radiometric ages. It is important to note that the early studies of Plio-Pleistocene geomagnetic polarity coupled K-Ar and magnetic data from widely spaced locations and, apart from the work in Iceland (e.g., Watkins et al., 1977). were not carricd crul in continuous slratigraphic sections. Plio-Pleistocene polarity s u b c h r o ~ ~were s initially documented in thc following publications! Jaramillo (Doell and Dalrymple, 1966), Cobb Mountain (Mankinen eet al., 1978). Olduvai (Gromme and Hay. 1963): Kaena and Reunion (McDougall and Chamalaun, 1966), Mammoth (Cox

96

6 The

Pl>occoe-Ple~alo~mr Poldrrty Record I

rt al., 1963), Cochiti (Dalrymple rt al., 1967). and Nunivak (Hoace ec al.,

1968). The two oldest suhchrons in the Gilben were initially identified by Hays and Opdyke (1967) from marine cores and by Pilman and Heirtzler (1966) from marine magnetic anomalies, Eventually they were identified from lava sequences in Iceland and llamed the Sidufjall and 'Thvera subchrons (Watkins et a(., 1977).

6.2 Subchrons within the Matuyarna Chron Subsequent to the recognition of thc existence of a normally magn~tizcd lava at 1.8 Ma by Grommg and Hay (1963)from Oldu\lai Gorgc. McDougall and Wensink (1966) reported a series of polarity zoncs (from base: R-NR-N) in a sequence of lavas iron1 Iceland wilh agcs of 1.60 2 .05 Ma. Yuc to the significant age difference between these normal polarily zoncs and that observed at Olduvai. Lhe normal inten,alsfmm Iceland were considered to be distinct I'rom the Olduvai subchron and were callcd the Gilsa subchrons. Magnetic stvatigraphy in deep-sea cores and oceanic magnetic anomaly profilcss,however, appeared to indicate a single normal polarity suhchron in the lowcr middle part of the hlatuyama Chron, closely coincident wilh the Pliocene-Pleistocene boundary. The question arose, lhereh e , as to whelherthis interval of normal polarity should be called the Gilsa or Olduvai suhchron. This icd to a great deal of confusion and conlroversy (Watkins e i nl., 1975) which was finally resolved in favor of thc name Olduvai. Recently. however. a short-lived normal subchron above the Olduvai has been documented hy Clement and Kent (1987) and was rillled by them ihe Gilsa subchcon (Fig. 6.1). Two short intervals uf normal polarity below the Olduvai were idcntificd in sea floor spreading anomalies and called anomalies "s'* and "y" by Heirtzler rr ul. (1968). The youngcst of these normal intervals probably coincides with a normally magnclized lava dated by McDuugall and Chamalaun (1966) at 2 Ma and named the Reunion Evcnl by Grommb and Hay (1071). This subchron has a duration ol'approximately 5820 ky (Clement and Kent, 1987; Khan et al., 1988). Mankiien el uul (1978) detected a shorl interval uf normal polarity receding .-the jar am ill^^ which t h ~ ycalled the Cohb Mountain subchmn. This suhchron has been observed in high scdimentalion sale cores lrom the Caribbean Sea (Ken1 and Spariosu, 1983) and the North Atlantic (Fig. 6.1; Clement and Kent, 1987). Anolher geomagnetic excursion has been reported between the Jaramillo subchron and the Krud~esiMatuyama boundary hy Maenaka (1987) and Mankinen and Crromm6 (198582) and it

Figure 6., Campari.;on ,,f ,he ma8,,e:ic stratigraphiEs or corcs VZX-170 and flole 609B iron1 rhc North ~ ~ l~h~~ corer ~ cover ~ ithe ~ . timc intervals hul they have iredimentation of ,nagnitudc (sfycr Shncklcton and Opdyke. 1477: Clemcllt rind rates dillrrinL! hy am

.,,.,,.

v2.-., 10Q7) RL,,L,

6.3 Magnctic Stratipmphy in Plio-plristocenc Sedimmtr

99

is thought to have occurred between 0.84 and 0.85 Ma (Champion el a/,, 1988). A subchron of comparable age was observed in a loess sequence from China by Heller and Liu (1984). In thc time scale presented here. (Fig. 6.2). only those subchrons recorded in oceanic magnctic anomaly records and in magnctostratigraphic section are included.

6.3 Magnetic Stratigraphy in Plio-Pleistocene Sediments

I I

'

,

The intensive magnctostratigaphic study of marine sediments of PlioPleistocene age began in the early 19ms utilizing the piston core collections at Lamont Doherty Geological Observatory (LDGO) and cores from thc Eltanin collection stored at Florida State University (Opdyke et a l , 196h; Ninbovitch er a l , 1966; Opdyke and Fnster. 1970; Hays c.1 ul.. 1969; Opdyke and Cilass, 1969; Hunkins et a[., 1971; Goodell and Watkins. 1968). 'The maximum age of the sediments in these studies was limited by the penetration of piston coring devices. The longest conventional piston core in the LDGO collection is about 26 m and, therefore, a stratigraphic record to the basal Pliocene was extremely rare. Cores rc&rding most of the PlioPleisto~xnemagnetic polarity record included E13-17 (Hays and Opdyke. 1967). R C 12-66 (Fostcr and Opdyke, 1970) (Fig. 6.2), and V28-179 (Fig. 6.1) (Shackleton and Opdyke, 1977). In early studies, the durations of the Plio-Plcistoccnc suhchrons were estimated through the magnetic stratigraphy of deep-sea cores by extrapolation of sedimentation ralcs from rcvcrsal transilions or by interpolation assuming constant sedimentation rates and adopting K-Ar ages for polarity chron boundaries (Fig. 6.3). Estimates were compared with those obtained from high-resolution nccanic magnetic anomaly profiles assuming constant sea floor spreading rates. The last decade has witnessed a dramatic improvement in the quantily and quality of the Plio-Pleistocene magnetostratigraphic records due, ill large part. lo the advent in the early 1980s of the hydraulic piston corer (HPC) and, subsequently, the advanced piston corer (APC). The HPC, first used during Leg 68 of the Deep Sea Drilling Project (DSDP), allowed recovery o l several hundred meters or sediment by piston coring, increasing the penetration of piston coring by more than an order of tnagnitude. The HPC was a considerable improvement over the standard rotary drilling techniques, which ortcn rcsulted in poor core recovery and drilling-rclatcd deformation. Notable among the Plio-Pleistocene magnetostratigraphies recovered by HPCIAPC are Lhose Crom DSDP Leg 68 in the Carribhean (Kent and Spariosu, 1983), DSDP Leg 73 in the South Atlantic (Tauxe et al., 1983e, 19Wi;Poore et al.. 1984). and DSDP Leg 94 in the North Atlantic (Weaver and Clcmcnt, 1986). Apart from theirvalue for biomagnetosh-ati-

101,

6:3 Magnetic Stratigraphy in Plio-plehtocene Sediments

0

1

2

3

4

5

6

7

8

9

10

11

12

1314

15

16

Depth in meters Figure 6.3 71mr v i d q h plot show~ngthc position and duratlon of the J a r a d l o suhchroii ~n dcvp-5ra plrlun ores (after Opdvke, 1972)

E ; ME: ' o

-

.Y.Lnl.n

m

.

LIP

1 1 1 ..*q,:3

n

*

.,

V

, . ,

d

R "i

. U -

i

C! 4

, I

~.~, .

' " .,

,j~

graphic correlations. the Plio-Plcistoccnc magnetic stratigraphies from DSDP Lcg 94 were important for two additional reasons. First, as mcntioned above, they resulted in the documenvation of two short-duration normal polarity subchruns within the Matuyama chron (Gilsa and Cobb Mountain subchrons) which had not previously been adequately recurded in sediments (Fig. 6.1) (Clement and Kent, 1987). Second, the correlation of oxygen isotope records to Leg 94 Plio-Pleistoccnc magnetic stratigaphies resulted in a fwsL attcmpt to astronomically tune this part of the GPTS (Ruddiman ez al., 1986. 1989; Raymo et al., 1989). The vast majority of Pliu-Pleistocene marine magnetostratigraphic records are Crorn pelagic deep-sea sediments. PIio-Pleistocene shallow matcr platfofm carbonates recovered from boreholes in the Bahamas (McNeill el al., 1988) and Mururoa Atoll (Aissaoui etal., 1990) have yielded excellent

102

6 The Pllocmo-Ple~stocene Polarity Record

magnclic stratigraphies despite pervasive dolomitization. These magnetostratigraphics provide a powerful chronosbatigraphic tool in an environment where hiostratigraphic control is inherently poor. It is important to balance these importam magnetostratigraphic studies in Plio-Pleistoccne shallow viatcr limestone3 with the many unsuccessful (and usually unpublished) magnctostratigraphic sludies in ancient shallow water limestones which are generally thwarted hy remagnetizatiun or very weak remancnce intensities. The Plio-Pleistocene stages in most geological time scales are derived from stralotype sections located in Italy (scc Riu eta!., 1991). Correlation of these stralutype sections to marine sequences outside the Mcditcrranean is cornplicatcd hy provinciality o f Mediterranean flora and fauna and by changes in lithofacics in the strarotypc scctions. Magnetic stratigraphics in the stratotype region provide the potential means of global currelation. The focus of Early Pliocene Mediterranean magnetostratigraphies has heen the Truhi Limestones, a fine-grained pelagic limestone recording the postMcssinian flooding of thc Mediterranean and cropping out in Sicily and Calabria (southern Italy). Magnetoslratigraphic rccords of the MiocenePliocene houndary and thc Gilbert Chron in Calahria (Zijderveld el ul.. 1986: Channclt ctnl., 1988) were expanded by the dcvelopment of a composile reference section in Sicily (Rosscllo composite) from the MiocenePliocenc boundary to the base o l the Matuyama Chron (Fig. 6.4) (Langcrcis and Hilgen,l991). The Truhi Limcslones are overlain by a more marly facies (Monte Narbonc F'ormation), the lransition occurring in the middle part of the Gauw Chron. Upward extension of magnetic straligraphies into thc Late Pliocene and Early Pleistocene has becn documented in Calabria (Zijdcrveld et al., 1991) and Sicily (Zachariasse et ul.. 1989. 1990). The Plio-Pleistocene houndary stratutype at Vrica (Calabria) originally studied magnetostratigraphically by Tauxe er nl. (198%) has becn restudied hy Zijde~velder al. (1991). and the stagc boundary placed close to. hut just above, the top of the Olduvai subchron. 'fhe land section magnetostratigraphic d ~ t ain southern Italy arc far better quality than those uhtained from rotary cores during ODP Leg 107 in the Tyrrhcnian Sea (Channel1 el ul., 1990b), although biostratigraphic control is better in thc Tyrrhe~iian cores duc to enhanced nannofossil prcscrvation (Rio er al., 1990). The hiomagnetostraligraphic correlations which have resulted from land section studies in Ltaiy and ODP Leg 107 cores (Fig. 6.5) facilitate export of PlioPleistocene stages from the Mediterranean and are an imporpant basis for global Plio-Pleistocene correlarions. Although the Early Pliocene (Zanclcan) type section is withiul thc Sicilian Trubi Limestones, other Pliocene stratotype sections ('l'abianian and Piaccnzian) are in the tcrrigeneous "argille azzurre" (hIue clay) facics

Pigure6A Cornposlte lnagnetlc strattgraphl from Capo Ro%rllrrregton. In S ~ c l l )(arlrr Lang erca and H~lgcn 1 Y Y l )

,'

Figure 6.5 Correlation of the Mcdirerranean and rxtra-Mcdircrrancan tt.opical biozonatinns to the GPTS. The variation of "STI"ST after Hodell w a/. (1991).

he Northern Apennincs (scc Kio el al., 1991). This is a more problematic es for magnctic stratigraphy due to the ubiquitous prcsence of iron su tides, although Pliocene magnetic stratigraphics are available from a fcw sections in this rcgion (Maryeral.. 1993; Channeller dl., 1994).Thc available magnetic stratigraphics indicate that sedimentation in the Northern Apennine foredeep in thc vicinity of the stratotypc scctions wasvery discontinuous and that the region is therefore not suitable for stratotype designations. The Miocene-Pliocene. boundary stratotype is a lithologic boundary the stratigraphic scqucnce a t Capo Rosello (Sicily) (Fig. 6.4). Benson and Hodell (1994) argue that it would be hotter to redefitle the Miocene1 Pliocene boundary and correlate it to the base of the Gilberl Chron in Morocco, because of the inadcquate biostratigraphic definition of the boundary in Sicily. l'hc Pliocene-Pleistocene boundary at the stratotype section (Vrica, Italy) is defined by the lirst occurrence of migrant tam such as Arnica irlrutdicn which cannot bc easily correlated to the open occan. These species first appear in the Vrica section just above the Olduvai subchron (Tauxc el al., 1983a: Zijderveld ur al, 1991). A definition of the Pliocenc-Pleistocene boundary a1 Lhc top of the Olduvai suhchron should perhaps be reconsidered. i Berggren r!t 01. (1980) gave 37 biostratigraphic datums for the PlioPleistocene of the world ocean and many more have subsequently been added (Fig, 6.6). The apparent correlation or biocvcnts to the GPTS can vary from sitc to site due to disconformities, fossil preservation, and the operalional method used lo dcfine the bioevent and thc species itself. Many of these datums have been correiated directly to the time scale. Emiliani (1955) produced the first long records of the change in 8'x0 in the occans through time in braminiferal tests relative to a standard (a belemnite from the PEE Dee Formation in South Carolina). The ocean water was shown to oscillate between limcs when the ocean had SIxOratios similar to those today (inlerglacial conditions) to times when the oceans were enriched in "0 (glacial conditions). Emiliani numbered these changes beginning at the present interglacial (stage 1) and the last glacial (stage 2). In this scheme, the inlcrglacials are indicated by odd numbers and lhe glacial stages hy cven numbers. Shacklcton and Opdyke (1973) extended this nomenclature to below the BrunhesMatuyma boundary and correlated this boundary to isotopic stage 19. Thc 6'" record was extended lo the entire Pleistocene by Shacklcton and Opdyke (1976) and to the base of the Matuyama by Ruddiman eta/. (1986), who extended the numbering system to isotopic stage 102 at the GaussIMatuyama boundary. The isotopic variations since -1 Ma were shown to have a dominant periodicity o l -100 ky (Hays erul., 1976).For the limc from the Late Jaramillo Lo the base ofthc Matuyama: thevariations in 6'" have beenshow to have a dominant

1

106

fi rile plioccne-Pleistocene Polarity Kccord

1

6.4 Astrochranolqk Calibration I$ the plio-Pleislocene GFTS

107

riodicity or -42 ky (Ruddiman d ul., 1986). The isotopic sequence has en extended by Shackleton etal. (1995b) to the base of the Gilbert Chron Site 846 lrom the eastcrn Pacific. The sequential numhcring system was scontinued belnw the GaussiMatuyama boundary and a letter prefix was used to label isotopic stazes in the Gauss and Gilbert Chrons (Fig. 6.7).

il

1

6.4 Astrochronologic Calibration of the Plio-Pleistocene GmS

!

I

!

I I

I

I 'imanauiorni i-rn

t ~ ~ d ~ ~ i Ln ~. , C; ~

Figurn 6.6 'me currelation to the Plio-~leistr~cc~~e CPTS of fim and last appcnrancrs of sclccted huna and Rora. FAD = first appearance datum, LAD = last appcarancr

Hays ef al. (1976) were the first to demonstrate the presence of orhital cycles of ccccntricily, obliquity, and prcccssion in palcoclimate proxies in deep-sea sediments. These authurs adjusted their initial time scale to bring the peak variance of the obliquity cycles to the same frequency calculated for obliquity by astronomers, and were thus the first to usc cyclostratigraphy as a means of tuning time scales. During the 1980s, astrochronology was largely restricted to thc Brunhes Chron and was utilized successfully to constrain the age of the oxygcn isotope record with a precision of a few thousand ycars (e.g., Martinson eruL. 1987). The turning point lor astrochronology/cyclostratigraphy came with DSDPIODP hydraulic piston core recovery at multiple holes for single sites with adequale sedimentation ratcs, oxygen isoiopic records, and magnctic stratigraphies. The hydrkulic piston corer allowed good undisturbed sediment core recovery, and the drilling of multiple holes allowed complete recovery of composite sections. Ruddian ~t ul. (1989) and Raymo et ul. (1989) compiled oxygen isotope data r the Pleistoccnc and Upper Plioccnc at DSDP Leg 94 Site 607 (North tlantic). Orbital tuning using the strong obliquity signal for the Matuyama Chron did not reveal any signilicant discrepancy with the slandard (Mankinen and Dalrymple, 1979) K-Ar polarity chron ages, apart from a signilicantly shorter duration for the Olduvai subchron. For over ten years, the Mankinen and Dalrymplc (1979) compilation of K-Ar ages of Plio-Pleistocene polarity chrons was generally accepted in spite of astrochronology indicatinga significantly older age fortheBrunhes1 Matuyama boundary (Johnson, 1982). The realization that the Mankinen and Dalrymple (1979) polarity chron ages may be in error came with the study of ODP Site 677 in the eastern equatorial Pacific (Shackleton et al., 1990). At this site, the oxygen isolopic record from benthicforaminifcra is dominated by the obliquity signal and thc planktic record by precession. Because the effect of precession is modulated by the eccentricily cycle, the precession-dominated planktic record provided a better meansof correlation to the astronomical dala than the obliquity record. The Site 677 oxygen isotope data could therefore be matched to the astronomical dala with more confidence than the poorly modulated obliquity signal at Site

108

6 The

.5 MY

1.0

'

so.

Plfocene-Plestoccnr P ~ l a n l yRecord

1.5

2.

Y.

Figure 6.7 Isotopic stratigraphy of thc Plio-Plcirtocenr and Lalo Miacme at 841 (allrr Shackleton ef aL, 1900. 1995h).

Sites 677 and

607. Indeed at Site 607. Ruddiman rt a / . (1989) apparently miscounted the number o l obliquity cycles in tlle lower Brunhes and early Pleistocene partly because they accepted the K-Ar age for the BrunhesiMatuya~na boundary. The Site 677 astrochronology (Shackleton et ul., 1990) yielded an age [or the BrunhcsiMatuyama boundary (0.78 Ma) a1 odds with the K-Ar age (0.73 Ma). The older astrochronological age fur the BrunhcsiMatuyama boundary from Site 677 coincided with the observation of discrepancies between

6 4 Astrochronologrc Cal~bratlonof the Pha-PLldoccnc

GPTS

109

conventional K-Arages and Pliocene aslrochrrrnology derived lromcarbonate cyclostratigraphy in thc southern Italian Trubi Marls (Hilgen ahd Langereis, 1989). Spectral analyisof CaC03 content in these Pliocene carbonates yielded periodicities of 15.5, 18.5, 35.0, and 335 ky, indicating a consistent discrepancy with the astronomical solutionsof 19.23.41, and 413 ky. Hilgen and Langereis (1989) computed the age of the Gilbert and Gauss reversals using an age of 3.40 Ma for the GaussiMatuyama boundary and tuning the CaCOl cycles lo the astronomical solulion. The Latc Pliocene and Early Plcistoccnc of southern 1t.aly is characterized by sapropels, dark organic-rich laminated layers. Hilgm (1991a) showed that for Late Pleistocene sapropels from the eastern Mcditcrranean (core RC 9.181). individual saprupels correlate to minimum peak values of the precession index and small- and large-scalc sapropel clusters correlate to eccentricity maxima related to the 100 ky and 400 ky ecccntricity cycles. Hilgen (1Y91a) applied this obsen7ation to the sapropel-bearing Latc Pliocene-Early Pleistocene sccticms in southern Italy which had magnetostratigraphic age control. The sapropel occurrences and sapropel clusters could be matchcd to the astronomical solution leading to a new time scalc for the 1.8-3.2 Ma interval. Hilgcn (1991a) il'lctrated how the time scales of Berggen ef a/. (1985b) atld Raymo et al. (1989) lead to a mismatch of the sdpropel occurrences with the astronomical solution. Below the main interval of sapropcl occurrcncc. Hilgcn (I991b) used the CaCO, cycles in the same southern Italian sections to generate an astrochronology for the Pliocene. Hilgen (IYYlb) used the relationship between the sapropel occurrences and CaC03 content in the Upper Pliocene, to infer that the gray marl beds denoting the small-scale CaCO, minima correspond lo minima in the precession index and that the larger scale CaCO, minima correspond to maximum amplitudc variations o l the precession index related to the eccentricity modulation. The recognition of both a precession and an eccenlricily signal in the C X O 3 record allowed the observations to be tuned to the astronomical solution with considerably more confidence than would have been the case in the absence of the eccentricity modulation. This important papcr rcsultcd in astrochronological ages for Gauss and Gilbert reversals (Table 6.2). Ncogcnc asfrochronlllogy has been exlcndcd inlo the Late Miocene using cores collected during O D P Leg 138 (Shackleton rt a/., 1992, 1995a). Multiple cores at each site allowed composite sections lo be compiled at each site using magneticsusceptibility and GRAPE (Gamma Ray Atlcnuatiou Porosity E\~aluaror)data (Hagelberg et a/., 1992). The GRAPE records reflect the ratio of calcitc to biogenic opal. and thc GRAPE records from the composite sections were matched with the orbital insolation solution of Berger and Loutre (1991). 'This matching has led to astrochronological

6 5 '"Ar/"~r Age C d r s t i o n nf the pho-Ple~\ior'cneDP I S

11 1

age estmates for geomagnetic reversals in the 3-6.2 Ma (top Kaena to c3A.112) interval. Conhtent estimates were der~vedfrom several holes/ sites. The estimates are within a few tens of thousands of years of the estimates given by Higen (1991a,b) from southern Italian scchons and differ from those glven by Candc and Kent (1992a) by up to several hundred thowand years. Fiaally. these authors (Shackleton et oi., 199Sa) generated a sea lioor anomaly tlmc scale beyond C3A.2n by utilizing theu age lor the top of C3A.ln (5.875 Ma), the radiometrrc age of Baksi (1993) for the top of C5n.ln (9.639 Ma), and the sea floor d~stancesgiven by Cande and Kent (1992a). How far back ~n timc the astrochronolog~caltcchnlque can be carried will resr on the availablhty of a suitable proxy climatic records in the oceans or croppmg out n land (see Hilgon el ai.. 1995).

6.5 40Ar$PArAge Calibration of the Plio-Pleistocene GPTS The advent of high-precidon ' % / ' 9 ~ age dating techniques and the implication from cyclostratigraphythat the conventional K-Ar ages for the PlioPleistocene (e.g., Mankinen and Daltymplr, 1979) were too yonng have resulted In a large number of "AriT9Ar studies aimed at testing the astrochronological agcs of Plio-Pleistocene polar~tychrons. The convenk~onal K-Arage for the BrunhesNatuyama boundary (0.73 Ma) has been modified to 0.78 Ma by a number of independent "'Ari3'Ar studieq (Lrrett and Obradovich, 1991; Baks~eraL, 1992; Spell and McDougall, 1992; Tauxc et a!., 1992; Hall and Farrell. 1993). T h ~age s is eonslstent with the cyclostratigraphlc estmates glvlng strong support to the new chronology (recent * h i 39Arage~ summari~edby Baks~,1994,seeTable 6.2). Thc astrochronological ages derlved by Shackleton eta!. (1990) and Hilgen (1991a,b) have now been auppurted by "Arp9Ar agcs for the Jaramillo (Glass er al., 1991: Spcll and McDougall, 1992: Tauxe eta!., 1992). Cobb Mountain (Turnn et a/., 1994), Reunion (Baksi ef al., 1993b3, Olduva~(Walter et a t , 1991; Baksi, 1994). Kaena and Mammoth (Renne et ol., 1993; Walter, 1994; Walter and Aronson, 1493), and Gilbert (McDougall et al., 1992; Baks~er or., 1993a) (Fig. 6.8). The wAri3qAragcs ol Pl~o-Pleistocenepolanty chrons (reviewed by Baks~.1994)have m generalconfirmcdthe astrochronolo~caldetemtnations, thereby ratifying the astrmhrunology (Tahlc 6.2) The ages for PlioPlei5tocene reversals based on oceanic magnetic anomalies glven by Cande and Kent (1992a) arenot consistent with the "h~r/'~Arorastrochronological estimates, part~cularlyfor the Gilbert Chron (Tahlc 6.2). D. S. Wilson (1993) analyzed the magnetic anomalies at various spread~ngcenters and concluded that the anomaly spacing is, in fact, consistent with the astrochronological esrimales. Cande and Kent (1992a) adopted the spacing oi Pl~a-

-

112

6 The Pliocene-Ple~stocene Polanly Record

. 3

-

2

-

Llnear correlallon coefllclent= 0 999

Late Cretaceous-Cenozoic GPTS

0

1 2 3 Polarlty uansltlon KiAr and ArlR ages (Ma)

4

Figure6.8 CompaWs<mufgeomagnelicpolarity r e w n a l ages b a r d onorbirvltuning(S11asklcIon el a / . IYYU: Hilgen 1001a.h) with reversal aces b a d on K!Ar and "'A~~"AT lilldspar ~ g e ~ s from Ihe Turkana Baxin (Spcll and .McDougail. 1992: McUou~alier oi.. 19%).

Pleistocene anomalics given by Klitgord er 01. (1975); which appears to be the source of the error. In Cande and Kent (19951, the astrochronological estimates.fot the age of the Plio-Pleistocene reversals were adopted (Table 6.2). The consistency of the astronchronological. 4"Ari-'YAr.and sea floor age cstinlates for Plio-Pleistoceneanomalies has ratified the astrochronological techniques. The resolution of astrochronological estimates for the age of reversal boundaries in the Plio-Pleistocene has reached one prccessio~i cycle (-20 ky), only a faclor of two greater than the time taken for field reversal to occur. Rennk et al. (1994) have taken the next step and used the astrochronological estimates to calibrate the 4nAri'9Ar dating standard (Fish Canyon sanidine, Mmhb-I)>thereby reducing the uncertainty to 0.6% for "'Ari3'Ar ages calibrated against this standard.

'1 i.

:

7.1 Oceanic Magnetic Anomaly Record -.

Vine and Matthews (1963) proposed that lineatcd magnetic anomalies observed at sea using towed magnetometers resulted from thc permanent magnetization of the oceanic crust. They suggested that this magnetic signature was a result of normal and reverse polarity gcomagneticfields and the sea floor spreading proccss outlined by Hess (1962). This hypothesis was essentially cunfirmed by Vine and Wilson (1965). Pitman and Heirtzler (1966). and Vine (1966). These papers demonstralcd that: (1) the. magnetic anomaly pattern was generally symmetric on both sidcs of a spreading ridge, and (2) the anomaly pattern emanating from the ridge reproduces in detail the mvenal sequence obtained by the radiometric dating of lava flows o n land. As a result, Vine (1966) and Pitman and Heirtzler (1966) extended the magnetic reversal pattern to 10 Ma and dated it by assuming conslanl spreading rate and an age for the base of the Gauss Chrou. It rapidly became apparent that correlatable magnetic anomalies generated by sea floor spreading extend for thousands of km away from the ridge crests of the North Pacilic, South Pacific, and South Atlantic oceans (Fig. 7.1). This led to the realization that it might h e possible to extend the geomagnclic polarity time scale (GPTS), as derived from these anomalies, back in time to the late Cretaceous with reasonable accuracy if the sprcading rates rcmainod constant. Heirtzler el a/. (1 968) compared anomaly patterns from all oceans. Taking inlu consideration what was known about the age of the sea floor at that time, they took a courageous step by deciding that the South Atlantic Ocean had spread at an almost constant raLe since the Late

7 1 O c r m ~ cMagnrllc Anomaly Record

115

1

Figure 7.1 Oceanl~rnagnetlc atlomdiv profiles from South Atlant~c(V-20). North Paclhc (V16), and South Pacific (S1-6. EL-19s) vllh svnthetzc model o r o h l~e ~T7mr r a l r,-,,m.t--.nad .- ~ ---.-.,..<.."*,bu assunling crlnstant sealluur spreading rates and an age 013.35 Ma for the h ~ s cof the Gauss Chron (afior Heirtrier el uL, 1968).

.

~~

Cretaceous. They then simulated the polarity pattern necessary to give the obscwed pattern of magnctic anomalies, assumed an age of 3.35 Ma for the base of the Gauss, and assigned ages by extrapolation assuming a constant spreading rate. With this bold and imaginative step, Heirtzler et nl, (1968) produced a dated polarity pattern for the entire Cenozoic and part of the Cretaceous and deduced an age of M) Ma for the KIT houndary. This geomagnetic polarily time scale (GPTS) became a vital tool for deciphering the age and tectonic history of the ocean basins. In their original treatment of magnetic anomalies, Heirtzler et nl, (1968) assigned numbers to certain prominent magnctic anomalies, from anomaly 1 (Brunhes Chron) at the ridge crest to anomaly 32 (Late Cretaceous) (Fig. 7.1). The labeling of magnetic anomalies was found to be very useful as an aid to correlation. As more detailed surveys bccamc available, however, it became clear that a finer division of the anomaly pattern was desirahlc. A study by Blakely (1974) of closely spaced magnetic anomalies in the Neogene Pacificled to the adoption of a numherilellcr schcmc for magnetic anomalies in thc Miocene. a change made necessary by the high reversal rate during that time. ,~Since the publication of Hcirtzlcr et u l (1968). most published geomagnetic polarity time scales (e.g., LaBrecque et a!., 1977; Berggren et al., 1985a,b: Harland er nl., 1990) utilized the hasic Heirt~leret ul. (1968) anomaly sequence interpretation (Fig. 7.2), albeit with some important modifications for anomalics 4A to 6 (Blakely, 1974), anomalies 1to 3A (Klitgord era/.. 1975). and anomalies 30 to 34 (Cande and Kristoffersen, 1977). The most recent and extensive revision of the geomagnetic polarity time scale for the Late Cretaceous and Cenozoic has been made by Cande and Kent (1992a. 1995). In most recently derived Late Cretaceous-Cenozoic geologic time scales, absolute ages are correlated to the GPTS, which is then utilized to interpolate between absolute age estimates. The template for the GPTS is derived from oceanic magnetic anomalies using the constant spreading rate assumption. The absolute agcs of geologic stage boundaries can then be estimated if the stage houndaries are correlated to the GPTS. In this way, the GPTS is central to the construction of geologic time scales and provides the means to corrclatc among the diverse measures ol gcologic time. such as biostratigraphy, isotope stratigraphy, and absolute ages. In some time scales, the link betwccn geologic stages and absolute agcs has been attempted without use of the GPTS as a means of intcrpolation (e.g., Odin er a/., 1982): however, it is now generally accepted that the GPTS should he the central thread of Late Cretaceous-Cenozoic time scales. Cande and Kent (IYYZa), aftcr surveying oceanic magnetic anomalies from the world's oceans, concluded, as had Heirtzler et ul. (1968), that the

7 Lale

~ T E ~ ~ C T O U ~ ~GPTS K ~ O ~ O ~ C

1L

7.1 Occanlc MagnBle Anornalv Record

117

South Atlantic magnetic anomalics were the most appropriate as the basis for a ncw GPTS. The South Atlantic has a well studicd set of magneric anomalies (Candc et uL, 1988) to anomaly 34. The South Atlantic also has both tbanks of the ridge crest preserved and has no1 suffcred from major plate reorganizalion. It does, however. have a rather slow rate of spreading. The procedure of Cande and Kcnt (1992a) was to use a set oI nine fi'nite rotation poles and delerrninc the distance to nine eonjugale magnetic anomalies. These were positioned on a flow line al approximately 3D'S latitude. The chosen anomalies (4A.5C. 7,13,20,24.31), 33. and 34) were positioned a t intervals of 150 to 300 krn along the flow line. These segments werc designated category '1 intervals. The details of the nnomaly pallcrn belwccn these calcgory 1 intervals were filled in by choosing undisturbed segments of anomaly sequences from individualprolilcs elsewhere in the South Atlantic. 'These sequences werc thcn continued downward 1.5 to 1 km, deskewed. and thc crossover points on the profiles were lhcn used to indicate the position of the reversal boundaries on the sea floor record (Fig. 7.3). Five to nine individual rccordn were stacked after adjustment to the previously determined tie points (category 1 intervals) (e.g., Fig. 7.4). Thc positions and widths of the anomalics wcrc thcn determined by averaging. This process yicldcd an average error of about 795 In order to uhlain Lhe fincr dctail of the anomaly sequence, known [ram fast spreading ridges in the Pacific and Indian occans, individual studies of anomaly sequences from these occans were inserted between the tie points. For instance, thc studies of Klitgordetul. (1975) and Blakely (1974) (Fig.7.51wereused to determine the fine-scale reversal scquence between the ridge crest and anomaly 6. Very low amplitude, but correlatahle, magnelic anomalics have been callcd "tiny wiggles" (Cande and 1.aBrccque. 1974: Blakely, 1974) which can be modeled as short polarity chrons or as intensity fluctuations of the geomagnelic field. Some have already been idcntilicd as short polarity chrons, such as the Reunion subchron (Gromm6 and Hay, 1971) and the Cnbh Mountain subchron (Mankinen el al., 197& Mankinen and Grnmmk. 1982: Clement and Kent, 1987). whichhavc durations of about 20 ky. Cande and Kcnt (1992b) have suggested that many of the correlatable tiny wiggles

Figure 7.2 Evolution rlf Lale Cretaceous-Ccnozoic gcnrnil@elic polarity time scales lrom Hicrtzlcr rr n i (1969 (HDHPLOK) to Cando and Kcnl (1995) (CKYS) (aftcr Mcad. 1996). Key: HDHPL6X: IIei~trlerCral (IY6X).Th176: Tarling and Mitchell (19761,LKC77: I.aRrrcqur PI u/. (1977). MD79: Mankincn and Ddrymple (1979), NLC80: Ncsq el n l (1981). LARI: Lowrie and Alvaicz (iY81), GTSRZ: Harland er 01. (1982). BKFVXS: Brrgqrzn d ul (198Za.b). HHVBX: Haq et al (1987). GTSSY: Harland cr n l (1Y90). CK92: Cands and Kcnt (1992a). CKYS: Candr and Ken1 (1995).

118

7 Latc Cretaceous-Cenozo~cOPTS

119

7.2 Numerical Age Conlrul

REVERSALBOUNDARIES

, 1 1

DESKEWEO

,

:

hiWdJ

1 1

1

1

South Atkrntic

ICIIFpUI) i di,Iillle,l

figure 7.5 Rluck models for anomaly I to anomaly 3A from iasl-spreading ccntcrs intcgratcd within thc B ~ u l hAtlantic (caregory 2) distance intervals (a11zr r a n d c and Kcnt. IYY2a).

DOWNWARD CONTINUED

afl(w

are relatcd to geomagnetic intensity vanations and have coined ihe term "cryptochron" for these short duration intervals.

ORIMNAL PROFILE

Figure 7.3 Prncedurc for constmciinb hlock models from occsmic magnetic anomaly profiles. Original prafile is projcned perpendicular to the strike of the lineations. bandpamcd. downward cunlinued, dcskcwcd, and reversill boundaries determined from zero n o n i n g s (affer Candc and Kent. IY)>a).

7.2 Numerical Age Control $+ rTT

Figure 7.4 South Atlantic Stack for anomaly 1 to anomaly 4. Dihtanccs in the stack (category 2 distances) are delrrrnincd hy averaging the widths of the suhinlen'als o n the individual pmfilcs (aftur Cande and Kmt. 1992a).

The Deep Sea Drilling Project, which began in 1968, provided a wealth of biostraligraphic data for sediments directly overlying magnctic anomalies, providing a minimum pale~ntologicalagc for many anomalies. Improved correlations oI magnctic anomalies to the geologic time scale have subscqucntly been achieved by correlationv ul magnctic anomalies to polarity zones in magnetostraligraphic sections with good biostratigraphic control. It is generally not possible to obtain radiometric ages on the sca floor lavas lhemselvcs, since they have suffered extensive alteration and argon does not completely outgas ai deep ocean floor pressures. Absolute ages are generally correlated to magnetic anomaly records by a Lwo-stcp process: correlation of magnetic anomalies to magnctostratigraphic sections and correlation of absolute ages lo magnetostratigraphic sections. The Cande and Kent (1992a) time scale (Table 7,1) utilized the following nine absolute age tie points for the last 83 Ma. (1) GaussIMatuyama boundary at 2.60 Ma from astrochronology (Shacklcton ct al., 1990). (2) C5Bn at 14.8 Ma through the correlation of foraminifera1 zonal boundary N9INlO to the GPTS (Miller et al., 1985. Berggrcn er a/., 1985b) and the age estimates for this l'oraminiferal boundary from Japan at 14.5 2 0.4 Ma (Tsuchi ei al., 1981) and from Martinique a t 15.0 -C 0.3 Ma (Andreieff et al., 1976). (3) Oligocene-Miocene boundary correlated to the middle part of C6Cn (Berggren et al., 1985b), chronogram agcs for this slagc boundary (23.8 Ma) from Harland ct 01. (1990). (4) EoceneOligocene boundary correlated to chron C13r (.14) in the Apennincs (Nocchi ef a!., 1986). with absolute age estimatc (33.7 1 0.4 Ma) from

120

7 Late Crctaceous-Cenodu>c GPTS I

Table 7.1 Normal Polarity intermis (Ma] PcllariQ'chron

CI" Clr.ln C2n C2rln C2An.ln C2h2n CZAtl.3n C3n.111 C3n.2" C3n.3n C3n4n C1An.ln C3An.2" C3Bn C3Brln C3Br2n C4n.l" C4n.2n C4r.ln ClAn C4Arln C4Ar.Zn C5n.l" CSn.2" C5rln Cir2n C5An.ln CSAn.2n C5Ar.h C5Ar.2n C5AAn CSABn CSACn CSADn C5Rn.ln C5Bn2n C5Cn.ln CSCn.2n C5Cn.3~ C5Dn C5En C611 C6Anln ChAn.2n ChAAn C6AAr.l"

,

Cande and Kcnt (19%)

U.OOW~~RO 0.984-1.049 1.757-19R1 2.197-2.229 2.600-3.054 3.127 3.221 3.325-3.553 40334,134 4.265-4.432 4.6114.694 4.812-5.046 5.705-5.946 6.078-6.376 (r744-6.901 bY46-6.981 7.153-7187 7.535-7.376 7.404~~7.892 8.047-8.079 8529-8.861 9.0h9-0.146 9.428-9.491 Y.592-9.735 9.777-1 0.834 10.940-10.989 11.378-1 1.434 11.852-12.000 12.108-12.333 12.618-12649 '12.718-12.764 12.941-13.UW 13.263-13.476 13.67414.059 14.164-14.608 14RD0-14.894 15.038-15.162 16.035-16.318 1h..i52 -16.511 16.583-16.755 17.310-17.650 18.317-18.817 10.083-20.162 20.546-20.752 21 021-21.343 21.787-21.877 22.166-22.263

121

7.2 Numerical Agc Control

Table 7.1 continued Cande

~~~t (1945)

u.u~o-~.~xu 0.990- 1.070 1,770-1.950 2.140-2.1 50 2.581-3040 3,110-3.220 3.330-3.580 4.180-4.290 4.480-4.620 1.800-4.890 4.980-5.230 5.894-6.137 6.269-6567 6.915-7.U91 7.135-7.170 7.341-7.375 7.432-7.562 7.650-8.072 8.225 8.257 8.6994025 9.230-Y.308 9.580-Q.642 9.740-9.880 9.920-10949 11.U52-11.099 11.476-11.511 11.935-12.078 12.184-12.401 12.678-12.708 12.775-1?.&19 12.991-13.139 13.302-13.510 13.703-14.(i'76 14.178-14.612 14.800-14.888 15.034-15155 16.014-16.203 16:327-16.488 16.556-16.726 17.277-17.615 18.281-18.781 19.048-20.131 20.518-211.725 20.996-21.320 21.768-21.859 22.151-22.248

(1Q95j 0.000 -0.780 n.ow 1.070 1.770-1.950 2.140-2.150 2,580-3.040 3.110-3.220 3.330-.4.580 4.180-4.200 4.480-46?(l 4.800-4.800 4.980-5.230 5.829 6.051 6.173-6.450 6.795-6.943 6.986-7.019 7.183-7.216 7.271-7.308 7.483-7902 8.055-8.088 8.543-8.887 9.106-9.191 9.490-9.560 9.670-9.827 9,874-1 1.089 11.214-1 1.273 11.738-1 1.806 12.314-12.495 l2.kZ812.901 13.256-13.294 13.37-13436 13.654-13.843 14.U50-14.312 14.555-15.021 15.147-15.677 15.901-M.007 16.178-16,320 17.289-17.592 17.628-17B00 17.872-18.051 18.617-18.955 19.603-20.074 20.321-21.295 21.633-21.814 22.047-22.324 22.703-22780 23.025-23.107

pdaiity chrun

Cande and Kent (1992a)

C:andr and Kent (1995)

Wci (1995)

C6AAr.2n C6Bn:ld ChBn.2n C6Cn.ln C6Cn.2~ C6Cn.3" C7n.ln Cm.2n C7An C8n.ln C8n.2~ c9n C1On.l n ClOn.2" Clln.ln Clln.2n C12n Cl3n Cl5n Cl6.ln C16n.2" Cl7n.l" Cnn.2n C17n.3~ C18n.ln C18n.2n Cl9n C2Un C21n C22n C23n.ln C23n:Zn C24n.ln C24n.2~ C24n.3~ C25n C26n C27n C28n C2Yn

22.471-22.505 22.599-22.760 22.81623.076 23.357-23.537 23.678-23.800 23.997-24.115 24.722-24.772 24826-25.171 25.482-25.633 25.807-25.914 25.974-26.533 27.W4-27.946 28.255-28.4X4 28.550-28.71 6 29.373-29.633 29.737-30.071 30.452-30.915 33.050-33.543 34.669-34.959 35.368-35.554 35.716-36.383 36.665-,37534 37.M7-37.915 37.988-38.183 38.500-39.639 39,718-40.221 41.353-41.617 42.629-43.868 46.284-47.861 48.947-49.603 50.646-50.812 50.913-51.6051 52.238-52.544 52.641-52.685 52.791-53.250 55.981-56.515 57.800-58.197 61.555-61.951 63.303-64.542 64.911-65.732 66.601-68.625 68.745-69.683 71.722-71.943 72.147-73288 73.517-73.584 73.781-78.781 83.00W

22.459-22.493 22.588-22.750 22.804-23.069 23.353-23.535 a.677-23.800 23.999-241 18 24.730-24.781 24.835-25.383 25.496-25.618 25.823-25.951 25.992-26.554 27.027-27.972 28.283-28.512 28.578-28.745 29.401-29.662 29,765-30.098 30.479-30.939 33.058-33.545 34,655-34.9411 35.343-35.526 3568536,341 36.618-37.473 37.60637.848 37.92W38.113 38.426-34.552 39.631-40.130 41,25741.521 42.536-43.789 46.264-47.906 49.037-49.714 511.778-50.946 51.017-51.743 52.364-52.663 52.757-52.801 52.903-53.347 55.901-56.391 57.554-57.911 60.920-61.276 62.499-63.634 63.976-64.745 65.578-67.610 67.735-68.737 71.071-71.338 71.587-73.004 73.291-73.374 73.619-79.075 83.000-

23.284-23.312 23.392-23.527 23.572-23.794 24,031-24.183 2d.302-24.405 24.573-24.673 25,191-25.235 25.281-25.579 25,850-25.982 26,136-26.247 26.284-26.784 27.214-28.100 28.400-28.625 28.690-28.854 29.514-29.779 29.886-311.228 29.576-31.103 33.313-33.812 34.922-35.200 35.586-35.760 35.9W-36.518 36.771-37.543 37.Mfl-37.877 37.941-38.112 38.389-39.382 39.451-39.892 40.8'18-41 135 42.W-43.245 45.731-47.511 48.77849.540 30.734-50.921 50.034-51.802 52.478-52.800 52.M7-52.947 53.057-53527 56.113-56.584 57.691-58.027 M1.850-61.188 62.359-63.47'1 63.821-61613

C30n

C31n C32n.ln C32n.2n C32r.l" C33n C34n

--

Odin er al. (1991). (5) An age of 46.R ? 0.5 Ma by Bryan and Duncan (1983) correlated to C21n by Bergyren er 01. (1983a). (6) An age of 55 Ma for the NP9lNP10 nannofossil boundary (Swisher and Knox. 1991) correlated to the PaleocenciEocenc boundary (Berggren et a / , 1985h). (7) Cretaceous-Paleocene boundary at 66 Ma (Harland er NL, 19YO). (8) Campanian-Maastrichtian bOund.dq at 74.5 Ma (Ohradovich and Cobban. 1975; Obradovich er al., 1986) bascd on the correlation of this stage boundary to the late part of C33n (Alvarez el a/., 1977). (9) CampanianSantonianboundary al84Ma (Obradovich er al., 1986: Alvarez er dl., 1977). In the revised version of their time scalc, Candc and Kcnt (1995) adopted the astrochronological eslimates (Shackleton rt al.. 1990; Hilgcn, 1991a.b) for a11 Plio-PlcisLocene reversals, and the 65 Ma estimale (as opposed to 6h Ma) for the Cretaceous-Tertiary boundary (Swisher ef al., 1992). Otherwise the absolute age tie points are as for Candc and Kcnt (1992a) (Table 7.1, Fig. 7.6). Wei 1995) has laken issue with cwlibrationpoints (2), (3). (4). ( 5 ) . and (6) used by Cande and Kent (L992a, 1995). In place of calibration point (2) for C5Bn3Wei (1995) proposcd a calihration point at 9.67 Ma for C5n (Baksi et ul., 1993a) and 16.32 Ma lor C5Br (Baksi, 1993). In place of calihration point (3) for the middle part of C6Cn at the Oligocene-Miocene boundary, Wei (1995) proposed a calibration pnint at 28.1 Ma lor C9r (Odin et nL, 1991). In place of calibration point (4) for the EoccneOligocene boundary, Wei (1995) proposed a calibration point of 35.2 M a for C15r. h i s is based on a linear regression of nine age determinations in the C13r to C16n interval. In placc of calihration points (5) and (6).Wci (1995) proposcd a calibration point of 52.8 Ma Cur C24n.lr. This is bascd on an "'Ari3'Ar age within a polarity chrun interpreted as C24n:lr in terrestrial sediments in Wyoming (Wing el al., 1991; Tame 01.. 1994; Clyde er a/., 1994). The resulting time scale differs significantly horn the cande and Kent (1992a. 1995) time scalcs. particularly for the Miocene (Table 7.1). Obradovich (1993) has revicwed "ArI3%r agc control on Late Cretaccous ammonite zones in the U.S. Western Interior. These zones can be correlated to the European ammonite zones to provide the best available numerical age estimates of AlbianiCenomanian and younger Late Cretaceous stage boundaries (Tahlc 7.2). Building on earlier work (De Boer. 1982 Schwarzacher and ~ i s c h r r 1982: , Wcissert et nl.. 1985; Herbert aud

Figure 7 6 Gcoma$nctic polarity time scalc for Late Cretacetru and Cenozoic (uRcr Candr and Kent. 1'195). Polarity chrons with duration lrsa lhan 3U ky arc omitied.

Pleistocene Pliocene

m,m

m

:

z

Miocene

Oligocene

19 20

(U

rn 0

-m m

Eocene

a

Paleocene

27

I

7 2 Numerical Age Control

125

Fischer, 1986), Herbert et a1 (1995) have determined the durations of the Aptian, Alhian. and Cenomanian stages on the basis of cyclostratigraphy in the pelagic scdimentsin central Italy. A consistent bundlingof limcstonemarl couplets in a nearly 5: 1 ratio, observed in Barremian-Cenomanian ltalian sections, first led De Boer (1982) and Schwarzachcr and Fischer (1982) to propose that the carbonate rhythms represent precessional forcing grouped into 100 ky increments by the eccentricity envelope. Herbert rt a/. (1995) combined cyclostratigraphic estimates for the duration of the Aptian, Albian, and Cenomanian with Obradovich's (1993) cstimale for the Cenomanian-Turonian boundary (93.5 2 0.2 Ma) to calculate ages for the Barrcmian-Aptian, Aptian-Alhian, and Albian-Cenomanian boundaries (Table 7.2). One clear discrepancy concerns the Campanian/Maastrictian boundary (see Table 7.2). This stage boundary is generally considered to coincide with the top of the Glohotruncanita calcarata foraminifera1 zone in Italian pelagic limestone sectionsl and this zonal boundary lies in the upper part of C33n. Therefore, the Cande 'and Kcnt (1995) time scale implies an age >73.62 Ma (age of top ofC33n)for the CampanianlMaastrichtian boundary, whereas the definition of the stage boundary in the Weslern Interior (at the base of the Baculilcs elimi zone) yields an age of 71.3 Ma (Kennedy et al., 1992; Obradovich, 1993). The younger age for this stage boundary appears to bc consistent with Campanian foraminifera-bearing limestones from Mexico (Rennb et aL, 1991). In the deep-sea fan deposits of the Point Lorna Formation (San Diego. Calilornia), the CampanianIMaaslriclian boundary lies in C33n and can be defined on the basis of ammonites and molluscs (Bannon et al., 1989). Due to the endemic nature of the macrofauna, direct correlation from thc Point Loma iaum lu the Weslern Interior fauna is not possible.

Paleogene and Miocene Marine Magnetic Stratigraphy

8.1 Miocene Magnetic Stratigraphy Research on the magnetic stratigraphy of the Miocene began with the study of conventionalpiston cores with relatively low rates of sedimentation (Hays and Opdykc, 1967). In central Pacific cores. Foster and Opdykc (1970) obtained magneticstratigraphiesfor sediments asoldas Chron C5N (previously known as Chron 11) (Fig. 8.1).These studics pushed the limits of depth, and hence age of sediment, that could be reached by con\,entional piston cores. In the absenceof suitable sections exposed on the continents, other methods of extendingthe polarity sequence and correlatingit to ihemicropaleontological record had to be found. Consequently, studies were initiated on piston cores from the central Pacific in which hiatuses were present. These studies resulted in the correlation oisiliceous fossilzones to scdiments as old as Miocene (Opdyke etal., 1974; Theyer and Hammond, 1974). The magneticstratigraphy of the M3ocene languished until the development of the hydraulic piston wrer (HPC) and advancedpiston corer (APC) by the Deep Sea Drilling Project (DSDP) and the Ocean Drilling Program (ODP). The magnetostratigraphic coverage of the Late Miocene is relatively complete and high quality data are availablc from land sections (Langercis et al., 1984; Krijgsman et a[., 1994a; Hilgen et aL, 1995) and HPCIAPC cores from North and South Atlantic (Fig. 8.2, columns I-23-4-5), Indian Ocean (column 7), and Pacific Ocean (columns 8-9). As aresult, corrclationsfor this time period are available for both high-and low-latitude sites. The magnctobiochronulogyfor the lateMiocene wasestablished by Berggrene!al. (1985b) and revised by Backman et al. (1990) and Berggrcn etal. (1995).

figure 8.1 M.dgnslic declination in KC1245 after aitcrnating field demagnrtiralion at peak fields 015 mT. The black (normal) and white (reverse) bar diagram on the left indicates the proposed extension of thc gcomagnrtiu time srale baaed on lhis early study (after Foster and Opdykc, 1970). Old polarity chron nomenclature given in brackets.

-

I

I

Ma

N 0

NORTH ATLANTIC & MEDITERRANEAN

1

SOUTH A T L A N T I C

II

/

INDIAN

PACIFIC

Table 8.1 Miiene RCXLunit

1

Silr 6llX

LO

r\ge r i i

Hr-COC

~ e ~ i n n

m

A

N Allmllic 14.84 -2:?.UY

NSF 6

NSI 1,s

M

NS\

1

S

W

r> R M

UM

AU

-

A

70 1

-

A

NMZ N C ~' i r ~ HI'

89

17

41

-~

PC:~

-

o

~ererences

1 Chrn",l

and

Rubinsan

(IYW 2 Poiomid3 Krriclln 1 SLfr 563 Sl(r 583 & Slfr 643 5 Slir 519 6 Ilulr52iA 7 Silr 701

Oil?+

Cmre

13112 +23.20

BIC9

N. ,\tlmntiO

BrCg RrCSP Rr~Cin

N. ,\rlantlt N. ,\tlantic 5 Arl#niie SAclrntlc S. Atlantic

+3ihd +?,;.64

4h~5C Ur C i A N

I67.X

1171 43.77 N l1 1 I . M

+

7

'23

3

1

3-3 m

1Ui 1 2 t 1 27"

I

1-4,

I

111-lm

1

3W 37

I - Oim

1

8%

-2s m P

I

im 124

-26.14 -26.07

-LO26

+O(n.S

-67.1

I I

~3 W

-61 R

I

iSY.1

2

-

-

K

'CA

-

-

I

1 1 I

5 m 5 rn

5

)

-

-

-

-

-

-

1 1

-

-

-

-

LH

B -

d

B

* A $

7

2

L

db 1

1

ZB

1 I VUI

LH

1

z-v

I 1

L-Y

A-1 A-l A

D~i

ZM

R

W 8

I 4

145

lh

1& 20

4

n

4X

3

Y 13

-

-

I

LsnpElcls noi ( l W )

-

2

tinan rc oi (IqRS) Milltrsr 01 (19R5) Mlf#l(19891 'Taukc ri or (1984) Hclltr a d (19R4) Haildoild md Clement

I

-

1 14 34

-

-

4

lli

-

51

-

-

W

-

-

5 5 5

2? as S d 1 0 Y ) - -

-

,,on.> ,.,,.,

1 swirs(lmj d Sehneider and Kent

11991XlI

130

8 Paleogene and Mioccne Marine Magnetic Stratigraphy

131

8 1 Miocene M a p e t ~ cStratlpphy

r

I I

i

Unquestionably, the best results from calcareous sediments of middle and early Miocene age were obtained by Clement and Robinson (1986) from the North Atlantic at DSDP Site 609. The magnetic stratigraphy from thissite is themost complete available for the Miocene. Asa result,ancillary studies such as 6180 and s7Sr/x%rstratigraphy (Miller el al, 1991b) have been carried out. Unfortunately for Miocene magnetobiochronology. the microfauna and Rora at this sitc are not representative of the tropical Miocene from which the standard foraminiferal and phytoplankton zonations are derived. Miller etul. (1991b) have, however, detectedrare tropical foraminifers in this core by processing large volumes of sediment, enabling them to correlate the fauna to the low-latitude zonation. The Miocene rnagnetobiochronology of Berggren et ul. (lS185b) relied heavily on South Atlantic cores from DSDP Leg72 (Berggren etal., 1983b). Unfortunately, magnetostratigraphic data were never published; only the interpretation was published as black and white bars. It is therefore impossible to assess the quality of the magnetostratigraphic data, which makes the magnetobiochronology suspect. The Indian Ocean magnetostratigraphic record recovered at ODP Site 710 is of good quality, although hiatuses are present in the record (Schneider and Kent, 1990b). Correlation of low-latitude foraminifera1 and nannoplankton zonations to the GPTS at this site has led to a revision of the Berggren et ul. (1985a) magnetobiochronology (Baekman
mcd

CPTS

132

R Paleogmc and Mlocrne Manne

Magnetlc Svatlpfdphphg

133

8.2 Paleogene Mapehc Strallgraphy

i

8.2 Paleogene Magnetic Stratigraphy

CONlESSA HIGHWAY SECTION AGE

POLARITY

YGP LATITUDE

909

Unlike the record [rum the Miocene. which has depended heavily on marine cores, the correlation of the magnetic stratigraphy to thc biostratigraphy of the marine Paleogene was h t accomplished in a series of sections in north central Italy by Alvarez et nl. (1977), Lowrie et al. (1982). and Napoleone er ul. (1983). in a series of papers that have become classics. The pelagic limestones in thesc sections at Gubbio and Contessa have magnetization carriod by magnetite and. in somccases. hematite. Both these magnetization components appear to have been acquired early in the history of the sediments. The antiquity of the magnetization has been verified by positive cold tests. Biostratigraphy in these sections required the use of new analytical techniques. Unlikemost studies on deep-sea cores, the foraminifera in thc. indurated sediments at Gubbio had to be identified mainly in thin section (Premoli-Silva, 1977). Initially the nannofossils were considered too poorly preserved to he adequately identified (Premoli-Silva, 1977). It was not until theworkof Monechi and Thierstein (1985) that anannofossil biostratigraphy was availahle in these sections. although the poor nannofossil preservation in this area remains a hindrance to the precise correlation of nannofossil events lo the GPTS. In the Contessa section, volcanic air fall ashes have been datkd by the 4 0 ~ r I ' q ~ r m e t h o(Montanarietal., d 1988).The magneticstratigraphy from the Contessa quarry section is shown in Figure 8.4. The sections of the refion are wellexposed in road cuts and could serve as type sections for integraled magnetobiochronulogy of Paleogene through Miocene time. The highest frequency of reversals of the geomagnetic ficld in thc last 100 My occurs in the Miocene. inhibiting correlation of Miocene magnetic stratigraphies lo the GPTS. The reversal rate decreases going back in time from the Late Oligocene to the Paleocene, facilitating the,correlation of magnetic stratigraphies to the GPTS. The better quality magnctic stratigraphies for the PaleogenC are plotted in Figure 8.5. The polarity zones from the Gubbio section have been given letter designations such as N+, which correlates to Chron 26n. In the time interval covered by the Italian sections, 33 normal polarity intervals occur in the GPTS, whereas 28 normal polarity zones are obsemed in thc Italian sections. The discrepancy of six normal intervals occurs in the Chron 15 to I 8 interval. Further sampling in the Upper Eocene may reveal the fine structure of this part of the magnetostratigraphic sequence. Thc magnetic stratigraphies from South Atlantic DSDP Sites 522,523. and 524 (Tauxe er al., 1983c) covcr most of the Paleogene except for a missing interval in the Eoccne from Chron 21 to 23, representing appraximately 6 My. The quality of the magnetic stratigraphy in thesc cores is

0

30N

Figurn 8d Magnetic stratigraphy and hiostrrjtngraphy of the Eoccnc part of the Contessa section (aRer Lowrie rr 01.. 19821.

exceptional, and their correlation to the GubbloIConteasa sections and to thc GPTS is unambiguous (Fig 8.5). Farther to the south in Antarctic waters on the Maud R ~ s e(ODP Leg 113), cxccllcnt core recovery at Holes hR9 and -..-690 ... - has facilitated the acou~sitionof high-quality magncttc Ftratlgraphics (Speiss, 1990). Hiatuses hinder correlation of the magnetic stratigraphy to the GPTS in some intervals; however, excellent results were obtained for Late Olieocene to Middle Eocene time (Chron 8 to Chron 21 in Hole 689B). An excellent record was also obtained in Hole 690B from Late Paleocene (Chron 26) to Early Eocene time (Chron 24). A short magnetostratigfaphic section has been reported from Hole 702B by Clemen1 and Hailwood (1991) in sediments of Middle Eocene age (Chron 18-21). Other data are available but are often discontinuous (Fig. 8.5). It is interesting Lo note that almost all of these data come from the Atlantic Ocean and Mediterranean area. Data are lacking from the Pacific and Indian oceans. Data are available from the middle Eocene of the Atlantic margin (Miller et uL, 1990), southern England (Townsend and Hailwood, 19851, and from

-

Rak unlr

I " A R C HI

RL~LUII

A

NCt

W5I

NSA

CSAD~C7 NAllrnlic +37.77 +hi* ClhL7Y Italy +433R tI2.V CI~~C?I~ S. ~ ~ l r ~ -2R.55 l i c 2.25 ~ 2 3 ~ ~ 3s .1~ ~ l . n U c I j O l -351 CD-C?L s k~iasllu l S U l 1 . 5 ~ 6 - C I ~S . n ~ ~ r n t ~ s7 6 l l I 3 CIK-CII S . A I L ~ ~ ! ~ ~ 50.95 C -26.31

?

27-1m 3-6m I-.J m IAm Id m

1

Pacdic 13211 lS7.n Atlinllc +im w S. Ailanrir IIXL? 6 5

0

M

D

KM

DM

AD

A

-

A

1

T

-

A A

V I ZliFv z ~ nI ZA L Z~BI

N U 2 NCh Y.R

RT

FCT Q

Refareme8

(IWl)

3 SifrSM Scadin s sttc SZi 6 site ils 7 Site 324 8 site 9 LIule ?(I28 4

iz?

ni-C;U

nrli TV~C26

I 4 Allrnlrs Cuailal PI. I S Alpinc Scaglir

3 '1 I

1

L 1

10

-

140.4 13i 290 -

I I

.A-.dm

I 1

1

mlvl

I

Zull 15s I

>W

2

s

.I ..s m

2

u

m

I I

no tin

2

Ern

1

1.14

-

- - -

-

-

-

-

-

-

-

A A A

A A A

lo .xi

I? I3

41

-

-

s ~h~~ n *I.

1

F+

1 lawie~lll(lOR21

27

9

n

21

Y

46

-

s s

46 54 51

R+ R+ R+ R+ R+

-

11

9

t

6)

13

Z ~ B1

9

i

r

:I

0

31 38

14 13

Z-B

z-tr

Z-v I 2-V

1

51 46 53

-

~

-

-

-

-

0985)

Tavra n oL (1YR4) n a1 (IOPJ) 5 ~ r u r eE I ~ (1W) . t raure ear (19~4) 6 Oement and Ilailwood iIW1) 6 Dleil (1989) 1 sprirl I~WJ I So&$ (l9W1 ti lux^

136

8 Paleogcnc and Miocrne Marlne Magnet~cStratigraphy

the late Eocenc and Oligocene of coastal Oregon (Prothero and A m e n trout. 1985). The magnetobiochronology of the Palcocenc has been discussed hy Bergpen er ul. (IYXSd), Aubry er nl (1988), and Bcrggren et al. (19953. The EoceneiOligocenc boundary has bcen placed by the International Stratigraphic Commission in the Massignano section (Italy) at the level of the last appearance of thc foraminifera Hankenina (Montanari el al., 1988). This position would place the s t a g boundary at C13r(0,14), close to the disappearance level of Honker~inaa t DSDP Sitc 322 (Poore et at., 19841, but above this bioevcnt in the Contessa Quarry scction. From the original paleornagnetic investigation a t Massignaoo, Bice and Montanari (1988) postulated the presence of a short normal polarity subchron immediately prior to thc EocenciOIigocene boundary. This subchron was not identified by Lriwrie and Lanci (1994) and the present consensus is that the oosition of the boundary should be placed at C13r(0.14). a? advocated b; Cande and Kent (IYY2aj The corrclation of the Paleocene-Eocene boundary to C24r(O.6) (Cande and Kent, 1992a) is thc same as that adopted by Berggren el a/. (1985a) and close to that observed in thc Umhritatl sections by Lowrie et a[ (1982). Berggren el nl. (1995) givc an uncertainly of about 1 My in the correlation o l this stage boundary. The earlyimiddle Eoccne boundary is correlated to the GPrS in the Contessa sections and correlates to the top of C22n. The evidence for the placement of the middleilate Eoccne boundary in the polarity s:quence is not so clear, and this boundary is placed in late C18n in the Contessa Highway section. Bcrggren ct 01. (1995) choose to place this boundary in C17n on the basis of fossils from DSDP Sitc 523 in thc South Atlantic (Poore el al.: 1984). This placcment is followed here: however, the correlalion oI this boundary to the GPTS dcpends on the palcontological criteria chosen to define it.

8.3 Integration of Chemostratigraphy and Magnetic Stratigraphy In Plio-Pleistocene sediments. 6'" stratigraphy provides a high-resolution slratigraphic cotrelationtool and a mcansof tuning the time scale and assigning absolute ages Lo polarity chrons (Ch. 6). The Cenozoic 5'80 record dcrived from Atlantic benthic hraminifera (Fig. 8.6) (Miller et a!., 1987) may be interpreted as changes in the tempcraturc of ocean bottom waler or ice volume changes, orboth. Shackletonand Opdyke (1973) have shown thatice volume changes dominated the Pleistocene henthic 6% record. Separating temperature change from icc volume for the early Cenozoic rccord is more

70

'

I "'"'

iceshcst: 0

2

4

ice free

I

k = -0.28 %

I

Figure 8.6 Composite benthic foraminiferal oxygeli isolopc rccord for Atlantic DSDP sitcs corrected t o Cibicidoides and reported to Pee Dcc Bclcmnitc (PDR) svandard. Chrtmortratigraphic subdivisions arc drawn after Borggrcn ?I 01. (1985a). The smoothed curbe is ahtained by linearly interpulating brhvren data at 0.1-My intervals and smoothifig with a 27-paint Gaussjan convolution filler, removing frrqurnciei higher Ban 1.35ikfy. The vertical line is drawn through 1.8960;values greater than this provide evidence for existence of signihcant ice sheets.The temperature scale is computed using the palcotempcrsturc equation, assuming Cihicidoidzs are dcplcted rclativc to cquilihrium by 0.64% Thr lower temperature scale assumes no significant icc sheets, and therefme aw = 1.2%"; the upper scale assumes ice volume equivalent t o rnr,dern valuer. and thrrrrore dw = 028%o (after Miller el a / . 1988).

difficult. In the Paleocene, the S1'O values vary about a value of O%O,then in the late Paleocene 6180 va1ur.s begin to become more negative, peaking a1 -1% in the early Eocene (Fig. 8.6). The St" values becomesteadily more

EOCENE

OLIOOCENE

MIOCENE early

mtddle

late

I

0.7089 0.70880.708781 srtBBsr0.70860.7085-

Figure 8.8 Pt

time

for the ~i~~~~~(after candc and Kent, 1~9s)correlated 10 foraminiferill roller and nannofossil zones (after Herggrerl rm,,rded in 6180 denoted by the prcfix "Mi"(Miller sful.. 1YYlh). Lo 6°C (after Woodruff and Savin. 19911, and "Srl

1995), (afier IIodeli and Wwdruli. 199411).

served in Paleocene and Eocene sediments as largely due to changes in bottom water temperature. They argued that ice sheets became an impor. tant factor only after the EoceneIOligocene boundary, and only since this time (-34 Ma) do oscillations observed in the 8"O denote ice volume change. This is supportcd in some cases by the covariance o l 8180 values from benthic and plauktic foraminifera.Times of sea level fall arc presumed to be coincident with positive shifts on the post-Eocene S"O curve, with changes to more negative values rcprcsenting deglaciation. It should be noted thal because of the smoothing employed in thc S1'O record shown in Figure 8.6, it is not possible to see the high-frequency Milankovitch cycles characteristic of the Plio-Pleistocene record. Miller ef 01. (1991a) havc prcsentcd a new and updated 8"O record for the Oligocene and Miocene (Fig. 8.7). They argue that contincnval glaciation began in Antarctica in the latest Eocene, at about 35 Ma. They have formally defined nine oxygen isotope zones in the Oligocene and Miocene which they have designated Oil and Oi2 and Mil to Mi7, ranging in age from Early Oligocene through early Late Miocene (Figs. 8.7 and 8.8). Miller et al. (1991a) interpret these isotope evcnts in terms of glacial episodes. It is clear that this 8180 stratigraphy does not have the resolution of the Plio-Pleistocene record and isotopic events havc agc uncertainties of 0.5 My using the sampling strategies employed by Miller er ul. ('1 991a). The question arises as to whether early Neogene and Paleogene records can, in the future. be used Lo yield a high-resolution 8'0 stratigraphy. Woodruff and Savin (1991) have studied in detail the mid-Miocene 8 ' 9 shift at DSDP Site 574, and it is possible to correlate isotope stage Mi3 to the section at Site 574in thc 150 to 156 mbsf interval (Fig. 8.7). It would appear that the potential exists for an extension of high-resolution ~ 5 stratigraphy ~ ~ 0 to sediments of Miocene age%and perhaps to thc cntirc Cenozoic. Sile 574 has no magnctic stratigraphy; however, the correlation to the Atlantic composite record is supported by the biostratigraphy. Shifts in the SI3C record have been utilized in scdiment correlation (Bergcr, 1982; Haq el al., 1980; Miller et a/., 1988, 1989). Changes in 8°C are thought to be due to burial of organic carbon, erosion of carbonate sediments (reservoir exchanges), variations in biomass (climate change), or changes in productivity (Berger et al., 1981). For the Pleistocene, inphase changes of @*O and 8I3C, with 8"' becoming heavier and 813C becoming lighter during glacial episodes, are well known (Shackleton, 1977). Such in-phase changes arc common throughout the Neogene beginning by Middle Miocene time (Woodruff and Savin, 19Y1). It is interesting to note that for two earlier SrnOevents thought to represent glacial events (Oil and Mi), the 6I3C shift is synchronous with the 8"O shift, but toward heavier values, not lighter ones.

/i i

/'

I

, I I

/I

I

I

A

l

l

I

,

I

I I

4 ; ; ;

/ I

;

,

I

,

,

,

I

I

) 4 " "

;,

I

I

I

I

/zgEeL f

R FEEP o o d d d

142

8 Paleugene and Mloccnc Martne Magnetic Stratigraphy

The besl documented GL3Ccarbon shift is the prominant uppcr Miocene SI3C shift of -0.7% (Keigwin, 1979) which coincides with Chron 3Bn at about 7 Ma (Haq et al.. 1980) (Fig. 8.8). This carbon shift seems to occur not only in the marine record but also in ancient soils in the land record (Quade et al., 1989). Thc land and marine occurrences are hoth correlated to the GPTS. Recent studies of O D P corm Off Greenland indicate 'that this event may correlate with the onset of glaciation on Greenland (Leg 152 shipboard party, 1994). This event is, therefore, an important link betwecn the marine and nonmarine records. A second important 6'" shift of about +1.5%0close to the boundary bctwccn the lower and middle Mioccne has been documented by Vincent et al. (1985). The middle Mioccnc is marked by the most positive 6I3C values of the entire Miocene and this inrerval has becn called the Monterey event (Vincent and Berger, 1985). This S13C shift is hypothesized to be the result of organic carbon being locked up in sediments around the Pacific Rim. Woodruff and Savin (1991) havc studied the G1'C fluctuations in the middle Miocene in dctail and have designaled SI3C isotopc stagcs 1 to IX. The relativc positions of the boundaries of carbon isotope stagcs I through V1 are shown in Figure 8.8. The cores which record these stages do not have magnetic stratigraphies and the biostratigraphic zonation is nonstandard, which makes correlation difficult. Nevertheless, future studies will undoubtedly result in the S I Trecord being a useful correlation technique for Miocene time. Miller cr 01. (1988) have provided a 8'" an8 Si3C record for the Oligoccnc from DSDP Site 522. which has good magnctnstratigraphic control. They suggcst that this record serve as a type record lor isotopic stages in the Paleogene. Two prominent 8°C excursions are seen in this core which are correlative with the G1'O events O i l and Mil (Fig. 8.7). One of the most promising isotopic correlation techniques to emerge in the last decade is the use of the changing ratio of *'Sri"Sr with time in ocean water (DePaolo and Ingram, 1985). It is particularly useful from the beginning of the Oligocene to the present, when "Sr/80Sr in seawater changed from .7077 to .7090 (Figs. 8.8 and 8.9). High-resolution data are available for Miocene and younger sediments (Hodell et a/., 1991). Miller et al. (1988.1991b) have studied deep-sea cores with magnetic stratigraphy to create a record of Sr isotope changes for the Miocene that can be correlated to the GPTS. The study by Hodell et al. (1991) used a combination of magnetostratigraphically dated cores and data correlated to the GPTS indirectly through biochronology. For some time periods where there is rapid change in 87Sr/86Sr,such as the Lower Miocene, it is possible to correlate magnetostratigraphic sections to the GPTS using the R7Sr/ "Sr record.

8 3 Integration of Chernostratlgraphyand Magnetr Strat~graphy

143

The reason for the strontium isotopic ratio changing with time is a matter of debate. Strontium enters the ocean through the erosion ofmountain belts, ancient shield areas, and oceanic ridge systems. Fortunately, it is not necessary to know the cause of the variation in ordcr to use it for stratigraphic correlation. The integration of chemostratigraphy with biostratigaphy and magnetic stratigraphy can enhance the precision of stratigraphic correlation in theCenozoic (Figs. 8.8 and 8.9). Berggren coined the term magnetobiochronology, which might now have to be amended to magnetobiochemochronology, which is a little awkward; but MBC chronology might be acceptable.

9.2 Nurlh American Neogene and Uualernary

Cenozoic Terrestrial Magnetic Stratigraphy

9.1 Introduction Onc of the long-standing problems of stratigraphy has been the correlation of nonmarine sediments containing vertebrate lossils to marine sediments containing invertebrate Cossils. I)uc to correlation difficulties, thc stratigraphy o l marinc and nonmarine sediments has developed more or less independently. As geomagnetic polarity reversals are eloballv svnchronous. -

Magnetostratigraphic studies of terrestrial sequences began in the early 1970s in a Plio-Pleistocene mammal-hearing scqucncc of tbc San Pedro Valley, Arizona (N. M. Johnson et al., 1975). This study, subsequently updated by Lindsay @t al. (1990). demonstrated the potential of magnetic stratigraphy in terrestrial sequences. The authors were able to correlate widely separated fossil localities and place them in stratigraphic sequence, Correlation of polarity 'oncs to thc GPTS providcd the time frame, The sediments of the basin were mapped using polarity zones and the rates of basin subsidence determined. In the 20 ycars since the publication of this paper, magnetostratigraphic studies of nonmarine Cenozoic sediments have become commonplace and have been used to solve problems in vertebrate paleontology, faunal migration, sedimentology, basin subsidence. and tcctonic history. Correlation of Plio-Pleistocene land mammal scqncnccs to thc GPTS has been accomplished in Norrh America, Western Europe, southern Rus144

145

sia, China, the Indian subcontinent, South America, and Africa (Fig. 9.1, Table 9.1). First and last appearances of mammal taxa have bccn correlated to the GPTS in many individual sections, some of which have radiometric age control. Magnetic stratigraphy of Plio-Pleistocene lake and loess sequences has bcen motivated by the desire to understand continental climatic response to Plio-Pleistocene glaciation. Good quality Plio-Pleistocene lakc sediment magncticslratigraphies exist for North Amcrica, Japan, Australia. and Europe (Table 9.2). The climatic record from lakes is often provided by downcore changes inpollen type and abundance. For locss, good quality magnetic stratigraphies have bcen ohtained from Europe. Central Asia, North America. South America, and, most importantly, China (Table 9.3).

#

9.2 North American Neqene and Quaternary For North America, magnetic stratigraphy of mammal-bearing sequences covers almost the entire Cenozoic and extends into ihc Cretaceous. The terrestrial mammal-hearing stratigraphy in North America is divided into 19 land mammal ages (LMAs) based on stage of cvolutiun and the appearance of immigrant mammal taxa from other continents. usually Eurasia, or at the end of the Cenozoic. from South America. Magnetic stratigraphy offers an independent means of ordering events in mammalian biochronology and thereby enhancing studies of tempo and mode in evolution. The Plio-Plcisrocene nonmarinc magnctostratigraphic record is weU known for most continents and the correlation of the mammalian mnes to the GPTS is reasonably well established (Fig. 9.1). In North America Lhc record is more or less complete, except for the Late Pleistocene and the earlicst Pliocene. The boundary between Blancan and Irvingtonian fauna occurs at the top of the Olduvai subchron, with the first appearance of Mammuth~rsin North America. The base of the Blancan LMA is tentatively placed in the Sidufjall subchron at -4.8 Ma. Figure 9.2 shows the correlation of first and last appearances of mammal taxa tu the GPTS in North America from Late Miocene time. The first appearances may be evolutionary first appearances. such as that of Equus at Nagermann in Idaho. or migrationary first appcarances such as that of Mamrnt~thus(elephant) just above the Olduvai subchron at Anza-Borrego in southern California. Considerable progress has bcen made in correlating the Miocene LMAs to the GPTS in North Amcrica (Fig. 9.3, Table 9.4). Initial studies on Miocene sediments by MacFaddcn (1977) and Barghorn (1981) have been followed by recent studies by Whistler and Burbank (1 992) andMacFadden er al. (1990b) that extended this correlation to sediments containing Clarendonian and Barstovian fossils of middle and early Miocene age. Whistler

146

9 Cenozoic Terrestrial Magnetic Strdhgraphy

and Burbank (1992) placed the boundary between the Clarendonian and Hemphillian m early C4Ar and thc preceeding boundary between the Barstovian and Clarendoman at the base of CSBr. MacFadden etal. (1990b) placed the Hcmingfordian-Barstov~an boundary w~thinchron C5Br The boundary between the Arikareen and Hemingfordian has been placed within CSEr by B J. MacFadden (personal communication, 1995).

9.3 Eurasian Neogene Eurasia has relatively long Neogene terrestrial sequences with abundant fossil content. The most successful studies have been concentrated along the southern deformed margin of Eurasia where Cenozoic sediments have been exposed by Alpinc-Himalayan orogeny. These areas extend from Pakistan and India, where much work has been done on the Siwalik sediments particularly in Pakistan, through the Caucasus and Turkey to the Mediterranean and Spain. Unfortunately, many of the arcas of Western Europe where classical faunas were initially described are poorly suited to magnetostratigraphic study bceausc of limited cxposurc and, in some cases. lack of knowledge of the precise location of fossil finds. The European early Pliocene land mammal age, the Ruscinian (Fig. Y.1), is younger than the flooding of the Mediterranean at the MiocenePliocene boundary, which is now well-dated magnetostratigraphically and astronomically ( ~ 6 6). . Magnctostratigraphic studies in Spain (Opdyke et of., 1989, 1996) place the Turolian-Ruscinian boundary (=Mn 13114 boundary) in the early Gilbert, a t about the level of the lower boundary of the Sidufjall subchron. The Vallesian-Turolian boundary (Fig. 9.3) is placed at 9 Ma by Berggren et ul. (1985b). This is close to thc placement of the ClarendonianHemphillian at 8.8 Ma (Chron 4A) (Tedford et a/., 1987; Whistler and Burbank, 1992). The position of the Astaracian-Vallesian boundary has been a subject of controversy because the base of the Vallesian is defined by the first appearance of Hipparion, several species of which migrated to Europe from North America. The appearance of Hipparion in the Siwaliks of Pakistan is well dated magnctostratigraphically at 10.8 Ma (Johnson et al., 1982). On the other hand, Berggren et al. (198Sh) prcfcrred an earlier age of 12.5 Ma based on radiometric ages from Germany that correlate with

Figure9.1 Selcctcd Ylioccneand Plristocrne Lerrrslriill rrconlaRorn iiorth America, Europe, Pakistan, China, Africa. and S w t h America. Numbers refer to individual studies (scS 'lahlc 9.1). Time scale from G m d e and Kent (1995).

-

--

52

--:s$ ~ c *s; ? < - fg-i s-< 8

i :

4

3

i 7 n " D . = - o

> e = - x : E * =

, E - ? ' * * =

s * $ * - d - - . .

A ns - ~ c

:5 . c

+.

151

9.3 Eurasian Neogenc

-

. , I

e: i "

3 4 $ 8 m z ,;: g g < <

,, g ;

wm$Srizur

ii

Ijnsrtotherium (MI.Pmslnemaw (MI, I Dlnahblus

Pleslwlo

(MI

INMI. Pliotarldea (NMI.

Figure9.2 Lowcst rlratip.phic &;llurn (LSD) andhishest straligaphic datum (HSD) ofNonh American Plia-Plrislocene mammals detcrminrd from magnelostratigraphic studies caner Lindsay ct a/., 19841.Lettcrs in hrackels rcfrr t o location nl fossil finds: A = Ari7cma. T Texas, M = Mexico. NM = New Mexico! C = California. I = Idaho.

=

early H~pparionfauna. Sen (1989) has suggested that H~pparronappears in the Medilerranean a1 approximately 11.5 Ma wthin Chron C5r. This age, which should correlate to ihe base ol the Vallccian, IS thc came as the age given by Tedford er al. (1 987) for thc base oI the Cldrendoruan. Whistler and Burbank (1992) have placed the base of the Clarcndonian in C5Ar.3 at 12.8 Ma, however i m m ~ y a nraxa t were nor used to defmc the boundary. If this placement of the boundary is accepted, then the base of thc Vallesinn IS about 1 My yuunger than the base oC the Clarcndonian (Fig. 9.3). The base of the Astararian, the interval preceding Valleslan in the Europe land mammal ronabon (Fig. 9.3). is placed by Krrjgsman ern!. (1994b) in CSACnaL about 13 Ma, much younger than previous estimates, and very different from the b a ~ of e the Barstov~an,whiehhas been correlalcd to C5Br by MacFadden el at. (19YOb). In North America, the boundary between the

~sriwddo,:) lue18 -!ruu~! ayl $0 amereadde aql uo paseq (EN 8 . 9 ~k[alew!xuldde) ~ , u 3 5 3 n o q ? jo ]led ,([lea aqi 01 s8e lewuem puy ue!hnlsrzg $0 aseq aql pa3eld Al1ua3a~seq (5661) Kespo!~ 'ureds u! Xlepunoq uepele~svpvtueallo pau!maiap Llmau aql qq.n pun auaso!m a[pp!ro a q u! 8u:roo~ s!iew![~ aiaAas q l ! ~~uap!su!m 'eW S.PI Te p a q d S! ue!AoIslea ale[ pur? klrm

I

T

.., <

-

AlikBleean Asenian

.

I

-

,., &, v 5. I " " . . I 1Hemingfold.l B a r s l o v l a n lclsrendonian I H e m p h i l l i a n Orleanian I Aotaracian l v a ~ i e r i a l T v r o l i a n

>

"

. . , -

154

9 Crnu~otcTerrcstr~alMagnet~cStratrgraphy

The magnetic stratigraphy of the OligoceneiMiocene boundary in Europe based on fossil mammals has bcon reported by Burbank ct 01. (1992a). The section studied covers 280 m in the lower freshwater molasse in the Hautc-Savoie (France) and includes the last two micro-mammal zones of the Paleogcne (MP29 and MP30) and the first mammal zone (MNI) of the Neogene in stratigraphic superposition. Unfortunately, a 50-m gap separates probable MP30 faunas from faunas assigned to MNl. The polarity zones are designated R 1 to N8 and the gap Calls between N4 and N2. The authors, however, choose to place the OligoccneiMiocene boundary at the base of NZ, which they correlate to C6Cn.3n. The magnetostratigraphic pattern. however, is ambiguous (Fig. 9.3, column 12). Another possible correlation to the GPTS would he N3 and N4 to C6Bn; and N5, Nfi, and N7 to C6AA. This would place the MP30lMNl boundary at the basc of C6Cn.In (Burbank et al., 1992a) and would place the base of the Agenian at this position. Barbcra et 01. (1994) have placed the OligocenelMioccne boundary at the basc of C6Cn.2n in a section in the Ebro basin of northern Spain. Unquestionably the best studied area of continental Neogene vcrtebrate-bearing sediments is the Siwaliks of Pakistan and India. These studies were begun in the mid-1970s and have continued t o the present. The region is wonderfully suited for magnetic stratigraphy because of well-exposed long sections of very fossiliferous sediments that became magnetized early in their history. In thjs region itis possible tobegin inBrunhesage sediments and systematically work back in timc to the onset of sedimentation in the basin (Keller el nl., 1977; Opdyke et 01.. 1979, 1982; Johnson er al., 1982. 1'185; Tauxe and Opdyke, 1982). In many rcspects the area has served as a laboratory for the application of magnetic stratigraphy to (1)vertebrate evolution and migration (Flynn el ul., 1984). (2) sedimentary processes (Raynolds and Johnson, 1985; Behrensmcyer and Tauxe, 1982; McRae, 1990b). (3) Leclonics (Burbank and Raynolds. 1988). and (4) basin dcvelop-ment (Cerveny et al., 1988). Almost all of these applications of magnetic stratigraphy to terrestrial sediments were anticipated by N. M. Johnson et al. (1975). In the Siwaliks, the magnctic stratigraphy is now so well known from multiple sections that vertebrate paleontologists usually correlate new fossil finds directly to a nearby magnetic stratigraphy and hence to thc GPTS. Thc correlation in China of the mammal-bearing sequences to the GPTS is known in considerable detail from the Late Miocene to the Pleistocene (Tedford el al., 1991; Shi, 1994: An el irl., 1987, Heller and Liu. 1982, 1984). The importance of the Chincse record since the carly Mutuyama is enhanced by the fact that the Chinese loe3s preserves a climatic record that can be correlated to Plio-Pleistocene climatic cyclcs (Kukla et ul.,

9.4 Ahcan and South American Nengene

155

1988; Kukla and An, 1989). A mandible and a cramum of H o m o erectus have been correlated to loess magnctic stratigraphy (An and Ho, 1989). Other data exist from Asiat~cRuss~abut unfortunately they arc not easlly accessible to Western workers (see Pevzner et u l , 1983).

9.4 African and South American Neogene Terrestrial magnetic stratigraphy in Africa stems from the great interest in dating the evolution of man. Long well-exposed fossiliferous sequences are available in central Africa, and magnetostratigraphic data exist for the Gauss to Recent interval in the Omo arca (Hillhouse ct aL, 1986), in the Shungura sequence (Brown er at., 19781, and in Afar (Renni et al., 1993). Other studies of importance arc by Tauxe r t al. (1985) and Deino el 01. (1990) in sediments from the Baringo basin correlative to the CSAA to C5N interval. The magnetic stratigraphy of Latc Oligocene fossilifernus sediments from thc Fayum has been reported by Kappelman et al. (1992). This important study is thc first from the tmestrial Paleogene of Africa. Magnctostratigraphic studies on South American mammal-bearing sediments have been concentrated mostly in the Neogene or Upper Paleogene in a series of studies by Marshall el al. (1979. 1982, 1986), Butler et al. (1984), and MacFadden et al. (1983. 1985. 1990a, 1993). South America separated from Africa in the mid-Cretaceous, allowing the South American mammalian faunas to evolve in isolation. During the Neogene, however, North America and South America approached one another and eventually became joincd through the isthmus of Panama, allowing thc mammals to move Irom one continent to Lhe other. The timing of the exchanges has been revealed by the magnctostratigraphic dating of sediments that conlain the oldest record of the exotic mammal groups. In South America, the oldest of the immigrants from North America, the Procyonidae (raccoons), appear at 7 to 7.5 Ma in lower Chron 7 (Butler rt ab, 1984). In this study, the boundary between Huayquerian and Montehermosan land mammal ages was placed at 6.4 Ma. The second wave of North American immigrants occurs in sediments of the Uquian land mammal age dated by Marshall et al. (1982) from late Gauss to above the Olduvai. In a study of the type Chapadmalal, Orguira (1990) correlated this important migration to the late Gilbert; however, the reverse, normal, reverse (R-N-R) magnetostratigraphic sequence is more likely to correlate to the Kaena and Mammoth subchrons, which would place this migration event in the middle Gauss. In North America, the first appearance of South American forms occurs at the beginning of the Hemphillian; however, the most important migration event took place at thc Gauss/Matuyama boundary (Galusha et al., 1984).

156

9 Cennro~cTerrestrial Magnetic Stratigraphy

9.5 Notth Amencan and Eurasian Palcngenc

157

It would sccm that there is an assymmctry bctween migrations across the isthmus of Panama, with the most important immigrants such as Equus arriving in South America during the middle Gauss, while the South American migrants arrive in North America near the GaussiMatuyama boundary. These studies illustrate the power of the magnctostraligraphic method in resolving problems of intercontinental faunal exchange. An carlicr intrusion of fauna into South America has also been recognized and recorded in sedimcnts of the Deseadan land mammal age. This exotic fauna is of considerable interest since it contains caviomorph rodents and the first South American monkeys. This is a most curious interchange sincc it is hclicvcd lhal monkeys originated in Africa, and it is not clear how they managed to make thcjourney to South America. Early Oligocene dates on the Deseadan from South America compounded the problem, bccausc the descendant species in South America appeared to be older than proposed ancestors in Africa. In a study of the Salla beds of Bolivia, MacFadden er al. (1985) were able to redate the Deseadan land mammal agc and dcmonstrate that it was originally dated 10 My too old. They were abIe to place the first appearance of monkcysat 27 Ma and to demonstrate that rodents appear in South America prior to 28.5 Ma. This study addresses the ancestor problem but does not address the problem of how the animals managed to travel from Africa to South America.

9.5 North American and Eurasian Paleogene Figure 9.4 shows the magnetostratigraphic studies of the Palcogcnc and Late Cretaceous of North America and Europe ranked in time and identified with author and ycar of publication (Table 9.5). The boundaries of the land mammal ages in the Paleocene have been updated by studies of Butler and Lindsay (1985) and Butler et al. (1Y87), and their placement of Paleocene mammalian boundaries is used here. In cases whcrc individual sequences have been restudied and reanalyzed, the most up-to-date reference is given (Table 9.5). Mammalian chronology for the Late Eocene and Oligocene has been reviewed and revised by Swishcr and Prothcro (1990) and Prothcro and Swisher (1992). Our correlation of Duchesnian to Arikareen North American land mammal agcs follows Prothcro and Swishcr (1992) except [or some minor revisions. The boundary between the Duchesnian and Chadronian is placed within C17n (Fig. 9.4). 7his placement satisfies the magnetic stratigraphy and new radiometric results from the Vieja section of west Texas (Tcstamarte and Gose, 1Y79). Unfortunately, the magnetic stratigraphy for this section is poorly defined. Prothero and Swisher (1992) placed

*

F i m 9 . 4 TerrestrialPaleogcne magnetic stratigaphies. Xumhers refer to individual studies (see Tablc 9.5).

9.5

> d-

f

*

&>> >

d,L>

>

>

> m m

$

a , ++<

<

4 ,

h46

+c I

b<

.,

PI

!-
>>

> I

A,,

C

ZI?EYI

2

d

Y

l

i

= e S$Y

""

A,

E l

" m

" I

2

>

-

>

& > L

>a>

>;2 " 2b$ <

,

k >

db ric ke

d cd c dc

b

2

< + b

4

IZL

r

2

x,-

,

Y

<

4

' 1 4

ZEg

2

5

P

," r\

o

3

1

%E$ E -,

I ;X , Y

z

2

-

Y S

$63

*a-

**"

" ","=

:g

, ,

rn

s; " +" -

6

8

--"

E"E 2 g

22

a

-r

4

g g z 8~ (

I

,

:=-$

1

7"?;;9 cc jUL.

. " 9

;r ; -e " 9 - N

I

,

P

"P-

- - *

*

d m -

p 7 r

8

~

/

I

4ss-5 z r I

- - %

I

I

$ 5 5 ; 5*- -5-,9z: f T P 4

<,-= r L L

r . 0 -

0

m y -

ccp

8

=$A YE;

ccq ? qcz C " b

. ,,, 4 G g ;

*k G Sn

+>

V , ? _

- C

n

z:%

a&..

-1:

" $ 826 ;

d3 s a%

4

- --

ra;

M E "

22"

-

, 55 4 (

. 3 r R

r i m m

E

-- s p r p r a

=.,

-

- z ---'

---

2"~

G

- 2 :e

-

: 7 +

, e

U

+

g

N

w S n

+

,

+ I 1

m" 2 n n * +

m e i

t

+

,

c rQ q q i "! " ;9zz (,(,>

*

o

.

I

":2

.+

?

#

Z -~5 L .;

-

C

I

2 ~ 25

.-

3

T

rr

"a-

$2

-m

--=

5-$

& L &

542

* -n,

-,

7

+

?," I

+

+

*, 3zrj SSPJ

7

2a :.Z G

.- ,2 a2 F.E5^5

g=

B'e,f$

5

5

z:z

-Erf

srn=

=::

;g$

Nonh Amencan and Eurasian PaIeogcne

159

the boundary within C16n on the basts of the radiometric dates and an early verslon of the Cande and Kent (1992a) time scale The Whitneyanl Ar~kareenbounda~yis placed at the C11IC10 boundary based on magnetic stratigraphy from the Toadstool Park-Roundtop section in Nebraska (Prothero and Swisher, 1992). The correlatton of the magnetic stratigrdphies to the GPTS (Fig. 9.4, Table 95) IS supporied in most cases by radiometric ages. An important magnetic stratigraphy from the Devil's Graveyard Fm of west Texas was correlated to MO (Walton, 1992). Amore reasonable correlation, supported by the Bridgenan to Duchesntan fauna, would be to correlate the section to the C21n to C17 interval. One of the problems that can be addressed by magnenc stratigraphy is the heterochroneity (time transgresston) of faunal datums. The land mammal age5 In North Amertca are based on stage of evoluhon and appearance of imm~granttaxa. Magnetic strattgraphy can test this hypothesis because the evolution and dispersal of mammals and reversals of the geomagnetic field are independent. Two studtes in parttcular have raised the possibrhty that faunal datum.; in North America of Paleocene age may be time transgressive. The first of these studtes is an extensive investigation of Cretaceous and Paleocenc sehments in the San Juan Bastn of New Mcxico (Butler et u l , 1977). Thls regton has a record of the turnover from the dinosaur fauna of the Cretaceous lo mammal faunas of Paleogcne age. The onginal interpretation correlated the Torrejontan faunas of the San Juan Basin to C26. The same authors (Butler et a l , 1981a) correlated the Tiffaulan of the Clarks Fork basm, Wyoming, lo C26. Thts led to the anomalous sttuation in which two distinctly diCferent mammal faunas, ihought to be sequentla1 in time, appeared to be coeval in the western Unlted States. Further study by Butler and Lindsay (1985) resolved the problem by showing that one of the normal polarity zones orignally proposed in the San Juan Bastn was caused by a viscous magnctic overprint. When this spurious polartty zone was eliminated, agreement was achteved between the two studies. The corrected stratigraphy leads to the placement of the CretaceousTertiary (KIT) boundary in reversely magnetved sediments of C29r. All other K f I boundar~esoccur in reverse polarity sedments cons~deredto correlate to C29r (Archibald et al., 1982; Lerbekmo et al. (1979), except for a core in New Mexico's Raton Basin wh~chwas studled by Payne et al. (1983) and was s a d to be normally magnetized. Tauxe and Butler (1981) cite a contrary optnion (as a personal communication by E. M Shoemaker) that the core is entirely reverse polartty. One hopes these data are pubhshed so that the results can be properly evaluated. Withln

160

9.6 Mammal Dlsper\al m the Northern Hem~sphere

9 Cenozoic Terrestr~alMagnetic Stratigraphy

the resolving power of the palcomagnetic method the K/T boundary appears to be synchronous. between marine and nonmarine sediments. A second study which initially seemed to indicate faunal diachroneity was a study of fauna and flora in Arctic sediments by Hickey et al. (1983). These authors originally claimed that planls and animals appeared in Arctic sediments several million years prior to their appearance in temperate North America. Thc Eureka Sound Formation, which yielded these results. has been restudied paleomagnctically by Tauxe and Clark (1987). who showed conclusively that the original study was flawed because of the incomplete removal of magnetic overprints and came to the conclusion that no diachroneity is apparent. The results of these studics confirm that land mammal ages are, indeed. useful chronostratigraphic units (Flynn et ul., 1984: Clyde el aL, 1994; Tauxe a al., 1994).

9.6 Mammal Dispersal in the Northern Hemisphere The fact thal mammals disperse from continent to continent is well known. and modern continental faunas are greatly influenced by these processes. The interchange of faunas betwcen the continents has been of great utiGty in the construction o l mammalian biochronologics within the different continents, particularly in North America and Europe. The ability of animals to move fromone continent to another is affected by the availability of corridors that allow animals to pass. Plate tectonics has had a first-order effect on thc inlerplay of faunas, since the different plates are in constant motion, changing geographic relationships between continents (McKcnna, 1975). The Age of Mammals began at the inception of the Cenozoic, at 65 Ma. Plate tectonic history since that time has been largely one of continental dispersal, which had led to the partial isolation of large continental areas such as Australia and South America. The other continents have a history of intermittent contact and subsequent isolation. Tectonic activity resulting from collision will tend to create links between continental blocks. For example, collision of Indiawith Eurasia during the Cenozoic and the continuing collision of the African plate with Europe provided opportunities for faunal interchange (McKcnna, 1975). Faunal dispersal is affected not only by changing paleageography but also by paleuclimatology, which can erect or destroy ecological boundaries, either expediting or impeding the migration of animals conlined to specific habitats. Insome cases raunas that appear in midcontinental North America and Europe may have been displaced climatically from the northern part of the continents, as they sometimes have no close relatives in either North America or Eurasia. Palenclimatology can also create or destroy dispersal

.

161

routes by glacioeustatic changes in sea level. These changes can occur rapidly, as has been demonstrated from the marine SIROrecord o l Pleistocene ice volume change, which implies rapid sea level falls of up to 120 m (Shackleton and Opdyke, 1973). This climatic record can be correlaled to the terrestrial faunal record by magnetic stratigraphy. Barry et 01. (1985) claimed that evolutionary events and dispersal first appearances in the Siwdliks could be directly related to the 8% record of the middle Miocene. It is interesting to speculate whether this correlation can be established outsidc the Indian subconlincnt. The change in S1'O for benthicCoraminSera from the AtlanticOcean (Fig. 9.5) (Millererul., 1991a) shows the overall climatic trend for late Eocene through Miocene time. Mammal migrations have occurred bctween North America and Eurasia throughout the Cenozoic (Fig 9.5), and indeed the geochronologic units (land mammal ages) were established using immigrant taxa as part of their definitions (Wood et at., 1941: Woodburne, 1987). It might be expected that the beginning of each new land mammal age would correlate with sea level drops bccause lowering of sea level would facilitate dispersal. It is unlikely, however. that all dispersals correlate with sea level change, since. in some cases, the corridors might stand so high a b o v ~ s e alevel as to be unaffected by that change. This is probably the case in the early Eocene when North America and Europe wcre connected (in thc region of thc present Norwegian-Greenland Sea), allowing unhampered migration between the two continents. Opdyke (1990) argucd that intercontinental mammal migration has resulted from sea level drops following the development of continental ice sheets. Since Neogenc land mammal ages are often defined on the basis of immigrant taxa, it was suggested that Ncogene land mammal ages in Europe and North America should be closely correlated if both were defined on the basis of i r n m i ~ a n taxa. t Miller er al. (1991a) have presented a new and updated 61X0record for the Oligocene and Miocene and argue that continental glaciation began in Antarctica in the latest Eocene (-35 Ma). These authors have formally defined nine oxygen isotope zoncs in the Oligocene and Miocene which they have designated Oil and Oi2 and Mil through Mi7, ranging in age from thc early Oligocene lhrough the early upper Miocene (Fig. 9.5). Miller er al. (1991a) argued thal the prePleistocene glacial stages may have been associated with glacioeustatic falls in sea level of 55 to 1811 m; however, they argue that the actual maximum drop in sea level in unlikely to have been as large as Pleistocenc values of 120111. The fall in sea levcl postulated above, will, of course, cause the development of unconformities on passive continental margins (Haq er al., 1987; Christie-Blick er ul., 1990). Pitman (1978) has argued that the rate of sea level change also controls the position of passive margin unconformities.

162

CCnozolc l'errestnal

Magnetic Stratigraphy

9.5 Carrelation uf mammal migrations to the GPTS (Cande and Kent, 199s) and oxygen isotope events (Miner et aL, 1YVla). Positions in lime of major mammal migmtiuns are indicated by arrows with the nurnher of immigranr taxa. Figure

Christie-Blick et al. (1990) have suggcsted that the timcs of most rapid drop in sea level should coincide with the formation of unconformitics. These passive margin unconformities may he detected using seismic reflection data (Vailetal., 1977; Haq etal., 1987). Evidence for sealevelfalls,combined

i

9 6 Mammal Dlapeml m the Northern Hemlrphcre

163

with physical stratigraphy. has been formally organized into sequence stratigraphy. If there is a causal relationship between icc volumc changes and sequence stratigraphy, then a corrclatiun should cxist between the isotopic record and sequence stratigraphy. Miller et al. (1991a) maintain that such a correlation does, in fact. exist in the Oligocene and Miocene. However. problems in correlating sequence. straraligraphy to pelagic isotopic records will remain until the magnetic stratigraphy 01continental margin sediments is rigorously documented. Webb and Opdyke (1995) ha%screexamined tho question of migratory events in North America and divide the migratory first appearances into first-order (>8 taua) and second-order (>5 taxa) migratory events. Cenozoic faunal migration to and from North America and Eurasia is essentially divided into two segments. the first taking place in the Paleocene and Eocene and the second being Miocene and younger. McKcnna (1975) and Opdyke (1990) have pointed out that tectonics and paleogeography are the primary controls on mammal migrations whereas climate-driven glacial eustasy is a secondary cffect made possible by changes in paleogeography. The large &st-ordcr migratory patterns are undouhtedly caused by changes in paleogeography. It is probable that throughout the ~al.?ogenc,mammals could pass between North America and Europe over the Thulean land bridge across the Norwegian-Greenland Sea. The route probably remained open until the rifting apart of Svalbard and Greenland, probably at the end of the Eoccne (Rowlcy and Luttcs, 1988). The major drop in the temperature of bottom water which occurs in the world ocean at the beginning of the Oliguccne (Oil. Fig. 9.5) is not accompanied by any important mammal migration. The reason is undoubtedly that migration across the Thulcan land bridge was n o longer possible. The other possible route from Eurasia l o North America, the Bcrirtg land bridge, presented n o apparent barrier to mammal migration during the Paleogene. Howrvcr, at the cnd of the Mesozoic the Bering region was at paleolatitudes above 80°N, almost on the North Pole. The Bering land bridge region has been slowly moving into lower latitudes during the Cenozoic. It seems possible that even though the region was moving toward lower latitudes (70"N), the climate was continuing to deteriorate at the end of Eocene and early Oligocene lime as indicated by icc-rafted debris around the Antarctic contincnt. It secms reasonable that the barrier that impeded mammal migration across the Bering land bridge was climatic. By Miocenc time, the Bering region probably reached sufficiently low latitudes to allow mammals to cross at times of sea levcl drop caused by glacial eustacy. Thc late Eocene "Duchesnian" interchange at about 40 Ma correlates with an unconformity on the continental margin (Miller el al., 1987) and with a major offlap event in the Vail curve (TA 4.1). In addition, this event

has also been correlated with an increase in isotopic values in both planktic and benthic foraminifera (Miller rt al., 1987) and with an cstimatcd sea level drop of about 80 m (Prentice and Malthews, 1991). The next major migration event occurs in the middle Oligocene in C9r (R. H. Tcdford, personal communication. 1994) where seven new immigrant species first appear coincident with, and following. isotopic evcnt Oi2 (Fig. 9.5). The third major migratory event takes place in thc lale Arikareen and early Hemingfordian from 21 to 18 Ma. This is the most important period of immigration in the Cenozoic. These migratory events correlate to isotopc events Mila and Milh (Miller et al., 1991a). At least four positive oxygen isotope changes in the 18-21 Ma interval suggest that thc lows in sea levcl at this time were comparable lo thosc in thc lalc Pleistoccnc (Prcntice and Matthews, 1991). Two sequence boundaries occur in this interval (Haq al., 1987), a type 1 sequence boundary at 21 Ma and type 2 scqucnce boundary at 17.5 Ma, and it is probable that this migratory event is cuinuidcnt with eustatic sea level fall. The faunal migrations within the late Mioceue occur within the Hrmphillian [.MA. The Hemphillian is characterized by three second-order migratory evcnts at -9.7 to -6 Ma, involving 6 or 7 taxa (Tedford et a/., 1987). The first oI these occurs near thc beginning of the IIemphillian in C4Ar.2 at about 9.5 Ma, wrrclating with 8IxO isotope event Mi7 (Wright and Miller, 1992);dnd with a typc 2 scqucncc boundary in C4A (Haq et a/,, 1987). The second of the migratory cvcnts occurs at 7 Ma, roughly coeval with the Late Miocene carbon shiit and a wcll-dcfincd 8''O event (Hodell et a/., 1994). The final migratory event within the Hemphillian is believed to have occurred at about 6 Ma, which is close in time to the position d Gilbert oxygen isotope stages. TG20 and TG22, astronomically dated at 5.7 and 5.8 Ma (Fig. 6.7) (Shackleton er aL, 1995b). Lindsay er al. (1984) ha\% placed thc HemphillianIBlancan boundary, which coincides with a first-order mi-' gratory event. between the Sidufjall and Nunivak subchrons, Tedford et al. (1987) placed the event at about 4.9 Ma, coincident with isotope stages Si, and Sib within the Sidufjall subchron (Fig. 6.7). The final episode of faunal migration is well known to be at, or close to, the GaussIMatuyama boundary at 2.5 Ma, and is clearly associated with the onset of the northern hemisphere glaciation (Fig. 6.7) (Shackleton et ai., 1984). A t this time the horse, Eqnus, migrated to Eurasia and spread rapidly. Lindsay et al. (1980) proposed this correlation from the Siwaliks of Pakistan (Opdyke et ui., 1979), and further research in Eurasia has tended to support the appearance of Equus at or near the Gauss-Matuyama boundary. Elephas appeanin sediments of middle Gauss age in the Siwaliks,

9.6 Mammal Dkpersal in the Northern Hemisphere

as well as in Eurasia and China (Opdyke er al., 1979; Tedl'ord etal., 1991). Faunal cxchange at this time is very important between South and North America, and the culmination of movemcnl of fauna from South America to North America is correlated Iu the Gauss-Matuyama boundary in North America (Galusha er al., 1984). Magnctic straligraphy has proved to be of great utility in dating and correlating the scdirnenlary record of tectonic events (Table 9.6). Onc of the best examples is the study of Burhank and Raynolds (1984, 1988) in which they were able to date the tcclonic developmcnl of the Himalayan

167

front using the magnctic stratigraphy of theredimenls of the Potwar basin (Fig. 9.6). This type of study has expanded to other parts of the world such as the Pyrenees (Burbank et al., 1992b.c) and Argentina (Jordan et ul., 1990). Magnetic stratigraphy has also bcen very important in studies of basin formation and subsidence (Burbank, 1983; Burbank and Johnson, 1983) and for dating of uplift episodes in mountain ranges (Cerveny el a[,, 1988). Magnetic stratigraphy has also enabled scientists to study the sedimentation process itself and to address problems of straligraphic completeness (Johnsoh etal., 1988; McRae, 1990a,b; Badglcy and Tauxe, 1990; Friend et al., 1989).

-

10 1 Oceann~Magnettc Anomaly Record

169

! KFATHLEY LI NEAT1 ONS

!

Jurassic-Early Cretaceous GPTS

27.00 26.00

Hawaiian

synthetic profile

Hawallan Dlock model

10.1 Oceanic Magnetic Anomaly Record When thc Hcirtzler et al. (1968) geomagnetic polarity time scale (GPTS) was constructed, it was noted that magnetic anomalics older than anomaly 32 have lower amplitude and are poorly lineatcd, in both the Atl.dntic and Pacific Occans. The lack of lineiiled magnetic anomalies in this so-called Cretaceous quiet zone implied either that there were n o reversals of the Earth's magnclic field for a considerable length nf time or that the magnctic properties of the sea Aoor were such that field reversals werc not recorded. The weight of evidence which became available from magnetostratigraphic studies of Crelaccous rocks on land (Ch. 11) dcmonstrated that the quict zone in the oceans corresponds lo a time of few or no reversals of the geomagnetic field. In the Atlantic, a Mesozoic sequence (M-sequence) of correlatable anomalies prior to this "quiet zone" was identified and n a m d the Kcathly sequence by Vogt rt ul. (1971) (Fig. 10.1). Larsnn and Chase (1972) succeeded in correlating the Japanese, Hawaiian, and Pheonix Msequence lineations, and Larson and Pitman (1972) were ahlc to correlate these with the Keathlcy scquence in the Atlantic. Thc key to this correlation was the recognition of the fact that the Pacific anomalies had been formed in the southern hemisphere and transported into the northern hemisphere hy plate motion. These correlations led to the extension of the GPTS into the Middle Jurassic. The anomalies in the mid-Mesozoic M-sequence were initially numhered from M1 to M25. Unlike the Late Cretaceous-Cenwoic part of the GPTS, anomaly nurnbcrs were mainly assigned to reverse polarity anomalies (Fix. 10.2); however, the labels M2 and M4 were assigned to normally

Figure 10.1 The Atlantic Kcathley lineations correlsfed to thr Pacific (Ha*aiian) block model (alter Vogl el al,, 1971: Larron and Pirman. 1972).

magnetized anomalies, resulting in a rather contusing situation. The magnetic anomaly hetwcen M10 and M l l is labeled M1ON. This anomaly was originally omitted from M-sequence anomalies (Larson and Pilman, 1972) and was included by Larson and Hilde (1975) (Fig. 10.2) and designated MlON after Fred Naugler, co-chief scientist of the NOAA Western Pacific Geotraverse Projcct. Jurassic anomalies older than M25 were suhsequenlly identified and the anomaly sequence was extended to M2Y (Cande era/., 197%). Recent accumulation of oceanic magnctic anomaly data in the MO-M29 interval from the Pacific has led to lhc. development of new block models for the Japanese, Hawaiian, and Phoenix lineations (Fig. 10.3) (Nakanishi et nl., 1989. 1992; Channel1 i.r nl., 1995a). Difficulty in correlation and low amplitude of oceanic magnctic momalies older than M29 led to the concept ol a "Jurassic quiet zone"; however. aeromagneticprofiles from the western Pacific have succccdedinidentifying lineatcd anomalies older than anomaly M29 and, as a result, the GPTS has bcen extended toanomaly M38 (Handschumaker etal,, 1988).The magnetic anomaly profiles from the western Pacific imply that the so-called Jurassic quict zone is an interval of high reversal rrequency. These magnetic anotnalies are. however, of much lower amplitude than younger magnetic anomalies, and the reason for this remains unclear: The M-sequence marine magnetic anomalies recorded by Larson and Hilde (1975) (Fig. 10:2) in the Hawaiianlineations have provided the means

170

10 Jurassic-Early Cmfaccous GPTS

10.2 Numerical Age Control

HAWAIIAN LINEATIONS PROJECTED PROFILES

f"LMM,s

-0 300

KM

C1219 050-

COMPOSITE

V2404

SITESITE 156 303

S I 3 5

. 1

, . d , , . I . . 171, M ,,,,, > ,on

SITE 100

'SY

,m

((0

30

SITE

b.3 7

304

PACIFIC MODEL

Figure 10.3 P~nfilesused lo consrrua ihe Pacific (Japanerr. Hawaiian, Phoenix) block models in Channcll el ol. (19Yja). Thc calculaled pgmtiic is the stimulation from ihe Larson and I lilde (1975) hlock model extendcd tu M29.

Figure 10.2 Magnet~canomalvprofiles used toconstruct t h e l m o n dnd H ~ l d c(1975) Hawanan block model (atter Larson and Hllde. 1975)

10.2 Numerical Age Control o l inlerpoIation between age estimates for most Oxfordian to Aptian tiple scales. The time scales of Kent and Gradstein (1985) (KGX5) and Harland el a/. (19W) (GTSX9) imply constant spreadingrate in thc MO-M25 Hawaiian oceanic anomaly hlock model of Larson and Hilde (1975) (LH75). The KG85 time scale uses the constant spreading ratc assumption to interpolate betwecn 119 Ma for the BarremianIAptian boundary (base CMO) and 156 Ma for the Oxfordian-Kimmeridgian boundary (CM25n). The GTS89 time scale is hased on chronogram ages for the Late Jurassic-Early Cretaceous stage houndarics. Thcsc agcs were used to make a linear recalibration of the KG85 time scale. thereby inheriting the constant spreading rate assumption. The chronograms of GTS89 for this interval are poorly constrained and therefore do not validate the constant spreading ratc assumption.

Sincc the KG85 and GTS89 time scales werc published several numerical estimates for the age of M-sequence chrons have bccome availahlc. Mahoney et al. (1993) obtained a 4 0 ~ r i " A rage or 122.3 f 1Ma for the basaltic basement of ODP Site XU7 on the Onlong Java Platcau. Tardunu ct 01. (1991) havc argued that Ontong Java volcanism is constrained to a short interval in the Early Aptian; therefore this age cstimate shouldhe applicable to the Early Aptian. This interpretation was based on ihc observation that seliments overlying basaltic bascment, and volcanoclastic sediments at distal sites, belonged to the Glohigrrinelloides hlowi planktonic foraminiferal zone, which was considered to b e confined to the Early Aptian (Sliter, 1992). Lower Cretaceous planktunic foraminifera1 hiostratigraphy in s e c tions dated with ammonites and magnetic stratigraphy (Coccioni et a/., 1992) has, however, indicated that the C;. hlowi zonc extends into the

172

10 lurasslc-Early Cretaceous GPTS

Barremian. Thc (>. blowi biozone does not. therefore, rcstrict Ontong Java volcanism to the Early Aptian. A t Site 878 on MIT Guyot, the oldest dated sediments (at 25 mm above basement) are from the lowcr part 01 the C, lirterarius nannofossil zone, based on the first occurrence (FO) of R. irregulurl\ and their location below the "nannoconid crisis." Pringle er a[. (1993) obtained a *)Ad3'Ar a_ee of 123.5 t 0.5 Ma (cnnvcrtcd lor standard Mmhb-l at 520.4 Ma) from these basalts. The remanent magnetization in the basaltic basement indicates a reverse polarity zone overlying a normal polarity zone. The reverse polarity zone is probably torrelalive to CM1.

1 120 125 130 135 140 145 150 155 Ages (Ma)

Coleman and Bralower (1993) obtained a U-Pb zircon age of 122 ? 0.3 Ma from a bentonite in the Great Valley Group (northern California). This level has been correlated to the C. litterurius zone based on the FO of C. litterarius, however X. irregularis (a more useful marker for the Barremian-Aptian boundary, see Fig, 11,6) has not been found in the section (T. J. Bralower, personal communication, 1993). C. litterarius has been reported from several Barremian sections and the 122 ? 0.3 Ma bentonite layer from the Great Valley Group may be Barremian. The 4"Ar/39Arage of 124 t 1 Ma obtained from reversely magnetized granitic plutons lrom Quebec (Foland er al., 1986) may be correlative to CM3, or tothe shorter duration CMO or CM1. Ten major igneous complexes with 4"Ar/39Arages tightly bunched around 124 Ma record only reverse polarity (Foster and Symons. 1979), and it seems likcly that the reverse polarity chron is CM3, which has more than three times the duration of

, 0 0 1 120 125 130 135 140 145 150 155 Ages (Ma)

Id

500 ' 1 i 0 ' 115 ' Ages (Ma]

120 125 130 135 140 145 150 155 Ages (Ma)

Figum 10.4 i h c selected radiomctric age determinations for MD (121 Ma), Mlb-Mli (137 Ma). and hT25 (154 Ma) plotted againat Ihc (a) Lamon and Hilde (1975) block model. (b) Hawaiian block model; (c) Krathlry hluck modcl, ~ n d(d) Japanese block model. MD is not prcscnt in the Japmese lineatiote and fix fd) the yuung cnd of MI is assigned an age of 123.2 Ma (after Channel c.1 a / , 1945a).

OBRADQ3 (Ma) Figure 10.5 OBRAUY3 time scale (Obradovich,1993) wmpared ro occanicm.xgnetic anomaly b l ~ models k

174

10 Jura\nc-Early Cretaceous GP L'S

any reverse polarity chron in the younger part of the M-sequence, The duration of the interval between CMO and CM3 is abnut3.5 My (Herbert. 1992). leading to an age for the base of CMO of about 120.5 Ma. A U-Pb zircon age estimate of 137.1 (+1.6!-0.6) Ma lrom the Great Valley Sequence of norlhem California has been correlated to polarity chron CM16or CM16n (Bralower era!., 1990).This is an indirect correlation to the polarity time scale using nannofossil stratigraphy. The interval containing the two dated tuff layers was attributed to the Upper Berriasian Assipetra infi-ncretaceusubzone based on three nannofossil events: (1)the first occurrence of Cretnrhnbdus angustiforatus, which was used to define the base of the A. infracretacea subzone, (2) the occurrence of Rhagodiscus nehulosus, and (3) the absence of Percivaliufenestrara. Close inspection of the ranges of the three marker species (Fig. 11.6) suggests that the correlation of the two

1M)

---c Hawaiian 150

140

a

The age eslimates mentioned above for the base of CMO (121 Ma). CM16CM15 (137 Ma), and the top CM25 (154-156 Ma) are not consislent with constant spreading rale in LH75 (Fig. 10.4a). Obraduvich (1993) used the LH75 block modcl t o interpolate between the same three age estimates: CM0 (121 Ma), CMl6116n (137 Ma), CM25n (156 Ma). The resulting time scalc (OBRADY3) implies an abrupt spreading rate change at CMl6 in LH75 and uthcr Pacific block models (Fig. 10.5). Gradstein er al. (1994) have presented an integraled time scale (GRAD94) fur Triassic to Cretaceous time. For the Late Jurassic-Early Crctaccous interval, constant spreading rate was assumed for most of the LH75 block model with an increase and then decrease in spreading rate in Velanginian and Berriasian time in order lo satisfy the three principal age constraints. S.qnchronous apparent spreading rate changes in different blockmodels may bc an artifact of GRAD94 (Fig. 10.6). A new Hawaiian block model (Nakanishi e f al., 19x9, 1992: Channell et ul., 1995a) may represent an improved estimalc of a constant spreading rate record for the MO-M29 interval. Thc three popular ages tor CMO,

-5 z (0

-2 VI

3

110

120

125

130

135

140

145

150

155

volcaniclayers to CMl6 or CM16nis too restrictive and that the CM 16-CM15 interval may be more appropriate (see Channell et ul., 1995a). The age of 154Ma for the young end of CM25 is basedon thccorrelation of this polarity chron boundary to the Kimmeridgian-Oxfordian buundary (Ogg et al., 1984) and to age estimates for this stage boundary of about 154 Ma from California (Schweickert er al., 1984) and Oregon (Pessagno and Blome, 19%). In the California Sierra Nevada, the ammonitebearing Maripusa Formation is a synorugenic Rysch supposed to contain the Oxfordian-Kimmeridgian boundary (Irnlay, 1961) and is affected by Nwaden orogeny. K-Ar hornblende ages and U-Pb zircon ages on dykes and plutons appear to constrain the age of the Mariposa Formation and the Nevadan orogeny to the 152-159 Mainterval (scc Schweickert etal., 1984); however, many of the age determinations arc not of high quality by modern standards. More recent "OAri3%r and concordant U!Pb zircon ages from the Klamath Mountains apparently constrain the OdordianiKimmeridgian boundary to the 150-157 Ma interval (see PessGno and Blorne, 199U). The definition of the OxfordianiKimmcridgian boundary in the Klamath Mountains is based largely on the overlapping occurrence of the radiolaria M i r w u s and Xiphostylus. This is a highly conlrovcrsial definition as Mirifrrsus and Xiphostylus are both present from Aelenian time in Mediterranean sections (see Baumgarlner. 1987).

10.3 Oxfordion-Aptian Time Scales

>

120

175

& E

130

10 3 Onford~an-AptlanTime Scales

1W

1W

GRAD94 (Ma) Figure 10.6 GKADY4 lime scalc (tiradsleln ef n l . 1994) cumpnrcd 14 ocean,' magnctlc anomall b l o ~ kmodel%

CM16-15, and CM25 are consistent with constant spreading in the new Hawaiian block model (Fig. 10.4b) and the resulling time scale (CENT94) does not imply abrupt or synchronous changes in spreading mte in the Pacific block models (Fig. 10.7). Figure 10.8 illustrates the relationship between CENTW andother populartime scales. For numerical agesorlate JurassidEarly Cretaceous polarity chrons and geologic stage boundaries. according to thc various time scales, see Tables 10.1 and 10.2.

10.4 Hettangian-Oxfordian Time Scales The Kent and Gradstein (1985) timc scale for the Hettan~an-Oxfordlan intervals was constructed by interpolation betwcen 208 Ma tor the bare of the Hettang~an(Armstrong, 1982) and 156 Ma for thc top ol the

120

130

140

150

160

CENT94 (Ma) Figure 10.8 PLn1 or vanou? hme tcalec, agamst CEYT94 LH7i (Larion and Hdde. 193, KG85 (Kmt and Gradsiern. 1985) GTS89 (Harland er 0 1 , IYYU), tiKADY3 (Gradstcm cr a l , 1995) and GRAD94 (Gradstem cr a l , 1994). CENlY4 (ChanncU ?r n l , 199%)

lowl'''l''''l''''l''''l,,,,l,,,,l,,,,l,,,,ll~ 120

125

130

135

140

145

150

155

160

CENT94 (Ma) Figure 10.7 CENT94timescals lChanncllr!ral., 1YYSa)comp~~edtooczanicmagnzticant~maly

block modclr.

Oxfordian, assuming equal duration of the intervening 50 ammonite zones from Hallam (1975) (Table 10.2). In this interval, the GTS89 time scale (Harland cr ul., 1990) utilizes the same age tie points (156 Ma for the top of the Oxfordian and 208 Ma for the base of the Hettangian) and thc samc slralcgy of equal duration of ammonile zones; however, the use of a different ammonite zonation (Cope et ul., 198na,b) leads to slightly different stage boundary age estimates. In GTS89, the Bathonian and younger Jurassic stage boundaries were also determined by interpolation from the magnetic anomaly record. The stage boundary age estimates derived assuming equal duration of ammonite zones, from magnelic anomalies, and from chronograms are listed in Table 10.2. For the Hcttangian-Oxfordian interval in the time scale of Gradstein et ul. (1994), maximum likelihood stage boundary age estimates were derived

178

I0 Jurassrc-Early Crrlacruus GPTS

'

Toble 1 0.1

~ a b l s10.1 continued

Compnrbon of M Sequer~eTime Scales Reversed chron

M0 (rcrp)

(base) MI M3 MS

M6 M7 MR M9 MI0 MIONn-1 MlONn-2 MIUN MI1 MI 1

179

10.4 Hcttangian-Oxfordian Time Scales

LH75 (Ma1

KGRT

CllSXU

(Ma)

(Ma)

108.19 l09Ul 112.62 llll4 '1 14.05 11675 118.03 118.72 118.41 119.06 119.27 119.79 IZ0.21 120.52 120.88 121.49 121.94 122.37 13-82 122.8X 121.31 123.34 723.73 124.07 125.10 125.68 125.74 126.22

118.00 118.70 121.81 122.25 13.03 12536 126.46 127.05 127.21 127.34 127.52 127.97 128.33 128.60 128.91 129.43 129.82 13u.19 130.57 13U.63 131.00 13102 131.36 131.65 132.53 133.03 133.08 133.50

124.32 124.88 l27.35 127.70 128.32 130.17 131.05 131.51 13164 131.74 131.89 132.15 132.53 132.75 132.99 133.41 113.72 134.n1 13431 134.36 134.65 134.67 134.94 135.17 135.92 136.27 136.31 136.64

MIIAn-1

GRAD93 (Ma)

GRAD94 (Ma)

119.15 120.10 122.56 722.88 123.45 125.15 125.96 126.39

120.38 120.98 123.67 124.04 124.72 126.73 127.68 128.19 128.33 128.44 128.59 128.98 120.29 129.53 129.79 130.24 13U,58 13u.90 131.23 131.28 131.60 131.62 131.91 132.10 132.70 133.05 133.08 133.37 133.M

126.60 126.80 127.25 127.68 127.98 128.33 128.88 129.31 129.71 130.13 130.20 130.60 130.62 130.99 131.26 131.81 L'32.12 132.15 132.42 132.66

m ~ m 4 (Ma) 120.60 121.U0 123.19 121.55 '124.05 125.67 126.57 126.91 127.11 127.23 127.49 127.74 128.07 128.34 128.62 128.93 129.25 129.63 129.YL 129.95 130.22 130.24 130.49 130.84 131.5U 131.71 131.73 131.91

LH75 (Ma)

KG85 (Ma)

GTS89 (Ma)

GRAD93 (Ma)

GRAD94 (Ma)

M17

136.42 138.16 138.82 13931 139.46 139.55 140.74 14128 141.64 141.71 142.47 143.47 144.74 145.29 147.11 147.17 147.23 147.29 147.38 14836 14851 148.72 149.15 149.48 149.51 150.24 150.63 151.06 151.09 151.33 151.48 151.79 152.21 152.39 152.73 153.03

142.27 143.76

143.61 144.80

138.51 t3q.86

140.86 142.51

144.33 144.75 144.88 144.96 145.98 146.44 146.75 146.R1 147.47 148.33 149.42 149.89 151.46 151.51 151.56 151.61 151.69 152.53 152.66 152.84 153.21 153.49 153.52 154.15 154.58 154.85 154.88 155.08 155.21 155.48 155.81 156.00 156.29 156.55

145.25 145.58 145.69 145.75 146.56 146.93 147.17 147.22 147.75 148.43 14930 149.67 150.92 isn.96 151.W 151.M 151.10 151.77 151.87 15201 152.31 152.53 15a.56 153.06 153.32 153.61 153.64 153.80 153.90 154.11 154.40 154.53 15476 154.96

140.33 14068 140.80 140.88 141.82 142.25 142.54 142.60 143.21 144.01 14592 145.46 146.92 146.95 147.01 147.06 147.13 147.91 148.03 148.20 14k.W 149.24 149.30 150.31 150.85 151.44 151.49 151.82 152.02 152.46 153.04 153.30 153.52 153.72

143.14 143.63 143.72 143.79 144.68 145.08 145.35 145.41 145.99 146.74 147.69 148.11 149.48 149.52 149.57 149.61 149.68 150.42 150.53 150.69 151.09 151.39 15142 152.10 152.46 152.86 152.89 153.11 153.25 153.54 153.93 154.10 154.31 154.49 155.51 155.69 155.83 156.00

138.89 140.51 141.22 141.63 141.78 141.88 143.07 143.36 143.77 143.M 144.70 145.52 146.56 147.06 148.57 148.62 148.67 14R.72 148.79 149.49 149.72 150.04 150.69 150.91 150.93 151.40 151.72 151.98 152.00 152.15 152.24 152.43 15313 153.43 154.00 154.31 155.32 155.55 155.80 156.05

156.14 156.29 156.77 156.85

156.19 15651 157.27 157.53

hi18 MlYn-1 M19 Mzon-l M20 M21 M22n-1 M22n-2 M22 M22A M23 M23 M24 M24 '

CENT94 (Ma)

Reversed chmn

M24A

MllA M24B M25 M26 M27 M28 M29

Key LH75, Larson and Hildc (1975): KG85, Kent and Gradstain (1985); GTS89, Harland el 135.94

141.85

143 28

llR2h

14114(1

1?K i l l

o l (1990): GRADY3, Gradstein el nl. (1995); GRAD94, Gradslein er al. (1994); and CENT94. Channel1 N a1 (1995a).

CRETACEOUS

20

r-7

Jurassic and Cretaceous Magnetic Stratigraphy

1 1.1 Cretaceous Magnetic Stratigraphy The majority of Cretaceous magnetostratigraphic studies havc been carried out on pelagic limestones i n the Mediterranean region, following rhe pionecring work of Lowric and Alvarez (1977a,b) on the Guhbio (Bottacione) section in Italy. ThE principal feature of Cretaceous geomagnetic polarity is the so-called Cretdceouslongnormalsuperchron,aninterval of prolonged normal polarity lasting about 38 My, beginning just above the Barremiant Aptian boundary and ending close to thc Campanian/Santonian boundary (Fig. 11.1.Table 11.1). a. KiT

Boundary Magnetic Stratigraphy

The position of the Cretaceous-Tertiary (KIT) boundary in the GPTS has always been of great interest becausc of the associated mass extinctions. The position of this boundary in magnetostratigraphic section was first determined a t Gubbio (Bottacionc) (Alvarez eta/., 1977). Since then, thc position of this boundary has been accurately determined in morc than 14 magnetostratigraphic sections both in outcrop and dccp-sea cores. Thc avcrage position of the K/T boundary in these studies implies a correlation to C29r (0.75) (75% u p from base of the i'evcrsc chron). However. the location of the boundary within the polarity zone correlative to C2Yr is not a good estimate of thc position of boundary in rime due to a substantially reduced sedimentation rate in the Paleocenepart of C2Yr. Cyclostratigraphy

Ffigure11.1 Summzryr,fiome ofthcmnrc~mporlan~ Crstaceousm~ncruirm1~gr~phcsiud1es For key sec 1ahlr 11 I

187

186

I1 JuraSslc and Crctaccous Magnetrc Stratigraphy

111 Cretaceous MagnetLC Strangraphy

based on carbonate contcnt, color dcnsity, and magnetic susceptibility From South Atlantic corcs and Spanish land ,sections was used to demonstrate that the boundary occurred in time almost exactly in the middle of C29r (Herbert and D'Hondt, 1990; Herbert er al., 1995). Nonmarine terrestrial sediienls from the western United Statcs (Butler er t ~ [ . ,1977, 1981b) as well as from Europe (Galbrun et at., 1993) have contributed t o the debate concerning the faunal crisis at the cnd of the Crctaceous by demonstrating that extinctions of marine and nonmarinc fauna wcre essentially synchronous (see Section 9.5).

b. Santonian-Masstrichtian Magnetic Stratigraphy The Latc Cretaceous polarity zoncs at the young end of the Cretaceous long normal interval were recorded and correlated to foraminiferal biostratigraphy in the original Alvarez eral. (1977) study of the Gubbio (Bottacione) section (Fig. 11.2) Subsequent nannofossil biostratigraphy at Gubbio (Monechi and Thierstein, 1985) has provided a correlation of thesc polarity chrons to nannofossil zones (Fig. 11.3). The correlations of polarity chrons to nannofossil and foraminifera1 binzonations at Gubbio have been ratified and refined in the Southern Alps (Channel1 and Mcdizza, 1981). Spain (Mary et al., 19Y1),and in cores from the NW Pacific (Moncchi er al., 1985) and South Atlantic (Hamilton et al., 1983: Chaw, 1984;Tauxe et ah, 1983c; Poore et aL, 1984: &eating and Hrrrero-Bervera, 1984). Due to the scarcity of ammonites in the pclagiclimestone sections, polarity chrons are generally correlated to the ammonite-bearing stage stratotypc sections through the micropaleontology. Direct correlation of the GPTS to European Upper Cretaceous ammonite ?ones which define stagc boundaries has not becn achieved. although the C33rlC33n polarity chron boundary has h w n correlated to ammonite biostratigraphy in Wyoming (Hicks rt al., 1995). C.

Cretaceous Long Normal

From paleomagnetic data available at the time, Helsley and Steiner (1969) postulated the existence of an interval of constant normal polarity lrom Late Aptian to middle Sautonian which could be cnrrclated to long smooth intervals in coeval oceanic magnetic anomaly records (Raff, 1966; Heirtzler et at., 1968). The presence of this long interval of normal polarity was confirmed by Kcating ef ul (1975) from deep-sea cores obtained from the Deep Sea Drilling Project (DSDP). Subsequent studies in Italian pelagic limestones rcsulted i n correlation o l the top of the quiet zonc (base of C34r) to the early Campanian, just above the base nf the Glohurr~~ncann elemtn foraminireral zone (Alvarez er ul., 1977;

limestone

marl ~i~~~~ 11.2sumrndv of ~ (ltalv) C~fteri.owrie r*/ lal. 19XOb).

normal

hiafus

0 reverse ~ rnuSncios~ra~igraphi~ ~ t studies ~a1 Guhhill~and Ci*mon ~

~

188

11 Jurassic and Cretaceouq Magnetic Stratigraphy

Figure 11.3 Correlallon of Lete Cretaceous-Eocene calcareous nannofossll and pldnkimlc foram~mfrraleventslzones lo p l a n t ) chrons at Gubbxo. Itdj (after Monech~and Th~rr-

stem 1985)

11 I Cretaceous Magnehc Stratigraphy

189

Lowrie and Alvarez, 1977a,b; Channell et al., 1979; Channell and Medizza. 1981) (Fig. 11.2). The entire Cretaceous long normal interval has been recorded in the Cismon section (northern Italy) and here the revcrse polarity chron at its base (CMO) is close to the BarremianIAptian boundary (Channell er al., 1979) (Fig. 112). The base of CMO has since hccn found to immediately postdate the first occurrence (FO) of nannofossil R. irregularis in two Italian land sections (Capriole and Pie' del Dosso) (Channell and Erba, 1992) and at ODP Site 641 on the Galicia margin (Ogg, 1988). In the absence of the ammonite Drshayesites, which formally defines the BarremianlAptian boundary. Ihc FO of R. irregularis is considercd the most reliable microfossil marker for this stage boundary. The base of CMO is usually close t o but slightly younger than this rannofossil event. There are a number of magnetostratigraphic records of post-CMO Aptian-Albian reverse poliirity chrons. The first such ohsenration in outcrop was by Lowrie et al. (1980a), who observed a reversely magnetized bed in the Late Aptian Globigerinelloids nlgerianus zonc at Valdorbia (Umbria, central Italy). This so-called ISEA reverse polar@ zonc (Fig. 11.4). also known as polarity chron CM-1 (minus I),was not ohsewed at two olhcr sections in IJmbriawhere the same foraminifera1 zone was prescnt in comparable or greater stratigraphic thickness: however, Tarduno et al. (1989) observed two samples with reverse magnetizations just above thc F O of G. a~gerzanusat DSDP Site 463. The documentation of this short reversc polarity chronis strengthened by two samples with reversc magnetization spanning 43 cm in a core from ODP Silc 765 (Ogg er al., 1991b). where the reverse polarity zone postdates CMO and is late Aprian in age. Sevcn reverse polarity zones have bcen observed by Tarduno er a/. (1992) in the middle Albian interval (Fig. 11.4) ( A . albianus nannofossil zone and top Ticinella primula to base Biticinrlla breggicnsis foraminifera1 zones) at the Contessa section near Gubbio (Umbria, central ItaIy). In thcse reverse polarity zones, the magnetization i s carried hy hematite and the reddening of the sediment may be Pale diagenetic in origin. Thc reverse magnetization components in this interval are not antipodal to the normal magnetization components. and the rev-ersc directions are offsel toward directions consistent with the Late Cretaceous or Paleogene. The duration or Lhe proposed middle Alhian polarity zones is well known from lithologic cyclostraligraphy (see Herbert er aL, 1995, and references therein). Of the scven polarily mnes recognized by Tarduno et al. (1992), the thickest (3.25 m) represents about 8Kl ky. This duration is greater than that esfimated for reverse chron CMO and about twice that estimated for CM1 (see Herbert, 1992), both of which are usually recognized by shipboardmagnetic anomaly surveys. It is, therefore, doubtful that thcse middle Albian reverse

190

11 Jura5slc and Cretaceous Magnetic Strdttpaphy

1

1

11.1 Cretaceous Magnetic Stratigraphy

191

d. Berriasian-Aptian Magnetic Stratigraphy

.'

Figure 1 1.4 Correlation of rmd-Crvlaceou~calcsreoua nannlrlossil and planktnnic inraminif. era1 events to polarity chrons (ailzr Ldnun ptnl.. 1493).

polanty zone? represent the geomagnetlc field at the time ot deposition of the sediments. Until the m~ddleAlb~anreverse polar~tychronr are recognlzcd elsewhere, we cons~derthat they should not be incorporated Into the geomagnetlc polarity tlme scale (GPTS)

"M-sequence'' polarity chrons are numbered according to the correlativc oceanic magnetic anomaly. These anomaly numbers, by tradition, corrclate to reverse polarity chrons except for M2 and M4, which corrclate to a normal polarity chrons. We use a prefix "C" to distinguish polarity chrons from magnetic anomalies and follow Harland et al. (1982) in labeling normal polarity chrons (othcr than CM2 and CM4) using the number of the next older polarity chron with the appendage "n." Using this nomenclature, .-CM9 denotes the reverse polarity chron correlative to magnetic anomaly M9 and CM9n denotes the normal polarity chron between CM9 and CM8. It should bc noted the magnetic anomaly bctween M10 and MI1 is labeled MlON. This anomaly was originally omitted from M-sequence anomalies (Larson and Pilman, 1972) and was included by Larson and Hildc (1975) and designated MlON alter Fred Naugler, co-chief scientist o[ the NOAA Western Pacific Geolraverse Project. Land section magnetic stratigraphy in the Mediterranean region has been the basis for the correlation of thc M-sequencc polarity chrons to binonations and hence to gcologic slage boundaries. The oceanic magnetic anomaly record from the Hawaiian lineations (Larson and Hilde, 1975) remains thc template for the M-sequence polarity pattern. Some pelagic limestone sections in Italy have recorded substantial portions of the Msequence pattern. Notable among these sections ai'e Gorgo a Cerbara (CMO-CM9) (Lowrie and Alvarez, 19841, Polaveno (CM3-CM16) (Fig. 11.5) (Channell and Erbal 1992; Channell et nl., 1995b), Capriolo (CM8CM16) (Channell et ul., 19871, and Bosso (CM14-CM19) (Lowrie and Channell, 1984) (see Fig. 11,1,Table 11.1).Tt is remarkable that the polarity pattern dcrived by Larson and Hilde (1975) from Hawaiian oceanic magnetic anomalies has been confirmed lime and time again by these land section studies, particularly in the CMO-CM19 interval. The only polarity chron observed in land section which is not included in the Larson and Hildc (1975) polarity pattern is a second reverse subchron betwcen CMll and CMI2. These two suhchrons werc recognized at Capriolo (Italy) (Channell et al., 1987) and have also now been obscrved in the oceanic magnetic anomaly record (Tamaki and Larson, 1988). The vast majority of the magnetostratigraphic studies in the CMOCM19 inrema1 have heen carried out in the Maiolica L i c s t o n e Formation of Italy. This is a Tithonian lo Aptian whitelgray thin bedded magnetitebearing pelagic limestone with fairly constant sedimentation ratcs (typically -15 mIMy), thereby aiding the recognition of polarity zone patterns. The biostratigraphic control in the Maiolica is mainly from calpionellids for the

I1 1 Cretaceous Magnettc Stratigraphy

VGP Latitude

r)

Figure 11.5 Virtual geomagnetic polar (VGP) latitudes from the Polaveno section ( I P ~ I ~ ) . corrclarionto Hle GPTS-and cnrrelalion ro nannofmsil and calpioucllid evcllts (after C h n e l l and F.rha. IYUZ; Channell ct u l , 1995b). In Lhe litholrrgic log, F indicates small faults (with minimal (~frsct); ~pccklcsindicatc chcrty intervals.diaponals i n a i a t r marlr interval, and hlxk bars indicate black shalcs. Horizontal thin hars indirate pelagic l&estonr. the dominant lithology.

193

Tithonian and Berriasian and from calcareous nannofossils for Tithonian tq Aptian. It has been demonstrated in numerous Maiolica scctions that polarity chrons correlate consistently to nannofossil events (Fig. 11.6) and calpionellid events (Fig. 11.7). For correlations to calpionellids. see Channell and Grandesso (1987) and Ogg et ul. (1991~);for correlations to nannofossils see Bralower (1987). Channell er ul. (1987, 1993). Bralower er al. (1989). Ogg e t a l . (1991b). and Channell and Erba (1992). Thc result of these studies is that the combination of nannofossil or calpionellid biostratigraphy and magnetic stratigraphy gives a very precise and useful integrated --stratigraphic tool for pelagic limestones of this age. Even for short scctions where polarity zone patterns are not distinctive, the nannofossilicalpionellid events indicate the approximate stratigraphic position relative to the GPTS, and the magnetic stratigraphy then gives precise stratigraphic control. Although the cmclation of polarity chrons to nannofossils and calpionellids is firmly established, the correlation to stage boundaries requires correlation to ammonite zones. Although ammonites arc very rare in the Maiolica limestoncs, recent finds place the HauterivianiBarremian boundary in the upper (younger) part of CM4 (Cecca et U L , 1994) and the Valanginiant Hauterivian boundary at the young end of CMll (Chamell crt al., 199%) (Fig. 11.6). Previous estimates of the correlation of these stage boundaries to polarity chrons relied on the correlation of nannofossil events to ammonite zones from Thierstein (1973, 1976). For the Berriasian-Valanginian boundary, the relevant ammonite zonal boundary has been correlated directly to CMISn at Cehegin (Spain) (Ogg et uL, 1988). For the Jurassic1 Cretaceous (TithonianiBerriasian) boundary, there are two alternative definitions with respect to ammonite zones, with no clear consensus. In the Tethyan realm, the boundary lies either at the B. grandis/E. jacuhi zonal boundary or at the underlying B, jacohi/Durangires zonal boundary. At Carcabuey (Spain), the B, juct~bi/Durungireszoi~alboundary lies in a normal polarity chron at the top of the section, which is interpreted as CM19n (Ogg ct al., 1984). Elsewherc, nannofossil and calpionellid events in the vicinity of the Jurassic/Cretaccous boundary have been consistently correlated to polarity chrons in land sections and deep-sea corcs (Lowrie and Channell. 1984; Cirilli et al., 1984: Channell and Grandesso, 1987; Bralower ef ul., 1989; Ogg er a/., 1991~).Thesc studies imply that the Jurassic/Cretaeeous boundary lies in the upper part of CM19n or at the CM19nlCM18 polarity chron boundary, and it has been advocated that this polarity chron boundary bc used to define the stage boundary (Ogg and Lowrie, 1986). At Berrias (France), thc Berriasian stratotypc section yiclded a magnetic stratigraphy that is correlated to calpionellids, nannofossils, and ammonite zones (Galbrun, 1985: Bralower rt a/., 1989): however, the base of the section is within the B. grandis zone and

194

11 Jurassic and Cretaceous Magnctic Stratigraphy

11.1 Cretaceous Magnetic braligraphy

rigvra 11.7 ~

~ ,of calpir,nz~~id ~ ~ ~ l~ ~to ~npolanty t ~r chronv i ~(aRer ~Lhannrll and Grindcrro. 1987: Ogg e6 aL, 1991~).

Figure 11.6 Summary of correlation of Oxfordian to Aptian polari~ychrons to nannafossil ~ " c n t ~ / ~ o nand e s ammonitc zones. Open b a n indicate the range of the nannofosail events with respect to polarity chrons (from Channcll ef 1w5a).

therefore the sedion does not include either of the two definitions o l the JurassicJCretaccous boundary. Ogg et 01. (1991~) have identitied ~ ~ in 1the 7 Purbeck Limestone of southcrn Britain, providing a iirst ~ . ~ magnetostratigraphic e ~ ~ ~ correlation ~ c out~ of the Tethyan into the Boreal Realm.

196

11 Jurashc and Cretaccou Magnetic Strat~graphy

JURASSIC

; i 1 1.2 Jurassic Magnetic Stratigraphy a. Kimmeridgian-Tithonian Magnetic Stratigraphy

As for the Early Cretaceous (CMO-CMl8), the Larson and H i d e (1975) oceanic magnetic anomaly block model is the template for gcomagneric polarity in the Tithonian and Kimmeridgian stages of the Jurassic (CM18CM25). Subsequent extension of the oceanic anomaly block model to M29 (Cande etaL, 1978),and toM38 (Handschumacher eral., 1988) has extended the record into the middle Jurassic interval. Land section magnetic slratigraphies have been correlated to CM18-CM2.5; however, correlations to older M-sequence chrons and between-section pre-Kimmeridgian correlations have not been adequately achieved (Fig. 11.8, Table 11.2). The problem for pre-Kimmeridgian magnetic stratigraphy is twofold. Firsl, preKimmcridgian calcareous nannofossil biostratigraphy does not yet allow precise correlation among Tcthyan sections; and second. pre-Kimmeridgian Jurassic facies in the Tethyan realm are often highly condensed andlor highly siliceous. The oldest polarity chron recorded by the MaiolicaLimestones of Italy is CM19. The Maiolica limestones are usually underlain either by siliceous limestones (Caleari Diasprigni) or by condensed nodular limestones (Ammonitico Rosso). The Sicrra Gorda and Carcabuey sections of southern Spain arc the key secttons for the correlation of CM19-CM25 to ammonite zones (Ogg er al., 1984). In these sections, the "Ammonitico Rosso"-type sediments have mean sedimentation rates of 2-3 m N y . The polarity patterns are distorted by variable sedimentation rates and it is not easy to correlate polarity zones to polarity chrons. Nonetheless, the correlations of Ogg et a/. (1984) placc the ?f. heckeri/HHhybonatum ammonite zonal boundary which defines the KimmeridgianlTilhonian boundary in CM23n. and the I. plunula/S, plarynota zonal boundary which defines the Oxfordiani Kimmeridgian boundary in CM25 (Fig. 11.6). Ammonites are lacking in this interval in the BeUuno Basin and Trcnto Plateau (Southern Alps, Ilaly), where the OxfordianiKimmeridgjan and Kimmeridgian~Tithonian boundaries lie in the unsubdivided "Saccocomu Zone." The lack of calpionellid or other microfossil events in the Italian sections precludes the accurate definition of these boundaries; however, the polarity zone pattern in these sections is more readable than in the Spanish sections. Estimates of

Figure 11.8 Summary of somc of the more mportant Jurd\\lc magnetostrat~praphtcstudlea For kcy, scc Table 11 2

F 200

11 Jurasac and Cretacroua Magnetic Stratigraphy

the location of the KimmeridgiantTithonian boundary in the Italian sections place it in CM22 (Ogg eta!., 19X4: Channell and Grandesso. 1987). Early work on Jurassic magnetic stratigraphy was carried out on the Kimmeridgian-Oxfordian Morrison Formation in Colorado (Steiner and Helsley, 1975a) and the Callovian Summerville and Curtis formations i n Utah (Steincr. 1978). Correlation Lo European sections is hindered bv poor biostratigrapby in thcsc terrestrial and near-shore sediments, complex magnetic behavior which compromises the iidclity of the magnetostratigraphic records, and variable scdimentation rates which distort the polarity zone pattern. Stcincr rtal. (1994) have made the case that magnetic stratigraphy can be used as a means of correlation in the Morrison Formation in Colorado and New Mexico.

b. Oxfordian-Callovian Magnetic Stratigraphy The correlations of European land section magnetic stratigraphies to MOM25 oceanic magnetic anomalies are fairly robust; however, magnctostratigraphic correlation to M26-M38 has not been well established. CM26 to CM30 have bccn correlated to Oxfordian ammonite zones in northern Spain (Steiner et aL, 198511986; Ju6rcz el at, 1994, 1995); however, the correlation betwccn land sections and oceanic anomaly records rcmains somewhat tenuous. The difficulty in corrclation is partly a result of discontinuous and low mcap ~edimentationrates (-1-3 m/My) in the condcnsed limestone facies. Similarly condensed Callovim-Oxfordian sediments in Monti Lessini (northern Italy) yield polarity reversals and sporadic ammonite control (Channel1 et al., 1990a) but the polarity pattern cannot be correlated lo the oceanic magnetic anomaly record. The existence of polarity reversals in the CaUovian-Oxfordian is confirmed by studies of short sections from the Krakow Uplands (Poland) (Ogg er al., 1YYla); however, herc again the lack of long, continuously deposited sedimentary sections does not allow a convincing correlation to the oceanic anomaly rccord. Although the magnetostratigraphic corrclation to the M26-M38 oceanic magnetic anomaly record has not been achieved,it is clear that thc Jurassic "quiet zone" in the central Atlantic magnetic anomaly record is not due to a prolonged interval of normal polarity, as bad been implied by magnetostratigraphic study of the Valdorbia section (central Italy) (Channell era/., 1984). It is now clear that the Callovian-Oxfordian interval. which ~~~- ~~~, ...... cnrre....latesto the Jurassic quiet zone, was an interval of frcquent polarity reversal (Fig. 118 , Table 11.2).

11.2 Jurassic Magnetic Stratiraphy

20 1

averagc reversal rate of at least 5.5 rcversals1My for thc Bajocian. The mean sedimentation rate in these sections is in the 1-4 m1My range. The highfrcquency of reversal and the lack of any discernahle "fingerprint" in the polarity zone patterns have precluded clcar correlation among land The elucidation of the magnetic stratigraphy in this interval will be aided by studies of more expanded sections. Such sections occur in central Italy but they lack ammonites and thcreforc await refinement of .- nannofossil and radiolatian biostratigraphies in this interval. Onc of thc most important magnetostratigraphic studies in the Jurassic isthat of Horner and Heller (1983) from the Breggia section (southcrnSwitzerland) (Fig. 11.9). The sedimentation rates in the nodular limestones at Breggiaare about 14 mlMy for thePlienvbachian and4-7 miMy oftheToar; cian and Aalenian. These sedimentation rates are several times greater than for more typical Ammonitico Rosso-type limestones, and this results in a considerable improvement in the clarity of the magnetostratigraphic record. In addition, the ammonite biostratigraphy for the Plienshachian-early Bajocian interval at Breggia has heen determined in detail by Wiedenmayer (1980). Marlon er ul. (1980) acquired a magnetic stratigraphy Iiom the

c. Pre-Callovian Jurassic Magnetic Stratigraphy For the Bajoc~anand Bathoman, %cvcralsections from southern Spain lndlcate a very Ingh frequency of polarity rcvcr5al (Stcmer eta!., 1987). an

Figure 11.9 Carixian (Pliensbachian)lo Bajocian magnetic stratigraphy and ammonitc zones at B r e g i a (Switzrrladd) (after Horner and Heller. 1983).

202

11 Jurasstc and Cretaceous Magncnc Stratigraphy

9-m-thick Bakonycsernye section (Hungary). Here sedimentation rates in the Pliensbachian are about 1.5 mlMy and 10 reversals are recognized in this stage. Pliensbachian sectionsin Italywithsimilar sedimentation rates yielded comparable numbers of reversals (Channell et a!., 1984). Howcvcr, at Breggia, the sedimentation rate for the Pliensbachian is an order of magnitude greater (- 14mIMy) and 39 reversals are recognized in the stage. Correlation between Breggia andBakonycsernye shows that seven normal polarity zones at Breggia have been concatenated into m e at Bakonycsernye. There is clearly a problem in resolving complete magnetic stratigraphies in the Ammonitico Rosso-type facies when sedimentation rates are a lew meters/~My~ probablybecause of frequent lacunae even in sections where ammonite finds imply continuity of sedimentation. TheToarcianstratotypesections at Thouars and Airvault (France) are gray marls and limestones, and their magnetic stratigraphies have been resolved by Galbrun et ul. (1988a). As the sections are about5 m thick and the entire stage is represented, the meansedimentation rates are less than 1 m1My. Nonetheless, a reasonable magnetostratigraphic correlation can be made h u m the stratotype sections to the Breggia section. The magnetic stratigraphy provides a means of correlation of the West European ammonite zonationin the stratotype sections to the Tethyan ammonite zonation at Breggia. For the Hetrangian and Sinemurian, three red nodular limestone scctions in Austria (at Kendelbach Graben and Adnct) have been studied by Steiner and Ogg (1988). Correlations among sections are hampered by the condensed nature of the sedimentation and the very poor biostratigraphic control in these sediments. The same problems affect graylwhite pelagic limestones of this age in central Italy (Channell et al., 1984). The best quality HettangianiSinemnrian magnetic stratigraphy is from a core drilled in the Paris Basin (Yang ef a/., 1996) indicating high reversal frequency in thc vicinity of the Hcttangian-Sincmurian stage boundary.

11.3 Correlation of Late Jurassic-Cretaceous Stage Boundaries to the GmS Late Jurassic and Cretaceous chronostratlgraphy 1s based on stage stratutypes defined by ammonite zones. However, duc to absence or uneven distribution of ammonite faunas m land sectlons and oceanlc cores, Cretaceous chronostrahgraphy IS often based on calcareous microplankton biostratigraphy and the supposed correlation of these events t o ammonites. A review of calcareous nannofossil events in scctions wlth ammonites, microplankton, or magnehcstratigraphy hasrevealedtheuncertaint~eslncorrelation of nannotossil and calpionclhd datum planes to stage boundaries (Figs.

11 3 Correlatlr,n of Late Iwass,c-~ret~ceous Stage Boundancs tu the GPTS

203

11.6 and 11.7). AU the nannofloral datums have been directly correlated to polarity chrons. Ammonite biozoncs have bccn directly correlated tomagnetic stratigraphy in parts of the Oxfordian-lowermost Valanginian (see Ogg era!., 1991~).Valanginian-Hauterivian (Cbannell ef al., 1995b). and uppcrmost Hauterivian-Barremian intervals (Cecca rt a!., 1994). The OxfordianIKimmeridgian boundary is correlative to the base of the Srctneriaplarynota ammonite zone, which has been correlated t o CM25 insouthern Spain (Ogg et al., 1984). The KimmeridgianKithonian boundary is correlative to the base of the Hybonoticeras hybonomm ammonite zone. -also in southcrn Spain (Ogg et nl., 1984). The TithonianiBerriasian (JurassiciCretaceous) boundary does not have a universally accepted definition. Many of the candidate biomarkers have been correlated to the polarity chrons (Ogg and Lowrie, 1986; Channell and Grandesso, 1987; Bralower et a!., 1989; Ogg et al., 1991~).The ba.e of CM18 has become thc generally accepted correlation to the Tilhoniani Berriasian boundary. The BerriasianlValanginian boundary is defined by thc base of the T. otopeda ammonite zone and falls within CM15n (Ogg et a/., 1988) and between the first occurrence (FO) of Cretarhahdus angusrLfi)ratrrs and the F O of Calcicalathina oblongata (Fig. 11.6). The ValanginianiHauterivian boundary coincides with the base of the A. radiatris ammonile zone and is close lo the F O of Nannoconlrs bfrcheriand at the base of C M l l n (Channel1 et al., 1995b). The base of the S.hugii ammonite zonc defines the Hauterivian/Barrcmian boundary, which falls between the last occurrence (LO) of Lihtraphidites bollii and the LO of Calcicalathina oblongata and in the upper part of CM4 (Cccca et a/., 1994; ChanneU et al., 1995b). The BarremianlAptian boundary was fomally defined at the first occurrence of Deshayesites. None of the nannofossil events proposed by Thierstein (1973) to define this boundary have proved to hc reliable. The F O of Rucinolithus irregularis is correlated to the uppcr part of the C. sarasini ammonite zone and is therefore slightly older than the Barremian1 Aptian boundary (Channel1 and Erba. 1992). The base of CMO coincides closely to this boundary, but direct correlation of this polarity chron with the ammonite biozone has not been documented. The Santonian-Campanian is fomally defined by the appearance of the ammonite Placenriceras hidorsaturn, the index specics of the oldest Campanian ammonite zone. Unlortunately, this species is exlremely rare even in the type area (northwest Europe) and the Santonian-Campanian boundary is often informally defined on the basis of the FO of the nannofossil Rroin~onia narca andlor the FOs of foraminifera Clobotruncana arca . and Boliv~nozdesstrigrllatus. The microfossil markers of the SantonianCampanian boundary result in a correlation of this boundary to the basal

204

I I J~rasslcand Cretaceou Magnetlc Stratigraphy

part of C33r, close to the top of the Cretaceous normal snperchron (e.g., Alvarez et a t , 1977; Channell et al., 1979; Moncchl and Thlersteln, 1985). The type area of the Campanian-hlaastnct~an boundary (northwest Europe) IS characleri~edby a malor hiatus, and hence there 1s more than average debate over the definition of thls stage boundary. In the pelagic realm, the LO of the loraminifera G~oborruncanztucul~urutais often used as an informal definition, and thls event usually appears in the top part of C33r (e.g.. Alvarez et ab, 1977).

Triassic and Paleozoic Magnetic Stratigraphy

12.1 Introduction The absence of preserved marine magnetic anomalies for times preceding the breakup of Pangea forces a change of strategy in the study of pre-Late Jurassic magnetic polarity stratigraphy. Irving and Pullaiah (1976) suggested the use of type sections to build the pre-Jurassic GPTS in thc absencc of the oceanic magnetic anomaly template. Up t o now, type sections for magnetic stratigraphy have not been established; however, wc believe that the typc section approach would be helpful for prc-Late Jurassic time. Type sections ior magnetic polarity sequences will have the same problems presently associated with type scctions in classical stratigraphy. such as unconformities, poor agc control, lack of diagnostic fossils. and inadequate descriptions. The ideal magnetostratigraphic typc section would have not only a well-defined magnetic stratigraphy but also high-resolution biostratigraphy facilitating correlation to geologic stages. Many limestones with good biostratigraphic control have very weak magnetization intensities or are remagnctized. Rcd siltstones and sandstones, on the other hand, often carry a well-defined primary magnetization but are often dilficult to correlate to the standard geologic stages due to lack of diagnostic fossils. In the Soviet Union, a hierarchical system of hyperzoncs (megachrons). superzones (supcrchrons), and zones (chrons) was established for Cambrian to Permian time (Khramov and Rodionov, 1981). Unfortunately, the original data on which this scheme was based are noL availabTc to Western researchers. and descriptionsof type localities were not given. The proposed GPTS (Fig. 12.1) is, therefore, diflicult to evaluate. In inlcrvals for which

more recent data are available, such as the middle Carboniferous, many more polarity chrons are known to exist.

Triassic magnetic stratigraphy was pioneered by Picard (19fi4) and Burek (1964, 1970) and expanded in clastic sediments from thc western United States by C. Helsley and his students in the early 1970s (e.g., Helsley, 1969; known from the lacustrine sediments of the Newark Basin. The initial magnetostratigraphic results from the Newark Basin were from industry boreholcs and outcrops (Mclntosh el a/., 1985). The upper pan of the sequence above the Watchung Lava flows, as well as the flows themselves. are considered Early Jurassic (Hettangian) in age on the basis of pollen stratigraphy. These units are normally magnetized and were labeled the Upper Normal Interval (McIutosh el al., 1985). This normal polarity zone extends down into the Upper Triassic (Rhaelian). h w e r in the section the sediments were found to be characterized by mixed polarity over a stratigraphic thickness exceeding 2500m. Later studies by Witte and Kent (1989) and Wine er a/. (1991). based on more complete demagnetization procedures and dcnser sampling, along with field tests for the age of the magnetization, led to a more complete documentation of the polarity sequence for sediments from the Newark Basin. From the Late Triassic

FQW12.1 Geomagnetic p 0 t ~ ~ i ttime )-

S C ~ I Cde.

rived from IhePaleo~oic~~Cthr USSR. ivor. ma1 polarit? (black). reverse polarity (white) (after Khramur and Rodionov. 1981).

The excellent results obtained from the Newark Basin outcrops led to a study of drill cores. A series of seven drill sites were located to obtain a complcte record of the sedimentary history of the basin. Core recovery was almost 100%and overlapping cores, each about a kilometer in length, resulted in a complete magnetostratigraphic record of lacustrine beds covering the time interval from the early Carnian to the TriassicIJurassic boundary. The magnetostratigraphic study of these sediments is enhaced by the presence of a well-defined cyclostratigraphy tuned to Milankovitch frequencies. The stratigraphy was divided into lithologic units, called McLaughlin cycles, which are believed to represent the 413-ky eccentricity cycle of the Earth's orbit. An age of 201 Ma based on 4"Ar13gArand UIPb techniques has been obtained for the lowest flow unit of Watchung basalts at the top of the sedimentary sequence (Sutter. 1988; Dunning and Hodych, 1990). It is, therefore, possible to date events within the drill cores by counting down from the Watchung basalts using the cyclostratigfaphy.

208

12 Trixssic and Paleozoic Magnetic Stratigraphy

The magnetic stratigraphy of these cores has been studied by Kcnt et al. (1995). Samples were taken from the drill cores at 2-m intervals. All samples were analyzed by progressive thermal demagnetization and oriented using the direction of the magnetic overprint. The magnclic stratigraphy from the Weeton core (Fig. 12.3) illustrates the high quality of the data. The sedimentary rocks in the Newark Basin dip to the north and the drill sites were locared so that the stratigaphies from each core overlap. This has resulted in ovcrlapping patlerns of polarity zones (Fig. 12.4). Thc borehole studies confirmed the previously determined outcrop studies, but the increased resolution led to an increase in the number of polarity zones from 12 1042, anincrease of almost afactor of 4. The Milankovitch cyclostratigraphy places the sediment sequence in a precise time frame and yields a reversal frequency of 2 My-' over late Carnian and Norian time (Fig. 12.2. column 19). Witte et al. (1991) have suggested that thc Newark Basin qualifies as the type locality for magnetic polarity stratigraphy of the late Carnian and Norian interval. The lack of a marine biostratigraphy is offset by the essentially wmplete magnetic stratigraphy and the prccise cyclostratigraphic timeframe. Middle and Late Triassic magnetic stratigraphics have been obtained from condensed marine "Hallstatt-type" reddish nodular limestones at two localities in southwest Turkey and from Austria (Gallet et al., 1992, 1993, 1994) (Fig. 12.2, columns 6, 9# 10, 17). The sedimentary sections are not thick (-30 m), and u?con~ormitiesare present. The qualityof the magnctostratigraphic data is excellent and field and laboratory tests for stability have been carried out. There are, however, discrepancies in correlation between the two Turkish sections which the authors attribute t o the sections originating in different hemispheres. The southernmost section has been comclated to a third section in Austria, the polarity of which is thought to be unambiguous as it must have originated in the northern hemisphere. Thc polarity sequence isnot easily correlated to that of the Newark Supergroup. Kcnt et a/. (1995) have offered a partial correlation of the Turkish sections to the NewarkBasin record. Thcy correlale theKavur Tepe magnetozones a+, e+, and g2+ (Fig. 12.2, column 9) to magnetozones d+, h i . and j+ (Fig. 12.2 columns 16, 19). They have also suggested a correlation of the Chinle section (Reeve and Helsley, 1972) to the Newark sequence. The Chinle section has four normal magnetozones (Fig. 12.2, column 1): the youngest of which (N4) is correlated to I+ of the Newark sequence (Fig. 12, column 16). N2 and N3 of the Chinle section could be correlated to j+ of the Newark sequence. The CarnianiNorian boundary based on pollen in the Newark sequence occurs between the upper part of polarily zone b+ and the base of polarity zone d + . This boundary, based on conodonts, is present in the sequcnce a1 1

,

Figure 12.2 Magnetic stratigraphy of Triasic sections. Numbers raper to individual studies in Tablc 12.1. Right-hand column is composite s ~ a i o n .

Bolucektasi Tepe just pnor to a boundary from reverse t o normal polarity, informally designated N11 (Fig. 12 2, column 6); prior to N8 in this column the quallty of the magnetic strat~graphydeteriorates and lhc VGPs fall to low latitudes. Kent et al. (1995) tentahvely correlated the base of polanty lone N11 lo the base of b+ In the Newark magnetic stratigraphy: however. a secure pattern match 1s not poss~ble.Unfortunatcly, the base of the

Table 12.1 Triassic R ~ ~ k u i l L U R ~ S H ~ Redun 1 chmli

2 chillli

Noel*

I0

17 18

IY Zn 21

2

5 - 9

1

147

1

30

R-

F

3

3

23

1

1 ~ 0 T

ZT

Nvraall-

Turkq!. 2

3m RW

1

1,s

-

-

B

563 I79

1

m

1

L T - I

21

A

RT

U-'I

USA

5

Tvrksy ~~t~~~

I 1

U 5 A Nhl

1

87

1

1

2W

I

110

1

5

1

5

I

179

-

30

-

1 T ILT~I TA

T

-

B

F-0-I

Z ~ P Y-Y

15 8

0

T

Z~PF.U-I Z-P P L I ~ I

1

'1

&P

I

1'

-

7-P F\' F~DL

3'

-

I

T

Z P Y-\'

4

-

F-V

-

0

6

-

EC.1

-

-

R+

+

57

R-

-

55

K t

FT

-

26

R+

n

n

02

n+

F+

PS

12

51

R+

F+ -

\Z

RAIL]

-

1 Mullol"ddilrW!

4

125

1

Russrr

+33l

+?2.R?

2

15"

I

I3 1hW)

USA

-4075

-75.21

3

4

7

IIW X

iurkey

3

+W

1

262

1

I2

-

-

T

Z ~ PY Y

3

cmlndu

I

3

342

I

IIM)

-

-

T

LP

F.DI

US&

iiO.?i

-753

1

1695

I

4RM K

1

T

Z ~ P F.l

G r m

+317

-23.43

1

36

1

Shim

121.9

1106.3

4

501

1

-

7

21

-

I

T

L-P

FV

8

9hO

-

-

T

L-P FDI

A?

%P

2

356

1

MO

-

-

4

2iyl

1

PtiO

-

1~0 I-A

Y M m m 5 el d. (1595) 3 Eurerlrh aod S l a w r i , ~(1QnSl Y Witm n a/ (1791) GmIIU r i d l1'19tl

oa

and sterner (lYY1)

IU

KrnLr~al(1995)

10

i?

65

R+

F+

I nnncrriol 11589)

1

16

.-

fl3

-

-

6 o ~ D m en1 "

7-P F

10

-

17

rornian Spin

-

R-

-

Tadrrrnian

-

51

IIU

.IlShl

16

6

5

16.95

-

19

75

Ausrrhs

R+

F

J

M a h

60

-

-1U6.5

Fm. Wurohrl ~~aiczu lbcrla

F+

-

M

135

X U ~ -95

llr

50

-

Aninian

I lets ~ p c r ~ ~ n ram,an ~ ~ p Rolussk!8rr Narlrn Tepc AXII~~ s~t.4" Archipcl?p Nrwafk Hct: Svpcr@nup Nmirn llldcs An;.qiro. ladinla" ~ e i ~ ~ a n g u r ns p r x ~ C r i r l

80

h hrwkc7aal (19m) 7 Sbivc rrol (1984) x ~ s l l cnt o 1 ( 1 ~ 3 ) 8 GaUst ri *I. (19%) 6 S t r l l l ~ alld r IUBS (IW2) 7 Hcllcrno1.<1988)

5

-

NEW&

F+ C+

T D ~ I

-

T.rerve

~4 R I

R

'I

12.2

Anislam

8 Mdin+G~~rar!d

K

1

Anlnm

4

51

T

136

carnmr

(lM11 3 Kccve ~l oi (1974)

K+

3

1.0

L

E Tnasrc Nnrizn

r+

49

5

a

23.41

Mayerhnn Anton C h l m

Rrevc and Heirley (In21

5

F

T

T

7 Mol-Gnnraol.

9

Z-P

ro

171.13

LISA. A"

5

C+

F

.

~ast timenlmd C!SA.NM

Ani SOB

-

9bR

DM A D

c~~~~~~

Rrlrrrl~ccs

NCh

RM

55

0

NMZ

11

3

Chug*abr 11% KAVU Tcpc

I ? ChiI S Rurris

M

-llH.5

Moenknpi

8 9 10 I1

NSA

+35

4

Moenkopi

NSI

USA. N M

Clmiam

ticmjng ~ , n d ~uc.ian

7

i33.m -1m.08

NSF

24U 10

3

6 KdSceklasj T=F

IISA.NM

A

-

nm

1 Turner era1 (l9W)

212

I 2 'lnasac and Paleurmc Magnrllc Strat~graphy

Straf~graphy Chrono-

VGP Latltude

Lltho-90"

0"

Polarlty 90"

Figure 12.3 Magnet~cscratlgraphy of the Wesron core Newark Basln (after Kent rr a l . 1995) Polant) chron~arc labclcd as normal and rcvenc pars, wlth thc rcvcnc ovcrl>rng normal.

Carnian is not identified in the Newark basm. In the Mayerhng sectlon (Austria) (Fig. 12.2, column lo), the LadinianICamian boundary occurs in thc normal polarity zone A+. At other sections, t h ~ sboundary occurs at unconformitic~(Gallct et uL. 1994). Middle Tnassic magnetlc strat~graphyi q less well established than that of thc Late Tnass~c.In the M~ddleTriassic Moenkopi Fm. of Arizona and New Mexico, the biostratigraphy must rely largely on rare vertebrate finds. The magnetic stratigraphy of the Moenkop~Formation is well known

214

12 Triassic and

Paleozoic Magnetic Stratigraphy

through the studies of Helsley (1969). Helsley and Steiner (1974). Purucker et al. (1980), and Molina-Garza et ul. (1991). Steiner er al. (1993) place the upper part of this formation in Arizona and Colorado in the early Anisian. Molina-Garza et al. (1991), on thc other hand, place a section of the Moenkopi from New Mexico in the late AnisianiLadinian, based on fossil vertebrates (Fig. 12.2, column 4). Steiner and Lucas (1992) have obtained a magnetic stratigraphy from the Anton Chicu member of the Moenkopi Fm. which they place in the Anisian on the basis of vertebrate fossils (Fig. 12.2, column 11). The polarity zone patternsfrom the Anton Chico member given by the two studies lrom New Mexico are very similar and are said to correlate to the Holbrook member of the Moenkopi Fm. in Ari~ona. The only studies of marine sections from the Middle Triassic are by Muttoni et al. (1994) and Muttoni and Kent (1994). The first of these studies delineates the magnctic stratigraphy across the AnisianILadinian boundary from a section on Hydra Island (Greece). The quality of the data is good and the stage boundary occurs within a normal polarity zone informally designated C + (Fig. 12.2. column 20). The other study is on limestones from the Southern Alps and the section is in scdiments of the latest Anisian age. The section studicd is entirely normally magnetized and probably correlates with magnetozone C + of the Hydra section. Magnctostratigraphic results have been reported from the Middle Triassic Muschelkalk Fm. of Spain by Turner et al. (1989). The quality of the magnetostratigraphic data from this study l;s poor. A large proportion (85%) of the samples did not yield interpretable demagnetization data, and this magnetic stratigraphy is not included in Figure 12.2. Lower Triassic results have been obtained by Steiner et al. (1989) from the Feixianguan and Jialingiang lormations of Sichuan (China) from a l-kmthick section which ranges in age from the Pcrmo-Triassic boundary to the Spathian (upper part of Lower Triassic) (Fig. 12.2, column 21). In a recent compilation (Steiner et uL.. 1993), stratigraphic gaps are shown'in the Spathian, but how these are recognized is not clear. Ogg and Steiner (1991) have studied the stratotype seclions of the Early TriassicfromEllesmere and Axel Heiherg islands in the Canadian Arctic. Thedata quality of the three sections is not as high as that from the Sichuan; however, the magnetic stratigraphy is rcadily interpretable (Fig. 12.2, column 13). It these two sections cover the entire Early Triassic then the correlation of polarity zones would seem to be straightforward, since nine normal polarity zones are present in each section (Fig. 12.2). More detailed biostratigraphy is needed to solve the problemof possible stratigraphic gaps in the late Early Triassic of Sichuan. Unfortunately the borealfauna of the Arcticcannotbeeasily correlated to the tropical fauna of the Chinese sections, inhibiting unequivocal magnetostratigraphic correlation. The magnetic stratigraphy of the boundary of the EarlyiMiddle

Triassic (SpathianiAn~sian)has been studied by Muttoni etal. (19%)infossiliferous limestones from Chios Island (Greece). The stage boundary, based on ammonites, falls w~thinnormal magnetozone C + (Flg. 12 2, column 14). The author? correlate normalmagnctozone C + at Chioswithnormal polarity "nne SnN2 - r ~ in the Arctic sections and withN8 ~nthe Sichuan section. although a correlation to N9 appears more likely. Mx~n~tnstratipranhic studies across the Permo-Triassic boundary indi" cate that a polarity zone boundary from revcrse to normal is present at, or shortly above, thc Permo-Triassic boundary. Studies in the former Soviet ,Union and China (Steiner et al., 1989: Heller ef al.. 1988) indicate that the basal Triassic normal polarity zone corresponds to most of early Grieshachian time. McElhinny (1973'). in an analysis of polarity bias, concluded on the basis of data available at that time that the geomagnetic field during Triassic time was 75% normal polarity. In Figure 12.2. the right-hand column represents a composite magnetic stratigraphy, and it can be seen that the magnetic ficld is not biased toward normal polarity. The reversal sequence given here for the Triassic is most reliable for the Early Triassic. upper Carnian, and Norian times and least secure in the Middle Triassic. L

12.3 Permian Irving and Parry (1963) werc the k s l to recoFize thc long period of reverse polarity spanning Late Carboniferous and Permian time. They named it the Kiaman Magnetic Interval, aftcr the name of the town in New South Wales (Australia) where rocks showing reverse polarity were first reported by Mercanton (1 926). The name Kiaman has been retained by Russian workers but was abandoned by Irving and Pullaiah (1976) and replaced by the PermoCarboniferous Rrverse Superchron [PCRS). The PCRS is known to end within the Late Permian. Irving and Parry (1963) originally placed the termination of the Kiaman at the. heginning of thc Naritbeen Chocolate Shales ul the Sidney Basin (Australia), which are Late Permian and Early Triassic in age. They called this polarity change the IUawara reversal. A restudy of this formation was carried out by Embleton and McDonnell (1981), who w n firmed a reverse t o normal polarity change in red claystones near the base of thc Triassic. The older sediments in this sequence are unstably magnetized so a complete magnetic stratigraphy could no1 he established. The Upper Permian and Lower Triassic sequcnces in North America. Europe, and in many o l the Gondwanaland continents often contain important unconformities. IJpper Permian marinescdiments are present, however, alongthe northern margin of Gondwanaland (Pakistan and India) as well as

216

12 Triassic and Paleozoic Magnetlc Stratigraphy

in China. Thebest quality magneticstrat~graphyforlaterpermian time comes from the Wargal and Chidru Formations of the Salt Range in northern Palastan (Haag and Heller, 1991) (Fig. 12.5,column 13).These scctions comprise mainly carbonate sediments and yield an excellent fauna allowing correlation to other sections. The magnetostratigraphic record spans latc Kazanian and Tartarian time, with six normal and five reverse polarityzones. The top of the PCRS was not determined as the lowerpart of the Wargal Formation is normally magnetized. Unfortunately,it has become apparent that there is an unwlnformity at the top of the Permian in this section: sediments equivalent to Changxingian time in the Chincse sequence are apparently missing (Wignalland Hallam, 1993).The biostratigraphicandmagnetostratigraphiccorrelation of the Salt Range sequences to China (Fig. 12.5, columns 11 and 12) (Steiner et al., 1989; HeHer et al., 1988) is therefore tenuous, although the correlation of the thick normal polarity zone near the top of the sequences in both China and Pakistan is reasonable (Fis12.5). PERMIAN

Figure 12.5 Magnetic shaligraphy of Permian sections. Numbers refer to individual studies in Tahlr 12.2, Right-hand column is cumpositc section.

12.3 Permian

The correlation of the Upper Permian marine sequence in Pakistan to nonmarine rcd beds which dominate the rock record in Russia, Western Europe, and North America and South America is not clear. A recent study o f t h e Ochoan (Tartarian) Dewey Lake Formation in northern Texas (Molina-Gana eta!,, 1989) shows at least two normal polarity zones in the Late Permian (Fig. 12.5 column 4). Khramov (1974) shows at least five normal polarity zones in sediments from the Upper Permian type section of the Ural Mountains (Fig. 12.5, column 3). The data quality appears to be high, and the results arc internally consistent. Unfortunately, these sections have an unconformity in uppermost Permian and lowermost Triassic, similar to the situation in the western Unitcd States. A considerable amount of magnetostratigraphic work has been carried out on the Permian Rotliegcnde and Zechstein Formations oI Germany, both in outcrop and bore holes tMenning et a/., 1988) (Fig. '12.5, column 9). There are normal polarity zones in the upper Rotliegcnde and throughout the Zechstein Formation. Menning et al. (1988) showed at least seven normal polarity zones in the Late Permian. This number of normal polarity zones is more or less consistent with thc results from the former USSR and Pakistan, where five and six normal polarity zones have hcen recorded in this interval. respectively. In the composite polarity column for the Permian stage (Fig. 12.5, right column). eight normal magnetozones are given for Lale Permian Lime, derived principally from the records liom China, Pakistan, and contra1 Europe. The position of the top of the PCRS is older than the Tartarian or Russia, the Zcchstein and upper RotliegcndeFormations of central Europe, and the base of the Wargal Formation in the Salt Range of Pakistan. The Wargal Formation is Murgabian in age, correlative to the Guadalupian or Wordian of the North America and to the base of the Kazanian stage in Russia (Haag and Heller. 1991; Catalano et a/., 1991). The upper houndary of the PCRS in North America lies beneath Ochoan units of west Texas, which have an associated '"Ar13'Ar agc of 251 2 4 Ma on volcanic ash (Molina-Garzaet ul., 1989). The boundary between the Permian and Triassic in China has been dated at 251 Ma by Claouk-Long et a/. (1991) using the SHRIMP ion probe at the Australia National University. Taken at face value, the date from Texas would place the Ochoan magnetic stratigraphy very close to the Permian-Triassic boundary. Harland et ul. (1990) have subdivided the Permian stage into the Rotliegende and Zechstein epochs in a bipartite division. The base of the Zechstein Epoch is given a numerical age of 256 Ma by Harland el al. (1990). Valencio et ul. (1977) in Argentina have reported a normal polarity zone lying ahove an intrusion dated at 259 2 7 Ma which they called the Q. Del Pimiento event. In the same region, Creer er 01. (1971) also noted normally magnetized lavas at 258 t 5

Ma. It is possible that these normal polarity zones may mark the end of the PCRS and arc not subchrons within the PCRS. Further sampling in the Argentine sections may offer an opportunity to date the end of the PCRS. D u e to the lack of reliable radiometric ages directly associated with sedimen~sof Late Permian age, stage boundarics of the Late Permian ate poorly constrained in time. The precise date of the end of the PCRS is therefore unknown. Haag and HeUer (1991) believed it to be older than 261 Ma, whereas M. Menning (personal communication, 1992) gives an age of 265 Ma. In Figure 12.5, an age of 262 Ma is given for this boundary; however, ii should be understood that nowhere is it directly dated. Over the years, reports of short normal subchrons within the PCRS have been puhlishcd in both Russian and Western literature. Helslcy (1965) found normally magnetized sediments in the Dunkard Formation of West Virginia, which is believed to be Early Pctmsdn in age. The initial study was done with alternating field demagnetization, which is notoriously ineffective For removal of secondaly overprints from hemalite-bearing samples. Samples from the normal polarity horizon were subsequently subjected to thermal demagnetization by Crose and Hclsley (1972) and were round to retain their normal polarity. Normal polarity magnetizations have been obtained from Lower Permian Pictou sediments of Prince Edward Island by Symonds (1990). The normal polarity zone in the Pictou heds may he correlative to that in the Dunkard Fm. of West Virginia. Menning era/. (1988). in their study ol the Rotliegende Formation of eastern Germany, found normally magnetized sediments in the Early Permian part of this sequence which thcy placed a t LhePermianiPennsylvanian boundary, implying a correlation with the Dunkard series of West Virginia. It docs seem certain that one or more normal subchrons occur within lower Permian strata, although it is unclear whether the normal polarity zones recognized at different locations arc correlative. Several sections have been sampled across the Permo-Carboniferous boundary in the western United Stales (Diehl and Shive, 1979,1981; Miller and Opdyke. 1985) and nonormally magnetized rocks were detected. Sinito et ul. (1979) gave evidence for a short subchron within the PCRS, positioned 100 m below a lava dated at 288 -t 5 Ma. This may be equivalent to the normal polarity subchron seen in the Dunkard Fm., although it was originally thought to be the base of the PCRS. Normally magnetized sites of Early Pcrmian age have been reported in several tectonic studies (Halvoryon et ul., 1989: Wynne el al.. 1983: Irving and Monger, 1987) and may be correlative with the Dunkard subchron. Although normal subchfons may occur within the PCRS, magnetostratigraphic studies by Diehl and Shive (1979, 1951), Miller and Opdyke (1985), and Magnus and Opdyke (1991)

I

I

.!

220

12 Tnas5lc and Paleomic Magnetic StrAtlgraphy

have shown that the PCRS is dominantly reberse polanty, as claimed by Irving and Pullaiah (1976) (Fig. 12.5, columns 6-9).

12.4 Carboniferous The age of the base of the PCRS has not been resolved. In North America, McMahon and Strangway (1968) reported normal polarity zones in sediments of the Upper Demoinesian Minturn Formation of Colorado. The section was restudied by Miller and Opdyke (1985), who did not find any normal polarity zones in the section. The study was extended throughout the Minturn Formation by Magnus and Opdyke (1991), who determined that the reverse polarity zone (PCRS) extended to the base of the Minturn Fm., which is earliest Demoinesian or late Atokan in age. Roy and Morris (1983) had shown that sediments with dual polarity are present in rocks as young as Westphalian A in the Maritime provinces of Canada. Inthe same region, DiVenere and Opdyke (1990,1991b) documented an internally coherent set of polarity zones in the Maringouin. Shepody, and Claremont Formations of Namurian and Westphalian A age. At least seven normal polarity zones arc present, giving a reversal rate of about 1My-' (Fig. 12.6). The age of magnetization is constrained to the Late Carboniferous by a positive fold test. An attempt was made to extend the magnetic stratigraphy to younger formations exposed along the shores of the Bay of Fundy. All samples were reversely magnetized; however, a data gap of 700 m occurs in the sequence (Fig. 12.6). These data yield the besl and most complete rcversal sequence now available for the early Late Carbonifcrous. The dual polarity for rocks of early Pennsylvanian age is supported by a study on red Morrowan palenols from Arizona (Nick el al., 1991). The age of the beginning of thc PCRS was originally determined to be 310 Ma based on the WAr age of the Paterson Volcanics in Australia. These volcanics were shown by Irving (1966) to be normally magnetized and were used to define as the base of the PCRS (Kiaman). Sites taken from the Seaham Formation, directly overlying the Paterson Volcanics, wcre reversely magnetized and were thought to be Late Carboniferous in age. Upper Carboniferous sitcs lrom the Currabubula and Lark Hill Formations were also reversely magnetized and passed the fold test. Thc base of the Kiaman was, therefore, thought to be firmly determined. However, a recent study of the Paterson Volcanics using single-crystal zircon dating by Claoui-Long et al. (1995) has yielded a date of 328 Ma, which places them within the Visean. much older than the normally magnetized sequences from the Canadian Maritimes (Fig. 12.6).Thereforc, the Paterson

Canadian

Maritirnes Nova

1

1

I I

1

' '

I

Mauch Chunk Formation Pennsylvania pottsv,lle

Jlm D o T e

Lavelle 111"

io,\is.ere and 0Pd"kP lB91bl

I

I

1

I

--

Figwe 12.6 Magnenc btrrrat~~raphy < the ,I Canadtan Marinmcprt~vlncesand the MauchChunk Fm (Pmnsylvalm)

222

12 Tr~assrcand Pnleo~o~c Magnehc Stlrat!gtaph+

12.5 Re-Carboniferous

Volcanics cannot be used to fix the onset of the PCRS. Reccnt studies in Australia (Opdyke et al., 1960) have located the base of thc Kiaman and dated it at approximately 320 Ma (Claoug-Long, pcrsonal communication). The other significant data set that bears on this problem is the magnetic stratigraphy of the Middle Carboniferous strata of the Donctz Basin in the Ukraine. Kbramov (1974) has presented a magnetic stratigraphy for the Upper Carboniferous which shows normal polarity zones as young as Upper Moscovian. Opdyke rr a/. (1993) restudied these sediments, and although it was possible to recover a prefolding magnetization which is probably Carboniferous in age, it was not possible to codinn Khramov's original magnetic stratigraphy. A t this time, thc only unimpeachable data that bear on the age of the base of the PCRS are those from Maritime Canada. The highest normal polarity zone in this sequence occurs within sediments of Namurian age and yields a duration for the PCRS of approximately 58 My (from 262 to 320 Ma). InNorth America, the reversalpattern dctel'mined in the LowerPennsylvanianhas been extendedinto the UpperMississippian (Visean)insediments of the Mauch Chunk Formation of Pennsylvania (DiVenere and Opdyke. 1991a) (Fig. 12.6). Opdyke and DiVenere (1994) have reported a minimum of nine normal polarity zones within the Mauch ChunkFormation (Fig. 12.7, column 7) with a reversal frequency of 1.5 My-'. similar to that of the early Pennsylvanian. The geomagnetic field appears to have been about 50%normal and 50%reverse in Viddle Carboniferous time (Fig. 12.7). Little is known of the polarity pattern for the rest of Early Carboniferous time; however, both polarities are present in sediments of Visean and Tournaisian age (Irving and Strong, 1985). The beginnings of a magnetic stratigraphy has been obtained from Lower Carboniferous lavas of thc Midland Valley of Scotland of Torsvik er al. (1989) (Fig. 12.7, column 8). Turner et a/. (1979) and Palmer et a/. (1985) have reported magnetic stratigraphy from Carboniferous limestones horn Great Britain; however, these limestones have been remagnetized (McCabe and Channell, 1994) and the magnctic stratigraphy is therefore unreliable.

12.5 Pre-Carboniferous Middle and Early Palcoroic rnagnetlc stratigraphies are largely restricted to the h m e r Sov~etUnion Lower Paleo~oicdala are mainly from the Siberian platform. In North Amenca and Western Europe, severe remagnctizat~onproblems have resulted in very few reliable paleomagnet~cpole position? and a pauclty or magnctortratigraphic data. A very condensed sequence across the FamcnnianiFrasn~anboundary of thebate Devonian appears to record several reversals of the geomagnetic

Figure 12.7 M.d~neticrtratigraphy of Cnrhoniferous sections.Right-himdcolumn isa compositc. Numhcm- refer to individual studirs in Table 12.3.

field (Hurley and Van der Voo, 1990) A reversal sequence of thls age does not appear in the Russian GPTS. Douglas (1988) has studied an -1 k m thick section of the R~ngerikeFm. of Norway wh~ch1s cons~deredto be Ludlow (Late S~lurian)In age. Two formations were studied, the Sundvollen Fm. and Stubdal Fm.3rost of the Sundvollen Fm. appears to be normally magnetized while the uppermost Sundvollen Fm, and overlying Stubdal

Em. are reversely magnetized.The normal polarity zone is well documented and the high unblocking temperature magnetization is most probably Silurian in age. Trhh er al. (1993) have compiled Silurian paleomag~tic_data ., from all continents. Theyidentified 19
~~

.

((66,) PP11"d

X

is

-

~i-

91

(MI)l o i*V*"llnaW ( ~ 6 1 )'"W@IYN

I

-

-

9?

YI

la

-

-

-

-

:P

-

tZ

-

B

i

-

-

ilnal) q z r r n pux ~ X?!ASIYI,

L9~b11 0 3 ~ P9U S ~ Y U U I Y (686~11 ~ 2 P"' ~ la\llw c

II

i.1

-

N

-

4

I

I

L d -

L

-

-

- -

I Oh

t-1 I

h€

I

O i

I

Y

-

UYL I?i

I

1

L

O ~ d d-Z

R 9 d

r

-

0 ~

Y

L

-

-

f 1

I

i"

5

YTZ-

.

.anl %mr.n axsrn~ p z g 'rcq srirng DL0 a l p p i x

i91-

-

'.

saluaJ31Sn

"urn


6U

-

Ox

Li

-

PL

$X<,

U,P.P5

i i ~ ~ i M,BU

SOCi

YY7d

'VSil

,op,r,

~U',,C~SXB."* YID"WA"I~<

-

-

-

0 LJ1 I.X

4JN

ZW

-

V

OV

- I Q U Y18

P

U

W

A

VST

-

-

ISN

,

3SN

v 6

O

"mn

-

-

Y

"08%

i

us,ll!"Suv1

PU,: J O A P ~ ~ Y B O %

I!SUIY

"1.4 'l!"LW

2-"p"W"I

.lvaa q=u;uj,ur pat!= '10 '"OWUIrm

lah!x

"a""',"

(laall .(S%I)

SLH Y U I A ~ ( ~ WJO~WIJ unzaq!s

>ope'".,

? H * v "1

13UU sux1

i

,

l!"'XJnU

UDlXAODlkOJ~IUD3

r

228

12 Tnasslc and

Palroruic Magnet~cStratigraph?

more complex than that proposed by Trench el al. (1991). A new study of Ordovician rocks from Arkansas (Farr et a l , 1993) has yielded 16 polarity zones spanning late Arenig and Llanvirn time (Fig. 12.8, column 7) and a section presented by Khramov for theMiddle Ordovician of the Lena River also shows many (-22) polarity zones. Unfortunately, faunal information is lacking from the Lena River section. What seems to be certain, however. is that the Middle Ordovician is a limc of relatively frequent reversal of the geomagnetic field (Fig. 12.8). A study of a Lower Ordovician section from the Lena River yields at least 16 polarity zones inTremodoc sediments 1 1984) (Fig. 12.9; Fig. 12.8, column 6). Thc base of the (Metallova ~ ' a!., Early Ordovician appears to be rcvcrsely magnetized in this section. It is clear, when the original data are compared to the polarity scale presented by Khramov and Kodionov (1981) (Fig. 12.1), that their scale was greatly simplified. The most important magnetic reversal studies in rocks of Cambrian age arc from the Sibcrian Platform (Khramov, 1974; Kirschvink and Rozanov. 1984). The stratigraphy presented by Khramov spans the time interval from the middle Cambrian into thc Ordovician. Data are now available from Australia from an expanded section at Black Mountain (Qucensland) for Latc Cambrian and Early Ordovician time (Ripperdan and Kitschvink, 1992) (Fig. 12.10). The section in Australia is 1 km thick, while the Sibcrian sections rarely exceed 100 m in thickness. An unambiguous corrcla:ion between these two regions is no1 possible, but there are some common features in the magnetustraligraphic records. Both studies show reverse polarity at the Cambro-Ordovician boundary and both show five normal polarity zones in the Late Cambrian. A more definitive correlalion bctwccn the two regions will require lurther study of the Russian sequences and a more complete presentalion of the biostratigraphy of the Siberian sections. Kirschvink and Rozanov (1984) have analyzed sediments Irom thc lowermost Cambrian of the Siberian Platform, and a magnetic stratigraphy across the TommotianAtdahanian boundary was obtsined. The magnetization camponenls pass a fold test; however, they yield a pole position which is far away from lhe Siberian apparent polar wander path, making polarity designations ambiguous. Kirschvink et al. (199'1) have inverted the inilial polarity dcsignations based on changing ideas of paleogeography. The correlation of this pattern to a similar age record from the Amadeus Basin of Australia (Kirschvink, 1978) is not possible, perhaps because of unconformities in the Australian record. A magnetostratigraphic rccord from this time interval given by Wu et ul. (1989) is now thought by the authors to be the result o l remagneli/.ation.

229

12 6 Polanly Bias ~nthe Phanerorotc

RLVER LENA

SIBERIA

E A R L Y ORDOVICIAN

n"

Figure 12.9 Magnetic stratigraphic study ofEilrly Ordovician redimeclrs from lhr Lena Rivcr. Sihcfiu (after Mctallova erul,, 1984). O and I denote dcclinaLiou and inclinauion. respectivet?. Polarity interprrlillion is reproduced in Fig. 12.8, column 6.

12.6 Polarity Bias in h e Phanerozoic The reversal record of the gcomagnetic field from the Carbonrferous to present 1s known well enough to allow uq to examine the overall reversal patlern for the last 330 My o l Earth history. 11 was previously thought that long segments of thts record were charactenzed by polar~tybias. Except

230

12 Trlasslc and Paleoro~cMagnetic Stratngraphy

t-Middle

Carboniferous

In WT boundary

4 11

40

38 32 28 24

m a l l o n (hlyr)

Figure 12.1 1 A cornpanson oldurauon olpo1antgehrtm.ifor fhc Ccnozo~cand the Carbonifcr ous-Cretaccou5 lnterval

Figure 12.10 Black Mountain section (Australia) across Cambro-Ordovician houndary (after Ripperdan and Kirschvink, IW2).

for the Cretaceous quiet zonc and PCRS (Kiaman), lherc is no evidence of substantial polar~tybias in the Phaneromic. At the present state of knowledge of the polarity structure of thepaleomagnetic fjcld, the durat~on of polarity chronr fromMiddle Carbon~ferousto the end of the Cretaceour i? rirmlar to the distnbut~onof durationsseen in the Cenozoic (Fig. 12.11). In the logarithmic histogram of polarity duration since 330 Ma (Fig. 12.12), polarity chron durations generally Lie in the 0.1-1 My interval. The PCRS (Klaman) and Cretaceous qmet zone (KQZ) la11 well away from the bulk of the data. Polarity chrons with durations greater than 2 My arc very rare. In Figure 12.13, the number of reversals was counted in 4-My bms and

Figure 12.12 Htilogram of duration of polar~tychrons (lag scalc), lor 330 Ma to present

232

Cenozolc

12 'Triassic and Palemoic Magnetic Stratigraphy

Triassic

Permian

Carboniferous

Secular Variation and Brunhes Chron Excursions

Figure 1213 Reversal rate from 330 Ma to the prcsent. The numher nlreveraals wasmunrcd in 4-My hins and smoothed with a running mcan.

smoolhed with a running mean. The number of reversals per million years ranges from zero to about 4 My-' in the late Cenozoic. The average reversal rate: excluding the two superchrons, appears to be about 2 My-'. If the rccord is considered as a whole it would seem that the geomagnetic field has two states, one in which it reverses at a rate of two or more per million years and another in which it does not reverse at all. The change between these two apparent geomagnctic states has been correlated to activity or mantle plumcs which rise rapidly from the D" layer and manifest themselves as large igneous provinces such as Ontong Java Plateau and the Deccan Traps (Courtillot and Besse. 1987; Larson and Olson, 1991). McElhinny (1973) analyzed the polarity rccord from paleomagnetic studies available at that lime and concluded that the ratio of normal to reverse polarity indicated that for long periods of lime the geomagnetic field seemed to favor one polarity. He recognized polarity bias in favor or normal polarity in Mesozoic rocks and in favor of reverse polarity in rhe Paleozoic. Intervals of polarity bias were postulated by Algeo (1996) for the Jurassic and Ordovician based on a global analysis of Phanerozoic paleomagnetic data. The analysis given hcrc shows no such bias in the magnetostratigraphic record. The apparent bias is probably caused by unremoved secondary magnetizations acquired during the Kiaman (PCRS) or during the KQZ and by inadequacies in amount of high quality magnetostratigraphic data, particularly for pre-Late Jura~dictime.

13.1 Introduction "Magnetic stratigraphy" or "magnetostratigraphy" most commonly refers to polarity reversal stratigraphy. However, directional changes associated with secular variation of the geomagnetic field are an important means of correlation in recent sediments. The technique has been largely restricted to lake sediments due to theirhigh sedimentation rates and lack of bioturbation. At typical deep-sea sedimentation rates of about 1 cmlky, secular variations are averaged out by bioturbation andlor the remanence acquisition process. However, marine sediments from restricted basins (such as the Mediterranean and Black Sea) with sedimentation ralcs of sevcral tens of cmky yielded some of the earliest secular variation records (e.g., Opdyke el al., 1972: Creer. 1974). The first convincing demonstration that secular variation of the geomagnetic field can be recorded in lake sediments is attributed to Mackereth (1971), who used this technique to provide a time frame for sedimentation in Lake Windermere (UK). He correlated remanence dcclination changes to observatory records, particularly to the early 19th century westerly declination maximum, and showed that the chronology is consistent with the older "C ages but not with those influenced by detrital carbon input from recent burning of fossil fuels. This work paved the way for the development of the field of secular variation magnetic stratigraphy, which can provide a Lime frame in lake sediments and is also our main source of information on the nature of Holocene geomagnetic secular variation.

234

13 Secular Vanafton and B~unhesChrun Excunlons

235

132 Scd~mentRecords of Secular Variauon

Whereas conventional (polarity reversal) magnetic stratigraphy depends on the rewgnition of duectional changes of remanent magneti~ation of about 180", secular variations generally d o not exceed a few tens of degrees. Hence sampl~ngand measurement tcchniques in secular vanation studies are particularly critical. The development of a pneumatic corlng dcvicc surtablc for lakc sampling (Mackereth, 1958) was an important step toward obta~ninglake sed~mentcorcs that were both long enough and sufficiently undeformed for secular vatiatinn studlcs In his pionecnng study, Mackereth (1971) used an astatlc magnetometer to mcasure subsamplcs sliccd from thc cores. The advent of pass-through fluxgate magnetometers (Molyneux er al., 1972) allowcd the declination of remanence of the whole core to be rapidly measured without suhsampling, and the more recently available 3-ax~spass-through cryogenic magnetomctcrs (Gorec and Fuller, 1976, Wecks rt ul., 1993) allow both declination and inclination to he determ~nedrap~dly.

AVERAGE

TRANSrnRMED DECLINATION

AVERAGE TRANSFORMED INCLINATION

13.2 Sediment Records of Secular Variation Although sccular variation records can provide a means of precise correlation, only about 20% of all lakes studied have ylcldcd secular variation records of bufticient resolution Lo define turnlng poihts in declination and inclination whlch can be used for correlation (Thompson and Oldfield, 1986). Some of the clearest and sharpest record? arc from Western Europe, yuch a7 those from Lake Windermere (Mackereth, 1971: Creer et u l , 1972), Lough Neagh (Thompson, 1973). Lake Vuokonjarvi (Stober and Thompson, 1977), Lac de Joux (Creer et u l , 1980a), and Loch Lomond (Turner and Thompson, 1979). The better dated B r ~ t ~ records sh have been merged to give a well-dated "master curvc" o l directio~alchanges for the last l 0 , m years (Fig. 13.1) (Turner and Thompson. 1981) which can be used as a calibration curvc. There 1s also a large body of data from North Amcrica whlch come7 mainly from thc Great Lakes (Creer et aL, 1976; Mothersill, 1979, 1981; Creer and Tucholka, 1982), from Minnesota (Banerjee et ul., 1479; Lund and Bancjee, 1985; Sprowl and Banerjee, 1989), from Oregon (Verosub et a l , 1986). from California (Lund et al, 1988; Brandsma et al., 1989). and from Bntlsh Columbia (Turner, 1987). The detailed 36-125 ka record from Mono Lake (California) (sco Lund et a/., 1988) rewrds the Mono Lake excursion at about 28 ka and a dist~nctivewaveform in the suhsequent declination and inclination record wh~chappears to recur every 2500-3500 years (Fig. 13.2). It was suggested that this dlstinct~vewaveform evolvcd out of the Mono Lakc excursion and the pers~stenceof the subsequent

Figure 13.1 Rnushmastcr secular varlatloncunr (afler Turner and Thompson, 1981) Transformed drcllnatlon and ~ncl~natlon avcragrd over ten corcs fur 0-7 ka and three cc~resfor 7-10

kd

waveform suggests along-term memory in the core-dynamo process (Lund. 1989).This hypothesisis supporled by the work of Negrini er al. (1994), who recorded a repetitive waveform in the secular variation in lake sediments from Orcgon which immediately postdate the Pringle Falls excursion. Comparison of the British Columbia (Mara Lake) record with the composite record from the Great Lakes indicates little correlalion for the

236

237

13 Secular Vdr~dltonand Brunhe* Chmn Excuruons

5000 and 2000 yr B.P. of 400 years over the 30 egree longitude difference between the two sites, implying a westward drift of the nondipole field ot.0.0785 degrcelyr (Turner, 1987). Such interprctions are highly dependent on the precision of available (radiocarbon) econtrol. The correlationof therecord fromFishLake,Oregon (Verosub al., 1986) with that Irom British Columbia is problematic (Fig. 13.3). with an apparent 800-yr difference in the age of corresponding features at the two sites (Turner, 1987). This may be explained by radiocarbon ages tare systematically 800 yr too old at the Oregon site (due lo the introduction of radiogenically old detritalcarbon) or by a time lag between deposiand stabilization of the postdcpositional remanence a1 this site (Turner,

DECL I NATIONS

I NCLI NATI ONS

Fish Lake Mara Lake

Figure 13.2 Secular variation rccord from Mono Lake srdiments. The Mono Lake excursion occurs in the 6-7 m intcrvsl and ia followed by several proposed repetitions o f the secular variation u~avelc,rm(after Lund el a/.. 1988: Lund. 1489).

Great Lake

Figurr 13.3 Declination and inclination records fn>mFish Lake. Oregon (Vcrosub et al., at Lakes (Creer and Tucholka, dcd to indicate proposed corre-

238

13 Secular Vanatwn and Brunhes Chron Excurs~ons

1987). This dilemma highlights the critical role of radiocarbon and other age determinations in efforts to establish master CUNeS of secular variation and the problems which can arise when radiocarbon dates are not backed by independent age control. The lake sediment records for the last 10 ky from Minnesota (Lund and Banerjee, 1985) and Fish Lake, Oregon ( ~ e r i . sub rt ab, 1986) can be convincingly correlated with radiocarbon-dated lava flows and archaeomagnetic records (see Brandsma et al., 1989). Elk Lake in Minnesota has yielded a high-fidelity secular variation record with a varve chronology independent of radiocarbon (Sprowl and Banerjee, 1989). Marine records of secular variation are confined to high sedimentation rate environments such asparts of the Mediterranean or Black Sea (Opdyke rt a/., 1972); Creer. 1974) and to coastal hemipelagic environments (Brandsma e! a/., 1989: Levi and Ka~lin,1989) and drift deposits (Lund and Keigwin. 1994) where sedimentation rates arc of the order of tens of centimeters per thousand years. Lake sediment accumulation rates are often of the ordcr of 1d k y , whereas pelagicmarine sedimentation rates are characteristically about 1 cmlky. The relatively enhanced fidelity of lake sediment records of secular variation is partly due to their higher sedimentation rates and partly due to the general absence of bioturbation. At DSDP Site 480 in the Gulf of California, the estimated sedimentation rate is 1 mlky and the 50-m core was sampled at 10-cm intervals for secular variation studies (Levi and Karlii, 1989). The record cannot be correlated in detail to the lake records although high-amplitude fluctuations in inclination in the 2050 ka interval may be related to low paleointensites at the time of the Mono Lake and Laschamp excursions (Levi and Karlin, 1989). In the Santa Catalina Basin (California Borderlands), estimated sedimentation rates range from 13 to 86 cmlky (Brandsma et a/., 1989). A 3-em sampling interval in 3-4 m-long piston cores has yielded secular variation records that can be correlated among three cores and with North American lake rccords (Brandsma rt al., 1989), thereby improving the chronology of the sediments lrom the Santa Catalina Basin for the last 10 ky. Lund and Keigwin (1994) noted that secula~variation records from the Bermuda Rise are relatively subdued for the Holocene (where sedimentation rates are about 10 cml ky) compared to the Late Pleistocene where sedimentation rates are 2-3 times higher. Secular variation records from Australia (Barton and McElhinny, 1979; Constable and McElhinny, 1985) are more difficult to correlate because of the generally lower amplitude of the directional changes. The records from Victoria (Barton and McElhinny, 1979) cannot he matched with those &om Queensland on a swing-by-swing basis (Constable and McElhinny, 1985); however, some features of the VGP paths can bc correlated.

13.2 Sedrmenl Records of Secular Vartallon

239

The original records from Lake Windermere and Lough Neagh were NRM rccords, and the observed oscillations in declination and inclination were considered to be periodic with a periodicity of about 2800 years. Partial alternating field demagnetization helped to produce the high-fidelity magnetic record from Loch Lomond (Turner and Thompson, 1979),which, together with an improved time scale, showed that the geomagnetic field changes did not follow a simple oscillationary pattern during the past 7000 years. Turner and Thompson (1981) noted that the calibrated master curve for Britain (Fig.,l3.1) has similarities with other European lake and archaeomagnetic records but differs from North American and Japanese records. Creer and Tucholka (1982) considered that the similarities between the UK and North American Great Lakes master curves were sufficient to indicate westward drift throughout most of postglacial time at about the historically observed rate. The general lack of similarity of record from Argentina (Creer et al., 1983b): Australia [Barton and McElhinny. 1979), UK (Turner and Thompson, 1981), and the North American Great Lakes (Creer and Tucholka, 1982) confirms that the secular variation is not due to wobble of the main dipole field axis but to nondipole components with different sources at different sites. For this reason, patterns of sedular variation cannot be expected to be correlative over large distances. This is also demonstrated by the lack of similarity of the record from the Sea of Galilee with the UK master curve (Thompson rt ul., 1985). The Western European record does, however, appear to match over distances of up to about 2000 km, for example, from Western Europe to Iceland (Thompson and Oldfield, 1986). Runcorn (1959) pointed out that westward-drifting nondipole sources will tend to cause the norfh-seeking geomagnetic vector to precess in a clockwise sense and therefore give a clockwise sense of looping of the VGP paths; however, this interpretation is not unique, and the converse can be true with some source configurations (Dodson, 1979). Presently available European secular variation data indicate a dominantly clockwisc sense or looping of the magnetic vectors and of the VGP paths implying that westward dr2t dominated the past 7000 years or so (with possibly a few hundred years of eastward drift beginning bchveen 1000 and I500 yr B.P.) preceded by a period of eastward drift (Turner and Thompson, 1981: Thompson and Oldfield, 1986). In the record from British Columbia, clockwise looping Irom 4.75 ka to 1.5 ka is followed by a period of anticlockwise Iooping (Turner, 1987). This is broadly consistent with the European record. In the Elk Lakc record, clockwisc looping predominates with anticlockwise looping in the 500-1500 yr B.P. and 6100-7000 year B.P. intervals (Sprowl and Banerjee, 1989).The Australian record showspredominantly clockwise looping, with periods of anticlockwise looping between 5.7 and 4.0 ka and

r

238

13 2 Scdimenl Records uf Secular Vanatlon

13 Secular Vanation and Brunhes Chron Excurs~ons

1987). This dilemma highlights thc critical role of radiocarbon and other agc determinations in efforts to establish mastcr curvcs of secular variation and the problems which can arise when radiocarbon dates arc not backed by independent age control. Thc lake sediment records for thc last 10 ky from Minnesota (Lund and Banejee. 1985) and Fish Lake, Oregon (Vero. sub et ul., 1986) can be convincingly correlated with radiocarbon-dated lava flows and archaeomagnetic records (see Brandsma etal., 1989). Elk Lakc in Minnesota has yielded a high-fidelity secular variation record with a varve chronology independcnt of radiocarbon (Sprowl and Banerjee. 1989). Marine records of secular variation are confined to high sedimentation rate environments such as parts ol the Mediterranean or Black Sea (Opdykc et al., 1972); Creer, 1974) and to coastal hemipelagic environments (Brandsma ef al., 1989; Levi and Karlin, 1989) and drift deposits (Lund and Keigwin, 1994) where sedimentation rates are of the order of tens of centimeters per thousand years. Lake sediment accumulation rates are often of the order of 1mlky, whereas pelagic marine sedimentalion rates are characteristically about 1 mlky. Thc relatively enhanced fidelity of lake sediment records of secular variation is partly due to their higher sscdimentation rates and partly due to the general absence of bioturbation. At DSDP Site 480 in the Gulf of California. the estimated sedimentation ratc is 1 mlky and the 50-m core was sampled at 10-cm intervals for secular variation studies (Levi and Karlin, 1989). Thc record cannot be correlated in detail to the lake records although high-amplitude fluctuations in inclination in the 2050 ka interval may bc related to low paleointensites at the time of the Mono Lake and Laschamp excursions (Levi and Karlin, 1989). in the Santa Catalina Basin (California Borderlands), estimated sedimentation rates range from 13 to 86 cmlky (Brandsma etal., 1989). A 3-cm sampling interval in 3-4 m-long piston cores has yielded secular variation records that can be correlated among three cores and with North American lakc records (Brandsma et ul., 1989),thereby improving the chronology of the sediments . from the Santa Catalina Basin for the last 10 ky. Lund and Keigwin (1994) noted that sccular variation records from the Bermuda Rise arc rclatively subdhcd for the Holocene (where sedimentation rates arc about 10 cml ky) compared to the Late Pleistocene where sedimentation rates are 2-3 times higher. Sccular variation records from Australia (Barton and McElhinny, 1979; Constable and McElhinny, 1985) are more difficult to correlate because of the gcncrally lower amplitude of the directional changes. The records from Victoria (Barton and McElhinny, 1979) cannot be matched with those from Quecnsland on a swing-by-swing basis (Constable and McElhinny, 1985); however, some features of thc VGP paths can be correlated.

239

The original records from Lake Windermere and Lough Neagh were

I. NRM records, and the observed oscillations in declination and inclination were considered to be periodic with a pcriodicity or about 2800 years. Partial alternating field demagnetization helped to produce the high-fidelity i magnetic record from Loch Lomond (Turner and Thompson, 1979),which, ; together with an improved time scale, showed that the geomagnetic field / changes did not follow a simple oscillationarypattern during the past 7000 I[ years. Turner and Thompson (1981) noted that the calibraled master curve for Britain (Fig. 13.1) has similaritieswith othcr European lake and archaecI plagnetic rccords but differs from North American and Japanesc iecords. Creer and Tucholka (1982) considered that thc similarities between the UK and North American Great Lakes master curves were sufficient to indicate westward drift throughout most of postglacial time at about the historically observed rate. The general lack of similarity of record from Argentina (Crecr et al., 1983b), Australia (Barton and McElhinny, IY79), UK (Turner and Thompson, 1981), and thc North American Great Lakes (Creer and Tucholka, 1982) confirms that the secular variation is not duc to wobble of the main dipole field axis but to nondipole components with different sources at different sites. For this reason, patterns of secular variation cannot be expected to be correlative over large distances. This is also demonstrated by the lack or similarity of the record from the Sea a£ Galilee with the UK master curve (Thompson er ul., 1985). The Western European record does, however, appear to match over distances of up to about 2000 km, for example, from Western Europr, to Iceland (Thompson and Oldficld, 1986). Runcorn (1959) pointed out that westward-drifting nondipole sources will tend to cause thc north-sceking geomagnetic vector to precess in a clockwise sense and therefore give a clockwise sense of looping of the VGP paths; however, this interpretation is not unique, and the converse can be true with some source configurations (Dodson, 1979). Presently available European secular variation data indicate a dominantly clockwise sense of looping of the magnctic vectors and or the VGP paths implying that wcstward drift dominated the past 7000 years or so (with possibly a few hundred years of eastward drift beginning between 1OOO and 1500 yr B.P.) precedcd by a period of eastward drift (Turner and Thompson, 1981; Thompson and Oldfield, 1986). In the record from Brirish Columbia, clockwise looping from 4.75 ka to 1.5 ka is followed by a period of anticlockwise looping (Turner, 1987). This is broadly consistent with the European record. In the Elk Lake record, clockwise looping predominates with anticlockwise looping in the 5N-1500 yr B.P. and 6100-7000 year B.P.intewals (Sprowl and Banerjce, 1989).The Australian record shows predominantly clockwise looping, with of anticlockwise looping betwecn 5.7 and 4.0 ka and j

240

13 Secular Vanallon and Brunhes G7mn Excunronr

between 10.5 and 8.8 ka. Within the age constraints, this sense of looping is consistent with thc Argentinean record (Constable and McElhinny, 1985). The record from Lac du Bouchet for the 10 ka to 30 ka interval is consistent with predominantly westward-drifting geomagnetic sources (Smith and Creer, 1986). Although thcrc is some consistency between infcrred drift direction in the northom hemisphere and UI the southern hemisphere. there is no clcar global consistency, which has led Creer and Tucholka (1982) to infcr that the looping may be largely due to "standing" sources which fluctuate in intensity rather than westward or eastward drift. Crosscorrelation of the Bulgarian archaeomagnetic record (Kovachcva, 19R3) with the Elk Lake (Minnesota) record yields a peak a t an offset of 520 years, implying a westward drift rate of 0.23"lyr: however, there is little consistency in the drift rate or the drift direction when the same exercise is performed for Eurasian andNorth American records (Sprowl and Banerjec, 1989). In a recent review of North Amcrican Holocene secular variation, Lnnd (1996) found no evidencc for westward or eastward drift. According to ~ u n (1YY6), d repetitive vector loops may indicate a recurring core generation process manifest in the secular variation record. A global compilation of local changes in declination and inclination for the past 400 years mainly from observatory records (Fig. 13.4) (Thomp-

j Vlml

Geomagnetic Poles (1600-1975 AD)

Figure 13.4 Historical vinual seomagoetic poles ovcr rhc Earth'? surfacc for thc 1600-1975 interval. The sense of motion is clockwisc cxccpt in the rryion of the Indian Ocean (arm Thompson and BarraclouEh. 1982).

24 1

13 2 S e d ~ m n Records t of Secular Var~at~an

son and Barraclough, 1982) illustrates that the amplitude of the secular variation and the timing of distinctive turning points in the records vary from place to place, with the most rapid high-amplitude secular changes and clearest turning points occurring over Africa and Europe. These are therefore the most promising regions for the use of secular variation as a historical correlation tool. Although high-quality relative paleointensity records have been obbined from marine sediments (see Ch. 14), paleointensity studies in lake sediments have generally been less succcssful. Roberts et ul. (1994b) pre_sentedapaleointensity rccordforthe65-105 ka interval from Lake Chewaucan (Oregon) which indicates an NRMIARM paleointensity high close to 80 ka. consistent with marine paleointensity records (Meynadicr el al., 1992: Tric et at., 1992). The paleointensity records from Lakc Baikal (Peck el al.. 1996) derived from NRMIARMin three 9m corcs yield anexcellent match to marinepaleointensity records for the last 84 ky. Inlake sedimcnts, anhysteretic remanence (ARM) and saturation isothermal remanence (SIRM) arc often inadcquate as a means of normalizing for variable concentration of ~manence-contributinggrains, possibly bccause magnetic grains included in a largernonmagneticgrainscontribute to ARMand SIRM but not to remanence (Turner and Thompson, 1979). Tucker (1980. 1981) has pointed out the difference in nature of the naturalDRM andof ARM and hassuggested the use of stirredremanenrmagnetization (StRM)asa normalizingparameter.Thouveny (1987) hasdemonstrated that StRMcan havethesamestability against magnetic cleaning as the NRM and that this can bc the most suitable normalizing parameter. Howevcr, laboratory redcposition of natural samples in known magnetizing fields may no! adequately mimic ihc relationship between NRM and magnetizing field because of the role of organic gels, which appear to bc important in stabilizing grain orientation in the natural wet sediment (Stoher andThompson, 1979). Clearly, itwould be highly dcsirable to have an adequate means of determiningpalcointensities. not only as an additional mcans of corrclation but also to have acompletc appreciation of the spatial and temporal nature of geomagnetic secular variation. Within an individual lakc, cores can be correlated using tie-points derived from lithological matching and magneticsusceplibility records. A general procedure is to transpose the depth scales l'or each core to that of a lake master core by Linear interpolation betwccn correlation tie-points. In order to correct for variations in tilt of the corer between holes and to find the optimal match between the directional secular variation records from the same lake, several workers (c.g., Constable and McElhinny, 1985; Turner, 1987) have used the cross-correlation technique suggested by Denham (1981). At eachstep in theiteration, the angle ofrotationbetwccn two records iscalculated to maximize thc sum of thescalar products bctween correspond-

-

242

13 Secular Var~atranand Brunhea Chron Excuntons

ing unit vectors. In order touse this method, data points of equal depth in the cores must be (linearly) interpolated for each record. The cross-comelation technique appliessimplerotationsto the whole dataset, making no alteration to the depth or time scales. Correction for thc tilt of the corcr is a criticalstep as Mackereth-type corers have operated efficiently at tilts up to 35" (Turner, 1987).Matched magnetization vectors from diffcrcnt cores can thenbe aver. aged to produce a smoothed master curve for the lake, with estimates of confidence limits using Fishenan statistics. Smoothing of individual records can be achieved by fitting polynomials or splines. Spline functions are generally preferred to polynomials for approximating smooth empirical functions. The commonly used cubic spline function comprises a piecewise polynomial of degree 3 joined at "knots." For a given sct oi knots, the parameters of the best-fitting cubic spline can be estimated by least squares. Thc number of knots determines the degree of smoothing and, if equally spaced knots are used, thc optimum knot spacing (bandwidth) can be determined by the cross-validation technique (Clark and Thompson, 1978), which involves the repeated comparison of the smoothed record with randomly chosen data points temporally deleted from the record. The optimum bandwidth is that which minimizes the average sum of sqnarcs of thc differences between cxcluded observations and the smoothcd curve. This process also permits an estimate of the variance of the error measuring the scatter of observations from the smoothed curve. and this is essential for thc estimation of confidence limits (Clark and Thompsbn, 1978).The method can be applied to the smoothing of scalar quantities such as susceptibility or to directional data. Smoothed records can then be wmelated by visual matching of distinctive features such as turning points in the directional records. Widely available computer programs such as Analyseries and Corepack are useful for optimizing core carrelation by matching distinctive features in the records. A less subjcctive method of correlation of betwccn-lake records involves determining the stretching function (transformation of depth scales) which results in the optimal fit between records (Clark and Thompson, 1979). The stretching function may be nonlinear, and the method can give valid confidence limits.

1988).These features may be either extreme excursions in secular variation, aborted reversals of the geomagnetic field, or short polarity subehrons with duration -lo4 years. In general, the excurs~onshave prwed to be too short in duration and difficult to detect to be particularly useful for stratigraphic correlat~onpurposes, but they are potent~allyvery important for our understanding the nature of the geomagnetlc field It is clear that some of the proposed geomagnetic excursions are based on spu~iousdata. For example, the Gothenberg excursion orig~natedfrom studies of Swedith varved sediments (Morner et a l , 1971). Attempts to *replicate the results in sediments from nearby lakes werc unsuccessful (Thompson and Berglund, 1976). Both the Gothenberg excursion and the Erieu excursion (Creer et al., 1976) can bc attributed to the effects of bottom currents and abrupt facies changes (Thompson and Berglund, 1976; Banerjee et al., 1979). Physlcal disturbance of sediments, Inversion of core segments. and self-revcrsal mechanisms in igneous rocks may all yield result? that could be erroneously rnterprcted as geomagnetic field behav~or. There is a major problem assoc~atedwlth the recognition of geomagand netic events which may have very short duration. local man~festat~on, differing character from one place to another. Inability to dupl~catcthe record of such geomagnetic events may not mcan that the events do not exist; however, the lack of duplication casts doubt upon the validity of the origmal record. Nonetheless, four Brunhes Chron excursions have been documented at a number of locations and are now generally cons~deredto represent perturbations of the geomagnetic field (Table 13.1). Mono Lake Excursion: The Mono Lake excursion, fir51 documcnted by Denham and Cox (1971) and subsequently Investigated by Lidd~coat and Coe (1979), is arguably the best documented excursion of the geomagnetic field during the Brunhes Chron. The excursion (dated at 27-28 ka) is followed by four successive recurrences of the excursion waveform (Fig.

Excurslan

13.3 Geomagnetic Excursions in the Brunhes Chron There is a largc number of papers in the literature advocating geomagnetlc "excursions" (Opdyke, 1972) with~nthc Brunhcs Chron, and the evidence for theu existence is controversial. Some have suggested that a full reversal of the Earth's magnetlc field has occurred many tlmes withln the Brunhes and that the excursions should be classified as subchrons (Champion el a l ,

243

13.3 Geomagnct~cExcursions In the Bruntlm Chron

AEC

Mono Lake

27-28.000 )ears B P

Laschamp

42.000 years B P

Blake

IOR-112.UUO years B.P.

Pnnglc Falls Big Lost

218,000 t lO$W years B.P. 565,000 yean K.P.

Rrfe~ences Denham and Cox (1971); L~ddicoatand Cae (1979) Bonhomel and Rehklne (1967). 1 em el Ql (19901 smith a i d Poster (196% Tric er 01. (1991h]; Zhu et "1. (19Y4) Herrero-Bervcra et a1 (1989. 1994) Negrini EI a(. (1987): Champion er al. (IYW)

244

13.3 Geomagnetic Excursions in the Brunher Chron

13 Secular VaTlal~nnand Rmnhrs Chron Excuralonv

13.2) (Lund et al., 1988). The directional changes have been closely duplicated at four sites on the shores of Mono Lake, allhough some sites at this locality with identical tephrostratigraphy Iailed to record the cxcursion, exemplifying the problem associated with the recognition of such events. The excursion is recorded as a declination swing to thc west by 60"followed by a swing to the east by about 40" and a swing in inclination to -30" (Liddicoat and Coe, 1979).The duration oflhe excursion has been estimated to be about I# years. Lack of recognition of the Mono Lake excursion at Clear Lake (Verosub, 1977b) or Pyramid Lake (Verosub et al., 1980). located about 300 and 200 km from Mono Lake, respectively, has been uscd as evidence that the Mono Lake record is spurious (Verosub el ul., 1980). These authors considered contemporancous nondeposition or erosion at Clear Lake and Mono Lake to be highly unlikely. This argument depends critically on the duration of the event; clearly, if the duration was short, recording of the excursion may be due to fortuitously rapid accumulation rates at Mono Lake at this precise time. Two lake records appear to yhow geomagnetic excursions of cornparable age to that recorded at Mono Lakc. These records are from Lake Tahoe (Palmer et ul., 1979) and from Summer Lake (Negrini et al., 19841, which are 1000 and 600 km from Mono Lake, respectively. Although there are no good age constraints or demagnetization dala on the Lake Tahoe record, the Summcr Lake excursion is precisely dated by tephrostratigraphy and the magnetic record is well documented. The character of the excursion at Summer Lake is similar to part of that recorded at Mono Lake. The fact that thc entire excursion is apparently not recorded at Summer Lake and that the record of the excursion has proved so elusive in lakes with a similar sedimentation history indicates that the excursion has very short duration, probably less than 103years. This in turn leads us to thc conclusion that this excursion will be of little nsc as a means of correlation, although it is wcll dated both at Mono Lake and Summer Lakc by tephrostratigraphy as being in the. 24-29 ka range (Liddicoat, 1992; Negrini el aL, 1984). The fact that this age correlates closely with the 27-30 ka age for the Lake Mungo excursion (Barbetti and McEUlinny, 1976) is an interesting coincidence, but most probably fortuitous. Laschamp Excursion: This excursion was originally detected by Bonhommet and Babkine (1967) in recent lavas of the Puy dc Dome region of the Massif Central, France (see also Bonhommet and Zahringer. 1969). The best estimate for its age seems to he 42 ka (Condomes et aL, 1982). Detailed studies on these lavas by Hcller and Petersen (1982) indicate that they possess sclf-reversing properties: therefore, the existence of the excursion may be in doubt. The observed directions of magnetization make large angles with the present geornagnctic field direction but are not fully

245

reverse polarity. Coring in nearby lake sediments which span the samc time interval failcd to detect the excursion (Thouveny and Creer, 1992). Recent studies of the Lonchadiere volcanic flow from central France yielded deviating magnetization directions and the lava is the same age as the basalts at : Laschamp and Olby. Icelandic lava flows which yield ages indistinguishable from those for the Laschamp excursion give paleomagneticresults which are in accord with the Laschamp excursion (Kristjansson and Gundmundsson, 1980; Levi et al., 1990). Based on thcsc results, the geomagnetic excursion appears to be confirmed. Blake Excursion: Another geomagnetic excursion within the Brunhcs i.. Chron which has been well documented is the so-called Blake Event, This excursion was first recorded in the Atlantic (Smith and Foster, 1969) and subsequcntly both in the Atlantic and in the Caribbean (Denham. 1976). This excursion appears to he a complete, albeit short-lived, reversal of the geomagneticfield and therefore may qualify as a subchron. The appearance of the Blake Event in the sedimentary record is variable, depending, presumably, on the completeness of the record rather than vagaries in the remanence acquisition process. It appcars as a single reversal (Tucholka et al., 1987) and two reverse intervals (Creer et al., 1980b;Tric et aL, 1991b) in Mediterranean cores and as three more complete reversals in Atlantic cores (Denham, 1976) and in Chinese loess (Fig. 13.5) (Zhu et al., 1994). The Blake Event has now been recorded in China and in the Atlantic and Mediterranean and therefore appears to bc a global event. Age estimates for the Blake Event from Atlantic cores (100 ka, Denham, 1976) compare ; closely to age estimates from Chinese loess (111-117 ka, Zhu et al., 1994). In Mediterranean piston cores, ncgative inclinations correlate to oxygen isotopic sllbstages 5e and 5d (Tucholka er al., 1987; Tric el ul., 1991h). Tephrochronology and S L 8 0stratigraphy in these cores indicate fully reverse magnetization directions over a 40-cm interval which the authors estimate to represent -4000 years. The event is centered at about 110 ka, close to the age originally suggested by Smith and Foster (1969), and best estimates of thc duration of the event lie in the 4-6 ky range. Pringle Falls Excursion: The three partially reverse polarity intervals recorded in lake sediments from near Pringle Falls, Oregon (Herrero. Bewera el al.,1989; Herrero-Bervera and Helsley, 1993) were first thought to be a record of the Blake Event. Thc excursion at Pringle Falls has now hecn correlated to similar directional records of Geld excursions at Summer Lake (Oregon) and Long Valley (California), and 40Ari39Ardating and tcphrochronology indicate an age of 218 ? 10 ka (Fig. 13.6) (HerreroBervera er al., 1994). Anomalous magnetization directions from the Mamaku ignimbrile in Ncw Zealand (Shane et al., 1994) yield virtual geomagnetic poles which lie close to the VGP path from OregonICaliforniarecords '

246

13 Seoular Vanation nnd Brunhn Chron Excursions

247

13 3 Geomagnehc Excurs~onsm the Bmnhes Chron

DECLINATION

Long Valley

I NCLI NATI ON

Pringle Falls

I

I

e

270go-

Decl~ndllon( " )

I

Figure 13.5 Record 01 I h r Blake Even1 from Chmese loesr (after Zhu

I

7

" 01

"

e( of,1994)

of the Prlngle Falls excursion. Isothermal plateau fission track age on glass (230 -C 12 ka) (Shane et a!., 1994) and the 40Ar/39Arage on plagioclase (223 ? 3 ka) (McWilhams, personal cornmumcation, 1995) from the Mamaku ignimbrite are consistent w ~ t hthe est~matedage of the Pr~ngleFalls excursion in OregonICalifornia Big Lost Excursion: Champion et al. (1988) have identified reverse directions in bore holes and outcrops from Idaho dated at 565 f 14 ka. They called this event the Big Lost subchron.The duration of this possible excursion (or subchron) 1s not known. A reverse interval recorded in Humbolt Canyon (Negrini rt al., 1987; Champ~onet al., 1988) and low-latltude poles from volcanlc rocks in southwest Germany dated at 510 f 30 ka have been correlated to thls excursion (Schnepp and Hrddetzky, 1994). Apart from those listed above, Champion et al. (1988) have postulated the existence ol' five additional subchrons within the Brunhes based on data from the Lake B~wacores (Kawai, 1984) and from old piston cores (e g., Goodell and Watk~ns,1968; Steanvald er al., 1968, Ryan and Flood, 1972, Watkins, 1968), most of which should not be used for this purpose. Early piston core data containing anomalous magnetization directions (e.g Ninkovlch et a l , 1966) were reinterpreted by Wollin et a1 (1977) as recording field-related behavior The cores in question had sedimentation rates of 1d k y or lessand it is, therefore, veryunlikely that short-term geomagnebc field hchavior would havc been recorded.

.

g d

k.4>

//

Summer Lake AR-C

\ , I

1

~8

!

I

I I I I I

I

11150

am Depth (cm)

Depth (om)

Figure 13.6 Declination and inclination of the magnetizstion in sediments from California and Oregon which record the Pringlc Palls eeotnagnctic excursion, with the correlarion of tephra among the sectiunr (after Hcrrrro-Bervera cf oL, 1994). Tephras are letter coded to aid conelation lrom the inclination to lhe declination records.

In recent years, abundant data have become available from Brunhes age sediments with rates of sedimentation above 1 cm/ky from both cnnventional piston cores and HPCIAPC cores of the Ocean Drilling Program. High-quality data are also available from loess scctions of China. In Figure 13.7, thcse are plotted as straight-line scgments assigning the top of the cores to zcro age and the BrunhesIMatuyama boundary to 780 ka. In at least one study of locss in t l i n a (Heller and Liu, 1984), dcnse sampling was carried out in an attempt to detect the Blake and Laschamp subchrons; however. no deviating diucctions were observed. The core with the highest rate of sedimentation is from ODP Leg lU7 (Site 650) in the Tyrrhenian Sea, which has a sedimentation rate of 56cmiky (Channel1 andTorii, 1990). Two zones of negative (upward) magnetic inclination are secn in this core:

248

13 Secular Varlatlon and Brunhes Chron Excursions

13 1 Crrwnagnet~cEncur\~cm*m t h e Brunhea Chron

j

:

Stratigraphic depth (m) Figure 13.7 Positiota or r e v e r r polarity intervals in Rrunhcs agc sediment corer and locrd sections (1.1 and L3). Brunher rccrinns range in length lrilm 8 to 400 m. Site 650 lrmn thc Tyrrheniim Sea is plotted from right to lcft with an independent depth scale increased hy a factor o [ 4. Thc p o s i f i o ~ ~ sthe o f reverge polarity samples or i n t r w l s arc indicated by the arrows or in the case of G I (Italy) by a gap in lhc straight line. It cam be are" that a correlation of reverse polarily inLerrals is not compellirlg. The pusitions in time o f t h e Muno 1,akc (Mo), ].aschamp (L).Blakz (EL), Jamaica iJa). Levantine (Lr). Biwa Ill (BilII). Emperor (Em). Big 1.nst (BL), and Delta rxcuniims arc shown.

the first (A) occurs at a depth of 10.39 m to 12.31 m and the second occurs at a depth of 100.16 to 100.45 m. When plotted on a time versus dcplh plot, event A would seem to correspond to the Mono Lake cxcursion. Event B occurs a1 an extrapolated age of about 175-ky, an apparent age older than expected for the Blake excursion. Hole h42B (Fig. 13.7) has zones of negative inclination which appear to corrclale with the Mono Lake and Laschamp excursions. The Blake Event, on the other hand. does not appear to correlate wilh negative inclinations in any of the cores and sections examined. This is also true of the Big Lost excursion. Studies of reversal transitions indicate that the lcngth of time taken for a transilion t o occur is -5-10 ky. The shortcsl subchron duration you could expect in which Lhe field goes cornplctely reverse would bc from 10 lo 20 ky. This is, in fact, the length of timc estimated for the Cobb Mountain and Reunion subchrons. If full reversals of the magnetic ficld can take as

~

249

little as 1 ky, as sccms to be indicated from the results by Tric et al. (1991b). then very short polarity subchrons may occur. For correlative purposes, however, such short subchrons would seem to be of marginal value except in regions of very high sedimentation ratcs. Excursion records and records of short polarity evcnts such as the Cobh Mounlain subchron (see Clcment and Martiuson, 1992) imply that the geomagnetic field cxhihits a complete spectrum of behavior frum high-amplitude secular variation tu full polarity revcrsals. Brief polarity reversals which never stabili7.c. where the dipolc field immediately reverts back to itsinitialpolarity, have also been recorded. high-fidelity records of the process of polarity reversal, recorded ~Detailed . in sediments with high accumulation rates, have occasionally indicated that the dipole configurationof the ficld is maintained during the reversal proccss for certain polarily transitions (see Clcment. 1992; Clcmcnt er al., 1995) and that the paleomapetic pole follows preferred longitudinal paths during the reversal proccss (Clement, 1991: Laj et al., 1991) (Ch. 2).

14 2 Magnet~cParameter$ Sensnttw to Concennntlon, Gram Size, and Mnncralogy

Rock Magnetic Stratigraphy and Paleointensities

14.1 Introduction Rock magnetic stratigraphy involves ihc utilization of magnetic of sediments and sedimentary rocks as a means of (1) stratigraphic corrclation. (2) idcntificetion o l sediment sources and transport mechanisms. (3) characterizationldetection of paleoenvironmcntal changt.. The first rock magnetic parameter to be widely used for corrclation was initial (low field) magnctic susceptibility (k). Its usefulness has long been recognized particularly for correlating volcanoclaslic-rich horizons in marine sediments (e.g.. Radhakrishnamurty era/., 1968). and it is now a standard corrclation tcchniquc in deep-sea sediments (e.g., Robinson et aL, 1995), lake sediments (e.g.. Thompson rl a!., 1975; Peck et ul., 1994), and in locss (c.g., Kukla et al., 1988). Magnetic susceptibility essentially monitors the concentration of ferrimagnetic minerals such as magnclitc and maghemite. and therefore the correlation method is sensitive in sediments which show variations in concentration ofeither of thcsc magnelic minerals. Magnetic susceptibility measurements are part of the standard core scanning procedure on RN Joirles Re.solutir,n and other oceanographic research vessels. High-resolution paleoceanographic studies require not only multiple cores at each sitc (cnsuringcomplctc recovery of the sediment sequence) but also the ability to make high-resolution correlation among cores. An important step forward in shipboard cure correlation was accomplished during ODP Leg 138 (Lyle era/., 1992; Hagelberg el a/., 1992). Specialized computer sollware linked lo core scanning systems recording saturated 250

251

bulk (GRAPE) density, magneticsusceptibilily, and digital color rcflcctancc spectroscopy produced optimal correlation of physical parameters between cores. This was tho basis for "real-time" core correlation within site, the construction of a composite section for the site, and correlation to downholc' logs. Real time within site correlation can be used to determine the number of holes necessary to ensure cumplcte recovery at a sitc and to compensale for the poorly understood phenomenon o l corc expansion (often up to 10,20%) during recovery. The field of rock magnetic stratigraphy has broadened considerably in the last 15 years with the use o l magnetic parameters which are sensitive not only to magnetic mineral concentration hut also to the grain size and mineralogy. These techniques were largely developed to study environmcntal changes in lakc catchments by monitoring changes in lakc sediment magnetic mineralogy which are sensitive to changes in land use, erosion, industrialization, and climatc (see Thompson and Oldfield: 1986). Similar methods have been applied to marine sediments (Robinson, 1986; Bloemendal et al., 1988; Doh et a/., 1988) and loess deposits (Maher and Thompson, 1991.1992) and have been shown to provide not only a sensitive means of corrclation but also information on changing detrital and hiogcnic fluxes. The magnetic parameters useful [or correlation andlor environmental modeling depend on the nature of the variations in magnetic mineralogy within the sediment.

14.2 Magnetic Parameters Sensitive to Concentration, Grain Size, and Mineralogy The magnetic parameters listed below and in Table 14.1 are the principal parameters used ~n ~tudiesof marlne and lake sediments (see Thompson etal., 1975. King el a l , 1982; Thompwn and Oldheeld, 198e Robinson, 1986: Bloemcndal et al., 1988, Doh el a l , 1988; Kmg and Channell, 1991; Stoner rt a1 , 1996).

a. Magnetic Susceptibility (k) Volume (initial) magnetic susceptibility. defined as volume magnetization divided by applied field, is often considered to be a measure 01 fcrrimagnetic mineral cunccntration: however. susceptibility (in magnetite) shows a grain size dependence (see Maher. 1988) increasing through the SDPSD-MD range. As the susceplihilily of the ferrimagnctic tninerals magnetite and maghemile is 3 or 4 orders of magnitude greater than that of common anlifcrroma-~eticminerals. such as hematite and goethilc.

252

14 Rock Magneuc Strat~graphyand Paleotntensltrrs

the susceptibility is usually dominated by the ferrimagnetic grains and therefore gives information on the presence or ahsence of this group of magnetic minerals. For magnetite- and maghemite-bearing sediments. magnetic susceptibility is a reasonable measure of the concentration of the magnetic mincral. and is not in itself a sensitive measure of grain size.

143 Rock Magnct~cStratigaphy m Msnne Scdlrnents

or hematite, Lrn,(determined using a back field of 0.3 T) would be more appropriate. The HIRM (.'hard3' IKM) is the difference between the IRM o,3 (IRM in a 0.3-T field) and SIRM divided by 2 (Robinson. 1986: Blocmendal er ab, 1988). HIRM-0,;

b. Anhysteretic Remanent Magnetization (ARM) or Anhysteretic

Susceptiiility (k,,) and Saturation Remanent Magnetization (SIRM) These parameters are also concentration dependent and, as for susceptihility, are dominated by the fcrrimagnetic grainspresent in the sample. These paramctcrs arc particularly useful because their magnitude tends to increase with decreasing grain size of the ierrimagnctic grains (see Maher, 1988). Normalizing these paramelcrs by initial susceptibility partially compensates for variations in concentration and hence the SIRMIk ratio (Thompson. 1986) and ARMlk or k,,,lk ratio (King ef al., 1982) vary inversely with magnetic particle size and are therefore useful granulometric paramctcrs. The former is likely to be more sensitive lo the multidomain grain size range and the latter to the pseudo-single-domain range. The granulometric method utilizes the fact that A R M and SlRM dccrcase with increasing grain size whereas susceptihility is relatively insensitive. showing slight increase with increaing grain size. Interpretation is complicated by the nonlinear response of these parameters lo changes in grain size. An additional drawback of all three granulometric ratios is that ultrafine superparamagnetic (SP) grains will have negligible remanence (SIRM, ARM, k,,,) but relatively high k, resulting in low values of the ratios (STRMIk ARMlk, or k,,,lk), comparahle to those seen in coarse mulridomain grains.

c. S-Ratiu and "Hard" IRM (HIRM) The S-ratlo (Stober and Thomp~on,1979, Bloemendal. 1983) IS found by measuring the SIRM and then plac~ngthe sanlple in a back held (typically 0.1 T) and remeasuring the IRM

253

= (SIRM

+ I R M ,,_r)l2

. This parameter is also a measure of thc concentration of antiferrumagnetic minerals with high coercivity. The two parameters (S-ratio and HIRM) compIement cach other and might be expected to show an inverse relationship in cores exhibiting a wide range of magnetic mineralogy and grain size, The combination of these two paranieters provides a means of monitoring changes in tnagneric rrrineralogy and magnetic mincral concentrations. Thc coercivity range in which the S-rado and IIIRM arc sensitive will depend not only on Lhc value. uf ihc back field hut also the magnetizing field used to acquire the SIRM. Note that, in practice, the "SIRM" is often acquired in a magnetizing licld of about is close to thc maximum licld attainable hy most small 1 T, electromagnets and pulse magnetizers. This field may not fully saturate the samplc, particularly in the prescnce of high-coercivity antifcrromagnetic minerals such as goethilc or hematite.

d. Frequency-Dependent Susceptibility The susceptibility of fine-grained magnetite. with grain sizes close to the superparamagnetic/single domain boundary, is frequency dependenl, dccreasing wilh increasing Crcquency o i ihc applied field (Stephenson. 1971). The susceptibility of single-domain and mulridomain g~ainsis relatively unaffected by changes in frequency. The frequency-dependent susceptihility (ktd) is often exprcsscd as a percentage, whcre krd= (kl,i- k~l.)lk~i and khrand k,, a r t Lhe susceptihility in the high-frequency (typically about 5-10 kHz) and low-hequcncy field (Lypically about 1 kIIz), respectively. The frequency-dependent susceptihility is a convenient mcasure of thc coi~cmtrationof very fine grained ( S P ) mugnerite in lhc sample.

14.3 Rock Magnetic Stratigraphy in Marine Sediments s-0~ is a

convenient measure of the proportion of (coarse) low-coercivity ferrimagnetic grains TO higher coercivily grains. As the proporrion of low-coercivity grains saturating in fields less than 0.1 T increases, this Sratio approaches uhity. For discriminating fcrrimagnetic grains (such as magnetite) from high-cocrcivity antiferromagnetic grains such as geothite

Over 15 years ago, the rccognition ororhital frequencies in NRM sediment records Icd to the proposal that geomagnetic field intensity was orbitally modulated (e.g.. Wollin r+r a!.. 1977). I t has since bccn demonstrated that these observations arc best explained by orhitally controlled climatic varia-

254

14 Rock MagnCtic Stratigraphy and Paleolntensittrs

ToMe 14.1

.

Generalized Table of Downcore Magnetic Parameters and Their Intwerprmions

Tmbk 14.1

conrlnued Parameter

-

Ralk Mamerrc Meostt,cmoir~ Natural Kcmancnt Magnutiration i)cpcndcnt nn mineralogy. concentruti<~n. and (NRXf): I b c fossil (rcmancnt) -erain size of the ~uaeneticmaterial ds wcll a. modc of acquisition of rzmanencc and intensity and dircnion o l the ~~~~~~pruc declination. inclination, and inrcnrity. fiuld. Volumetric Marnctiu Lsci~tihi111\' k is a first-ordcr mcasurc of thc amount o i . lk): , . 4 measurc uf the cnnccntratlnn of fcrrimagnctic material (c.8.. magnetite): i. ih magnrtizahlr material. Dvfind er thc particularly cnhanccd by supcrparamaenztic ratio of induced ma$nrli,ati<m (SP)magnI0 & m ) Whcn the strength of the applizd weak field coficentralicm r,f lrrrim~gneticmaterial is (H): k = M!H. low. k responds In an~iilnumagnrlic[e.#.. hematite), paramagnetic (ug.. Fe. Mg silicates). and diamaenetic marerial leg.. calcium carhol~ntc.silica) a,hich may cllmpliciile the infcrprctation. Isothermal Remanent Magnetization SIRM primdrily depends on thc conccntration (IKM): .Magncric rcmancncc ~ c q u i r e d of nlagnetic. principally rrrrimagnelic, undur thc influcncc of H strong L>C matciial. It is grain sire d c p m d m l , being ficld Cummrmly cxprcssed as a particularly Jcnsitive t o magnetirc grains ~dturstiunIRhI or SIRM when a iicld smsllcr than a fcw tcnr 61 microns. gvrater than i T i > used. A hack lirld IRM (BIRM) is that i l q u i r r d in J, =versed DC field alter SIRM acquisition. ~

S ralior: These arc dcrivrd by impaning

~~

hissing I1C ficld within a decreasing hou~evcr.it is also stronzly grain size illternling field. Commonly dcpendcnt. k n u prcfcrcntially responds to expressed as anhysteretic smsllcr magnctitc g r n ~ nsizcr (<1U wm) and susceptibility (kAxx,) whcn ir ir\cful in the derrlopmenL rtl grain s i x nlrrmillircd hy thc hiasjng field dependent rntios. used. Conrinrrrcd iMngnrilr ParanzNl.rs The "hard" IRM (JIIRhl): This ir H l R M i$ a measurc uf the concentration of dcri%d by impirrtins a back lieid. mil~nrticmalerial with higher ccrrrcivitv than tvpically (1.1 or 3.0 T. on a samplu the-back field. Thir confnlbnly gives prrviuusly givcn a S l R M The information on the concet~tratiocof rraulling RIRhI. which has a nqativc anrifcrromagnctic ( c . ~ .hcmatito) . or vely sign. is used to derive Lhc HIRM by Rnr grained fcrrimagnrlic (r.g., magnctitc) the fornula: HIRM = (SIRM + grains dvpendinp on the back field used. BIRM)/?.

255

14.3 Rock Magnetlc Stratigraphy m M a m e Sediments

.

lnrcrprrtalion

Tbc S ratios can be wed to estimate t h c

magneric mineralogy (e.g., magnetite o r hemulite). Downcorr variations may bc associatcd with changing mineralogy. Values d o s c tu -1 indicate lower coer~ivitgand a fcrrimagnetic mineralogy (e.g.. magnctilr): values closer to zero indicatc a higher comcivity. possibly an antiferroma@etic (r.g., iiemafitc) mineralogy. S ratio with a back tield 111 0.1 T may bc scnsilive t o minrralugic and gmin-sire changes. where* the S ratio with the D.3 T back field is more sensitive to n~inrralugiclchimgrs (c.g.. proportion of magnetite to hematite). II the magnetic mineralogy is dominantly k,,,lk: Indicatcs changes in magnetic magnetite, kAR~!kvaries utverscly with grain sire, il Ihe magnetic mineralogy magnetic grain ske. particularly in ihe 1is dominantly mngnctitr. 10 +m grainsize rangc. Hnwever. thc interpretation of this ratio may hr. complicatcd by signiticant amounts of supcrparamagnetic (SF) o r paramagnrlic material. If thc magnetic mineralogy ir; dominantly SIRMlli: lndicatcs c h a n ~ e in s mametic magneGte, SLRMlk miles invrrsely with g a i n s i m , il the magnetic minc~alagy magnctic particle sizc. StRMlk is marc is dominantly mgnetite. sensitive than kanw!k t o changcs in the ) SIRMI proportion uf large (>ID ~ m grains. k may also be compromi.md by SP o r paramagnetic material. SIRMlkaRMincreases with increasing magnetic SIRMlknRM:Indicates changes in grain size but is less sensitive and can be magnetic grain size,if the magnetic more difficull to interpret than the tu,o rados mincralogy is dominantly magnetite. abo5.e. A major advantage of S I R M I ~ A R is M that ir only reaponds to remanence carrying magnetic matcrinl and is therefore not afiectcd by SP or paramagnetic material. k, is used t o indicate the presencc o f SP Frcqurncy Dependent Magnetic matcrial SP material in high concentratirlns Susceptibility (k,): The ratio oilawcan compromise ihr grain-sizc interpretation frequency k (U.47 kHz) to high madc using k ~ a d and k SlRWk. frequency k,,, (4.7 kHz) calculaled by kt = 100X(k - khi)ik. Hystrrcsts Measrrrcmcnif M,,/MM decreases with increasing magnetitc Saturation Magnetization (M,): M, is grain sire in thc submicron to few tenr of the magnetization within a saturating microns grain size rangc. field: Saturation Remvnence (M,,): M,*is the remanencc remaining after removal of the saturatine ficld. a back field, typically U.1 or 0.3 T,on a sample previously giviven a STRM. The resulting BIRM, which has a ncgative sign, is normalized by the SIRM;S = RIRMISIRM. This provides a measurc of the proportion of saturation st the back ficld applied.

~

256

14 Rack Magnetlc S m t l g a p h v and Palet,mtcnslt~rs

Table 14.1 continued Parameter Coercive Force (f?,.]:Thc back field requited irr rulate saturation magnrtirarim lu zero within a n applied field; Ccirrcivity o f Rernanence (fir.): Thc hack ficld requlred to rotate saluratiun magnetization to zero rrmanmcc.

257

14.3 Rock Magnetic Slratigraphy in Marine Sediments

O w cycles k V 1

lntcmctat~on For ma.~gnelite, thc ratio H,,/H, incrcase with increasing grain size (in the submicron to several hundrcd micron grain-sirr range) due t o the strong grain-size dependence of both pu'amsters, rdrticularly H,. Hi,is a useful guide l a magnetic mineralogy.

Note: After Stoner cf 01, 1996. NHysteresis parameters provide a mwnr of monitoring grain-sire variatirms in magnetite. Sediments should he homopenrnu~because ui thc small sample sire (-:OOi g) typically use11 for hystcrcsis measurements. Mired magnelic mineral asscmblad,es greatly mmplicatc the jnterprctstions, i h c g~neralizedinterpretations lislrd ithove arc based 00 a ferrinmgnetic (r.g.. magnetite) mineral assemblage.

tions affecting magnetic mineral conccntrations, which in turn affect NRM intensities (Amerigian, 1974; Chave and Denham. 1979; Kent, 1982). In deep-sea sediments, the principal climatically controlled variable which influences NRM intensity is carbonate content. which dilutes the magnetic mineral flu (Kent. 1982). Detailed studies of the magnetic mincralogy of marine sediment Cores (Robinson, 1986: Bloemendal et al., 1988: Doh et al., 1988) have shown that the climatically induced carbonate dilution 01magnelic mineral concentrations is not the only paleoclimatic process influencing the magnetic properties and that changes in eulian, biogenic, and ice-rafted fluxes of magnetic minerals can be deduced from changes in magnetic properties. Susceptibility rccords a t ODP sites off west Africa and in the Arabian Sca are related to terrigeneous (eolian) flux (Bloemendal and deMenocal, 1989). Spectral analysis of these Late Plioccnc/PTeistocent:susceptibility re' cords reveal 100 ky, 41 ky, 23 ky, and 19 ky periodicities (Fig. 14.1). Thcre is an apparent shift in power at about 2.4 Ma, coincident with the onset of NorthernHemisphereGlaciation. Priortothistime, the23 ky and 19 ky (prccession)pcriodicities dominate, and afterthis timc the 41 ky (obliquity) periodicity dominates. The prescncc of the 41 ky periodicity during thc Maluyama Chron is also afeature of the North Atlantic SIROrecord (Raymo et aL, 1989). The shift in powcr in thc orf-Africa susceptibility records implies a change in monsoon intensity andlor eolian source area climate related to the onset of Northcrn Hemisphere Glaciation (Bloemendal and dcMenocal. 1989). A similar change in power a t this timc (-2.4 Ma] has been noted in foraminifera] abundance data from the Mediterranean (Sprovicri, 1992).

41

19

Penodlcnjes (Ky)

$7

Figure 14.1 Susc~~tibility recuds from the central Allalltic (Site 661) and Arilhian (Sire 721) martr 111Alrica. Speclral analysis ove~U.4My intervals indicates poacr in the M~lankovitch prriodicilier (afler Bloemendal and deMcnocal. 1989).

In the King's Trough Flank area of the North Atlantic (41-43"N). peaks in ~Usceptihility(k) tend to coincide with glacialintervals (Fig. 14.2) as indicated by the oxygen isotope record, cold watcr foraminifera1 assemblages. and lows in calcium carbonate concentrations (Robinson, 198% Rohinson rt al., 1995). In core BOFS-5K, locatcd further north at 50.6'N (Fig. 14.2), the susceptibility record is dominated by the ice-rafted detrilus (IRD) associated with Heinrich events, as observed in other piston cores from this region (Groussct el aL, 1993). Robinson ef al. (1995) have used Ihc ratio of snsceptibiIity a t the Last Glacial Maximum (LGM) (-1819 ka) to susceptibility in thc Holocene as a mcans of mapping t h e distrihu-

14 3 Rock Mametic Stratlpraphy m Mannc Sediments

259

tion of IRD during the LGM (Flg. 14.3). The ice sheet instability which leads to high-susceptibility Heinrich (IRD) events m the North Atlantic (Broecker et al., 1992) has a d ~ l e r e n manifestation t in thc Labrador Sea. In t h ~ sregion, detrital carbonate (DC) and low detrital carbonate (LDC) layers, some of which correlalc to Heinrich events and other IRD layers in the North Atlantic, have been associated with turb~ditlcactlvlty in the North AtlanticMid-Ocean Channel (NAMOC) (Stoner et al., 1996). Some

5 a a'" z 2

Ez 51 * sR

:L 0

g2

-3 a=. h

g8 F

e4

"M

53

Figure 14.3 Ratios of rnapnctic wsceptibilily at the 1.ast Glacial Manimum (LCM. -1819 ka) to rusceptibility in thc Holocene. used to map trrrigeneous ice-raftcd detritus during the LGM (aftcr Robinzon el ill, 111951.

260

14 Rock Mag~~etic Straiigraphy and Palcointensilies

D C and LDC events arc characterized hy high susceptibility (due to their association with IRD) although the detrital events are more easily recog. nized hy fernmagnetic grain size-sensitive ratios such as k,,lk and SIRMI k (Fig. 14.4). Such records provide a link between (Laurentido) ice sheet variability and the marine record which arc critical to the understanding of climate change during the last glacial cycle. Parameters which are not strictly concentration dependcnl and thercfore indepcndcnt of calcium carbonate content; such as the Sratio and the

14.3 Rock Magnrlic Slriltigraphy in Marine Sedimcnla

261

ARMIk and ARWSIRM ratios, show variations which rellcct not only the principal glacial/interglacial variations but also additional fine-scale variations. particularly within interglacials (Fig. 14.5). Robinson (1986) interpreted these datain terms of three magnetic mineral fluxes: (1) antifcrromagnctic mineral flux mainly of eolian origin associated with vigorous atmospheric circulation and aridity in glacial periods: (2) coarse ferrimagnetic mineral flux associated with ice rafting in glacial periods; and (3) a constant flux of fine-grained (single-domain) magnetite of eolian or biogenic origin, thc effect of which is noticcable when fluxes ( I ) and (2) are low in intcrglacial periods. Changes in k and k,,,,,/k at Termination I and I1 for a core located off the southern tip of Greenland :trc interprcted in terms of progressive deglaciation as the continental ice shcct melted hack to ihc shoreline and then retreated into thc continental interior (Stoncr rt nl., 1995a). At Termination I in this core, k,,,lk perturbations record the Younger Dryas (YD), Heinrich Laycr I (HJ. and an additional melt event (ME1) prior to a broad

Depth (cm) Figure 14.4 Tn Labradm Sea pirlon unrc P-094. the grain slLe-srnritivc parameters (k,,lk and SIRbiik) arc morc sensitive than suaieptihility (k) for rrecognition <,I dctriial carbonate IDC) and low dc~ritalcnrbonatc (LUC) layers taller Sronct. rr 01.. 1995a).

Figure 14.5 Magnetic pilrametcr; (%ratio. ARMIk. and ARMISIRM) lor a pistnn core (:SX. 79~4)from the ~ i rough ~ region ~ af .the Xorth ~ Ailanlic (aRcr Robinson. 1986).

262

14 Rock Magncts Slratlgraphy and

Paleoinlcn~~ties

low in the k-lk record {Fig. 14.6). ME, is interpreted to signify detrital influx a t the ttme of initial detachment oi grounded sea ice from the shelf. The subsequent broad low in k,,,,lk is interpreted as the detrital influx dunng continental deglaciation on Greenland (Stoner et al., 1995a). Thts

0 ~ , , , , . , , ., , ,

0

100

, 200

, , "

,,

300

4W

500

Depth (cm) Figure 14.4 The grain rile-rensstive pmamctcr (kv,ik) as a ntonitor of continental derived rnelrwater detritus during [he Ins1 deglacialion in piston corc P-013 from off SW Greenland (aftcr Stoner PI a l , 1YY5a). The timing of the drcrrase in k,,,,/k (incrcasc in ferrimagnetic grain size) coincides with the retreat of antinenlalice lrcrm the mwtline into thc continental inlnior ol Greenland.

,

1 4 3 Rock Magnct~cSuat~graphyjn Manne

Sednments

263

interpretation is supported by radiocarbon age control tying the Iow in k,,,lk to geomorphological evidence for the timing of continental ice retreat on Greenland (Funder, 1989). Bloemendal eF al. (1988) used seven rock magnetic parameters as a means of detailed correlation in Quaternary corcs from the eastern central Atlantic. Rock magnetic parameters correlate to oxygen isotope showing the strongest correlation records, with HIRM and k,,/SIRM with .the oxygen isotope curve. Magnetic parameters are computed both as magnetic accumulation ratcs (computed by dividing the parameter by the estimafed time taken lor each individual sample to accumulatc; Bloemcndal, 1983) and as accumulation rates of the noncarhonate content. This allows the variations in magnetic properties independent of total sedimentntion rate and carbonate sedimentation rate to be nsscsscd. An antiferromagnctic and coarse ferrimagnetic dctrital flux is dominant in glacial intervals, and a fine-graincd ferrimagnetic flux is largely cunfincd to warm interglacial events. The principal indicator for this latter component (k,,,) shows a strong 21-23 My precessional variance in one of the corcs. Bivariate scatter plots of k , , against k and S !,.; against k,,,/k have been used as a means of discriminating the magnetic mineralogy of late Neogenc and Pleistocene sediments from different sedimentological environments (Bloemendal er ul., 1992). The k,, against k plot is largely controlled by the grain size spectrum of ferrimagnetic grains (magnetite and maghemile), and the S o.3 against k,,/k plot by the ratio o l antiferromagnetic (hematite) to ferrimagnetic grains. Differing sedimentary environments tend to rcveal characteristic distributions in these bivariate plots. For example, colian detritus is rich in antzerromagnetic grains. ice-rafted detritus is poorly sorted and rich in coarsc ferrimagnctic material, and current-controlled deposition tends to bc relatively well sorted. Doh et at. (1988) sludicd a 25-m core (LL44-GPCZ) from the central NorthPacific whichcovers theentire Cenozoic.The lithology is theunfossiliferous pelagic clay which is characteristic of thc Pacificmidlatitudes (Davics and Gonlinc, 1976). The stratigraphic control in this facies is particularly poor due not only to the almost total lack of microfossils hut also to ihc unstable naturc of the rernanent magnetization. The magnetic parameters show coherent fluctuations downcore which arc potentially useful for correlation. In addition, fluctuations in concentra6on-dependent parameters (such as HIRM) plotted as accumulation rates correlate well with the sedimentologically determined total eolian accumulation rate (Fig. 14.7), indicating that these parameters can he used as a proxy indicator for colian

264

14 Rock Magnatlc Shaligraphy and Palenlntcnsltle:,

Eolian Detritus [mg cm'P(l0lyrr'l

14 4 Rock Magnetrc Stratigraphy in Luess Oeposw

265

1974; H. P. Johnson et al., 1975) or alternatively the growth of authigenic ferrornanganese phases during diagcnesis (Henshaw and Merrill, 1980). The critical observationis whether thc boundary bctween the stable and the unstable remanence record, which is controlled by the magnetic mineralogy, occurs at a consistent time horizon or stratigraphic level. Available data indicate that it tcnds to occur very close to the GaussiMatuyama boundary (Opdyke and Fostcr, 1970; Prince er al., 19RO; Yamadaki and Katsura, 1994 although synchroneity has not been well established. The change in magnetic mineralogy in thc late Pliocenc may be either directly or indirectly due to change in colian flux, which in turn may be related to the onset of Northern Hemisphere Glaciation. If so, Ulcsc mineralogical changes represent time lines suitable for regional stratigraphic correlation.

HlRM (xlOJAlm.y.)

Paleocene

14.4 Rock Magnetic Stratigraphy in Loess Deposits

70

Figure 14.7 HlRM n proxy fur n,lian drlrlldl flux in rcd clays from the equatorla1 Psitlic (nfter Doh ur 01, 19881

flux (Doh era/., 1988). Changes in eolian flux arc related to changes in paleogeography a n d paleoclimate. as the Pacific plate moved northward. and can provide the basis for local stratigraphic correlation; howevcr. the precision of the correlation technique depends critically on whether thc magnetic properties are due to changes in detritus or to later diagnctic growthlalteration of magnetic minerals. In tore LL44-GPC3, polarity intervals (stable magnetizations) are recorded above 4.33 m (late Pliocenei Pleistocene) and below 19 m (Paleoceneiearly Eocene) (Prince era/., 1980). The intervening interval is characterized by unstable. highly viscous rcmancncc. 'fhc intervals of stable remanent magnclization coincide with low values oIARMik, indicating that increases in grain size oCmagneticminerals may be the cause of remanencc instability (Doh er al., 1988). In thesc intervals. magnetic grain s i x (from ARMlk) and eolian grain size (from grain observation) are high and vary consistently, suggesting that most of the magnetic mincrals in these intervals are of eolian origin. This is not the case for the interval of unstable remanence, where eolian grain sue fluctuations are not recorded in the ARMIk profile? suggesting diagenetic altcralion of magnetic minerals in this interval. Indeed. the instability of the remanent magnetization in this particular facies has bccn attributed to oxidation of primary magnetite to maghcmite downcure (Kent and Lowrie,

Hellcr and Liu (1982) pointed out that the susceptibility variations in the Chinese loess match the paleosoliluess lithologic variability. and Hcller and Liu (1984) matched the record to the Brunhes oxygen isotope record from the Pacific of Shackleton and Opdyke (1976). Kukla era/. (1988) showed that thc susceplibilityrecord closely parallels the orbitally tuned SPECMAP oxygen isotopc record derived from deep-sea sediments (Fig. 14.8),indicating a close linkage between eolian flux, global ice volume, and climate. Rock magnetic data have played an important role in understanding the origin of the climatic record in Chinese loess. The present consensus is that the higher susccptibility of the paleosols in the Chinese locss is the rcsult of "magnetic enhancement" associated with authigenic production of SP and SD magnetite duringpedogenesis (Zhou et aL, 1990; Heller et a/., 1991; Maher and Thompson., 1991,1992; Banerjee er at.. 1993: Evans and Heller, 1994). Beer etal. (1993) showed that the pedogenic and detrital susceptibility components in the loesscan be quantitatively distinguished by comparing the susceptibility variations with "Be concentrations. According to these calculations, the susccptibility in Lhc paleosols due to magnetitc formed in situ is often -50% greater than the detrital susceptibility. Taking prcsentday precipitali~mlsusceptibilitydata (Liu el al., 1992) as the means of calibration. Heller el al. (1993) used estimates of lhc pedogenic susceptibilily contribution To estimate paleoprecipitation. Maher et 01. (1994) constructed apalcoprecipilationrecord based on bulk susceptibility differences between paleosols and intervening unweathered loess and a logarithmic relationship between rainfall and susceplihility in nine modern paleosols across the Chinese loess plateau.

266

14 Rock Mapctle Stratigraphy and Paleomtms~t~r~

ODP 677

Xiteng

Figure 14.8 Correlation ofsusccptibility records from Chinese loess wlrh the benthic omen isrllupr r e e d irim ODPSile 677 (aitcr Heller and Evans. 1995). Shading in lithology indiwtzs

paleosols separated hy k,ras.

The pedogedicmodcl which explains the magneticvariability in Chinese locss cnnnot be applied to all locss deposits. The susceptibility record in Alaskan locss has orbital variability, but the intervals of high susceptibility correlate to cold climate intervals, opposite to the relationship in the Chinese loess (Beget and Hawkis, 1989). A close relationship between mean' loess grain size and susccptibility values implied that the high-susceptibility inlervals can be attributed togreater windvelocities during glacial intervals. which transported Iargcr rnagnclitc grains which dominate the suscepfibility record (Beget et al., 1990).

14.5 Rock Magnetic Stratigraphy in Lake Sediments Magnetlc parametcrs prov~dea baas for correlat~onand for determ~ning changes of magnetic mineral fluxes in both marlne and lake sediments. However, because lake sedimentatton rates arc typically 2 or 3 orders of

14 5 Rock Ma@ncl~c Slral~graphym Lake Sedmcnts

267

magnitude greater than those of deep-sea sediments, the time scale of Lhc observations is very different. Whereas magnetic mineral flux changes in the marine setting can give us information on long-term climate change. the samc information fromlake sediments is generally giving US information on Holuccnc and recent anthropogenic environmental change. There are. however, lake sediment corcs which extend back through thc last glacial cycle such as those from Lac du Bouchet (Thouveny et al., 1994) and Lake Baikal.(Peck et al., 1994). In thesc studies the susccptihility record is an important means oC monitoring lithologic changes associated with the .,glacial-interglacial cycling. Volume susceptibility measurements are a very important. rapid, and nondestructive means of core correlation in lake sediments (Thompson et al., 1975; Blocmendal et al., 1979: Peck et al., 1994). Distinctive features in the records, which are controlled by variations in concentrafions of ferrimagnetic minerals. can be matched to monitor variations in sediment accumulation. The method can be useful cvcn in lakes which show wide variations in sedimentation rate (Thompson and Morton, 1979). Other magnetic paramcters, such as SIRM, can also be uscd for core correlation and may b e useIul when susceptibility values are uniformly low, but thcy may require subsampling of the whole-core and therelore measurements may bc considerably slowcr than whole-core susceptibility mcasuremetlts. The usc of magnetic susceptibility profiles for correlation in lake sediments led to the observation. at Loch Neagh, that progressively increasing susceptibility values parallel pollen variations signifying progressive dcforestation (Fig. 14.9) (O'Sullivan era/., 1972). Thompson et ul. (1975) interprcted these data to indicate increased erosion and a resulling flux of ferrimagnetic grains from the basaltic bedrock. It has since been demonstrated that the susccptibility of topsoil can be greater Lhan that of the deeper soil horizons and of the bedrock. This "magnetic enhanccment" has been attributed to the lormation of magnetite and maghemite in the surface soil (iizdemir and Bancrjee, 1982: Mullins. 1977; Maher and Taylor, 1988: Evans and Heller. 1994; Eyre and Shaw, 3994). Increases in susceptibility associated with pollen evidencc for deforestation would now be interpreted to reflect increased erosion of topsoil as a result oT changes in land usc (e.g.. Ilearing et nl.. 19861 Thompson and Oldfield, 1986). It has been demonstrated that susceptibility of lake sediicnts can be a very sensitive indicator not only of changes in erosion style in the catchment but also of changes in climate (Thompson and Oldfield, 1986). changes in stream discharge (Dcaring and Flower. 1982), natural burning or burning for land clearance (Rummcsy et dl., 1979 Rummery, 1983), and industrial burning oIfossil fuels (Oldfield era/., 1983). Burningcontributes to the lake sediment susceptibility record because magnetite is an important product of both soil burning (Longworth er ul.,1979) and industrial burning (Chaddha and

268

14 6 Future Prospects tor Rock Magnetic Strattgraphy

14 Rock Magnehc Stratigraphy and Paleoinlcns~t~e~

-

269

1

application, t h ~ r is e a need to establish a more formal, quunlitative approach to the study of magnetic mixtures, in ordcr to delineate the proportion of individual magneLic constituents. Thompson (1986) has used the Simplex search method to deterrninc optimal mixingmodels on the basis ofparticular magnetic characteristics (such as IRM acquisition curves) oI individual sediment samples and of possible source materials. Alternatively. the magnetic characteristics of the scdiment uafl be matched to combinations or standard curves based on synthetic samples. /

14.6 Future Prospects for Rock Magnetic Stratigraphy

Pollen Zones

u

2

v,

Figun 14.9 6NO-year record from Lough Neagh 1Northern Ireland) Forest clearance and iamrngare~n&catedbylncreasesin plantams,%rarses, and cerealpollcn lncrcascs msusccptlbllltv are attnbnted to Increased soti eroslon m thc lakc carchrncnt (aftcr Thompson nnd Oldfield, 1986)

Scchra. 1983). As many of these anthropogenic processes produce changes inferrimagneticinffux, concentration-dependent parameters such as susceptibility and SIRM are most applicable (rather than mineralogy-dependent magnetic parameters). However, for detailed studies of erosion history and land use changes, a more complete characterization of sediment source material .and of the sediment itself is necessary. Magnetic parameters can be used as tracers to determine changes in sediment sourcc and thcrcby provide markers of environmental and land use change. In this type of

Until recently. rock magnetic stratigraphywas confined lo the use of susccptibility as a convenienl method of core correlation in volcanogenic marine sediments. During thc past ten years, rock magnetic studies of lakc scdiments havc demonstrated that rock magnetic parameters are not only useful for core correlation but can provide delailcd information on changes in magneticmineral flux that can bc related to change in land use (Thompson and Oldficld. 1986). The techniquesused in this developingfield of environmental magnetism are now being applicd to the marine and terrestrial sedimentary record. Susceptibility data from Chincse loess deposits (Kukla et al., 1988) and from thc coliatl-dominated equatorial sites off Africa (Bloemendal and deMenocal, 1989) have variance in which the dominant "heartbeat" is very similar to that sccn in marine oxygen isotope records, providing a fascinating link between marine and terrestrial climate records. Rock magnetic methods are gcnerally conducted on subsamplcs from sediment corcs often collected as 8-cm' plastic cubes. Thc widely used 2G755R cryogenic magnetometer (with 3.2-cm-diameter access) is designed [or discrete sample measurement with homogeneous response of SQUID scnsing coils over a volume large compared to the discrete samplc volume. Weeks ef (11. (1993) desnibc a modified 2G-755R magnetomctcr in which the SQUID sensing coils are arranged for optimal spatial resolution of magnetization rrom continuous "u-channels" tracked through the sensing region. Deconvolution of the signal in the time domain (Constable and Parker, 1991; Oda and Shibuya, 1996) gives resolution comparable to that achieved by back-to-hack discrete sampling. The u-channels (sec Tauxe er ul., 1983h Nagy and Valet. 1993) are typically 1.5 m in length with a 2 X 2 cm cross-section. Stepwise demagnctization of NRM, acquisition and stepwisc demagnetization of IRM and ARM can be carricd out on a single automated measurement track (Weeks et al., 1993). Rock magnetic measuremcnt tracks with thjs capsbilitg will greatly i~lcreas,cthespeed a t which

272

14 Rock Mugnet~cS t r a t ~ ~ r a p hand u Palcomten~~l~cs

10

LD

30

40

50

60

70

I

80

Time ( k a ) Figure 14.1 1 Comparison ol srdirnmlary relalive ppaleoinrcnsity rccords from Mediterranean Sea (continuous shaded record) wilh palnlinlenrily dnta from volcanic rocks and archzomag~ nctic artefacts (points with error b a n ) (aRer Tric ui nL, 19921.

Deep-sea sediment cores hold considerable promise for giving long continuous records of relative geomagnetic paleointensities. The method relics nn thc proposition that the intensity of DRM is a linear function of ambient field ovcr the range of geomagnetic field intensity (Kcn~,1q73; Barton er a/., 1980: Tucker, 1980) and that variations in concentrations of magnetic minerals in the sediment can hc cumpensated by a suitable normalization factor (see review by Tauxe. 1993). Thc ideal sediment for paleointensity determinations is one in which the remancncc is a D M carried by PSD magnetite in the 1-15 p m grain size range and in which the concentration of magnetite does not vary by more than a factor of 20 or 30 (Kinget a/.. 1983). In most studies. the paleointcnsity record is derived from the NRM intensity (after partial A F demagnetization Lo eliminate VRM) normalized lor varialiuns in magnetite concentration by dividing by ARM, k, SIRM, low-field IRM, or SLKM (stirred remanencefrom redeposition experiments) (Kent and Opdyke. 1977: Tauxc, 1993). Mure than one normalization factor can be used, strengthening thc record in the case ol concordancc. ARM is often the normalization factor of choice as it is particularly sensitive lo SDIPSD magnetite grain sizes, whereas SIRM and k are more sensitive to MD magnclile grains. Meynadier er a/. (1992) demonstrated a similarity in coercivity of thc NRM and ARM in three piston cores from the Somali Basin, thereby showing that thc grain size fraction of magnclite carrying the ARM is similar to that carrying the NRM. For these 5 cmlky scdimentation rate cores, the NRM,,,,/ARM

ratio was used as the measure of relative paleointensity in the 20-140 ka interval. The Somali Basin records arc consistent among the three cores and compare well with the 80 ky record of Tric et ul. (1992) from lhc Mediterranean Sea (Fig. 14.11) based on stacked NRMm,,rlk data from four cores with sedimentation rates of about 10 cmlky. These records can be comparcd with variation seen in Thellier archaeomagnetic and volcanic records for the last 40 ky (c.g., McElhinny and Senanayake, 1982), with relative paleointensities for the last 20 ky from box cores (Constable and Tauxe. 1987), and with longer records from Ontong-Java (Tauxe and Wu, -- 1990; Tauxe and Shackleton, 1994) Sulu Sca (Schneider, 1993). western equatorial Pacific (Yamazaki and loka, 1994), central North Atlanlic (Weeks et ul.. 1995; Lehman et at., 1996), and Labrador Sea (Stoner el a/., 1995b). The correlatihn of the marine paleointensity record? to a 84 ky record from Lakc Baikal (Peck er al., 1996) is quite straight fonrrard, but correlation to Lau dn Bouchet (Thouvcny et al., 1990) is less clear. Comparison of the Labrador Sea record with a composite record compiled from the Mediterranean and Somali Basin records (Fig. 14.12) leads lo two conclusions. (1) Relativc paleointensity records lor the last 100 ky' can be correlated over large distances, implying that marine sediments can record paleointensities of the gcomagnetic dipole field. (2) The AMS 14C age control for the Labrador Sea record is inconsistent with the oxygen isotope age control from the McditerraneanlSomali basins. Paleointensity studies in marine sediments, which hitherto had been restricted to the last -200 ky, have been greatly expanded by an important study of ODP Leg 138 cores from the equatorial Pacific (Valet and Meynadier, 1993). The record covers the last 3.9 My in corcs with mean sedimentation rates not exceeding 2.5 cmlky (Fig. 14.13). Relative palcoin: tcnsities were converted into virtual axial dipole moments (VADMs) by calibration to Thellier determinations of paleointensitics for the last 0.5 My. Several important featurcs are apparent in the records: (I) paleointcnsity lows in the Brunhes Chron appear to correlate in agc to geomagnetic duectional excursions; (2) a sawtooth pattern of paleointensity fluctuations appears to be related to rcversal boundaries: reversals tend to coincidc with lows in the VADMs with abrupt increases immediately post rcversal. The longer the duration of thc polarity chron, the greater the abrupt increase in VADM at its onset; and each polarity chron appears t o be characterized by decay of VADM leading t o the next reversal. This has led to the suggestion that the reversal frequency, and the occurrence of polarity excursions, is ultimately controlled by the intensity of the dipole field, which also controls Lhc stability of a particular polarity state. The correlation of widely spaced sedimentary paleoinlcnsity records for the last 140 ky suppers the contention that they are reliable records

:

( ~ 6 6 'IarP l -vrv(qq pue lqe* rai~e)ma30 3gr3e~poolenha aql u~ ( I = 211s pue 8 ~ airs) 8 get 821 dao u o q proaar (wav~ 'lwamour alodrp T P L Xleninh) ~ &~sua~u~ua[ed a1rso&wv3 CL'QL an6y

VADM (X 10% in2) '(4~661'70 P lauols ram) .q*awuqi E3S ropwqrrl aql no panold p w n!l lanqh 3qsn qaais eaS ~ o p z ~ rqu upaie[om= ~ y a r ~ si~!snalu!oapd O I (2) ~ . a [ w s a u ! ~OIW 341 uo pallold puv a!\ lens!& Bu!sn yawls OIW oi p a i e l a ~ ~ o93es r Li!srralu!oalsd EJS loperqe? (q) .L8o[ouolq3nlqdsr pue 0,,9 DO paseq saae qsm ( ~ 6 6 1,.," ra '70 JJ Ja!peu.aJ oeg q i q w l d e pun nalep J,, swv moq dlmeru pan!lap lapnu 2% qi!m (JET paqsep) y a e Li!suaiu!oa[ndan!1~~31 ~ ~ 3 ~so p e ~ q ea? u (e) ~ [ ' Q Ia n 6 3

(nq) sa!Zolouoq:, @npm!puI 4

-

0

.

011 .

OL

Ilh

OE

?>enueaao ue!puI neaunr?l!pow-

?>PIS "2s

~ o p e ~ q......... el

PLE

276

14 Rock Magncrle S~rattpmphyand Paleotntenvtacs

Bibliography

Figure 14.14 Comparison of the rela~ivepdlrrtintensity stack (Sinf-2W. Guyodo and Valet, 1996) with the P-01.7 palr
of theVADM. The presence in the paleointcnsity rccord of orbiral periodicities (see Meynadier er al., 1992) suggests. however, that climatc-controllml litholngicvariahility may he influencing the records. Tauxe and Wu (1990) advocate use of a coherence function to test whethcr paleuintensity records are coherent (a1 lh; 95% confidence level) with the normalizing lactor used, and with other magnetic paramctcrs (such as ARM or k) which may hc affected by climate-controlled lithologic cydcs. Recent paleointensity studics Irom the Allanlic and Indian Oceans for the Jaramillo Lo early Brunhes interval (Valet el nL, 1994) appear to confirm the general structure of thc Leg 138 record (Valet and Meynadicr. 1993). indicating that paleointensity records will not only provide us with significant insights into the nature of the geomagnetic ficld and the mechanism for polarity reversals and polarity excursions but also provide us with a new means of high-resolution stratigraphic correlation. A n example of such a correlation is depictcd in Figurc 14.14. The relative palcointensity stack compiled by Guyodo and Valet (1996). based on paleointensity records that can he directly correlaled lo the oxygen isotopc chronology. is correlated to the paleointensity record from the Labrador Sea (Stoner r l nl., 1995h) whcrc oxygcn isolope data are compromised by lackof benlhos and meltwater influence on the planktic record. The correlation in Fig. 14.14 allows the oxygen isotopc chronology to be imporled into the Labrador Sea, heralding a new era of high resolutiotl magnetic stratigraphy based on gcomagnetic paleointensity.

eustasy. Earth Plonel. Sci. Len. 97, 102-112. A1geo.T. J . (1996) Geomagnetic pularily bias patfcrnr thruugh ihe Phanerozoic. I . Crophys. Res. 101,2785-2814. Alvarer. L. W.. Alvarer, W.. Araru, P., and Michcl. H. V. (1480). Extratcrrcrtrial u u w lor the Ctrlaceous-Tertiary cntinaion. Science 208, 1095-1108. Alvarc~.W.. and Lowric, W. (1978). Uppcr C~ctacrolripalcomagnctic stratigraphy at Mona (Urnhrian Apcnninrs, Italy), verificarion of the Cuhbio section. Crophy.~.J R. Astrun. Soc 55, 1-17, Alvarrz, W., and I.owrir. %'. (1984). Magnetic stratigraphy applied 10 rynsedirnentary rlumps, turbiditics and basin analysis: 7hr Scaglia linlesronc at Furlu (Italy). Geol Snr. A m Bull. 95, 321-336. Aluarcz. W.. Artlrur. M. A.. Fischrr. A . G.. Lowric. W.! Napdeone, ti.,Yrcmoli-Silva, I.. and Roggenthen, W. M. (1977). Upper Crctacaous-Paleocene magnctic stratigraphy at Guhhio, Italy. V. Type sealioh for the Latc Crrlilceous-Paleoccnc gcomapnelic reversal time scale. CcnL .TI,<;. Am. BUN 88,383-389. Amedgian, C.(1974). Sra-floor dyrlamicprocrssesils the possible causcoim~rrrlationsb e t w c n paleoclimntic and pnleon~agncticindicra in deepsea sedimentary cores. Earth Pia~lel. Sc;. Lett. 21, 121-326. Amcrigian, C.(1977). licasurcmrnt 01 the effect of partidc size variation On thc DRM Ir, ARM n t i o in somc abyssal sediments. Earth Planet. Sci. Lwr. 36,434-442. An, Z., and Ho. C . K. (1989). New nlagnetostrarigraphic dates of Lantian Homo trsctus. Q. Rcs. 32. 21 3-221. An. 2..Bow1cr.J. M., Opdyke,N. D.,Macumhzr,P. G..andFirman. J. R. (19R6j.Paleomagnctic stratigraphy uf Lake Bungunnia: Plio-Pleistocene prccurrr,r "I anidity in thc Murry Basin. xoutheaster~lAustralia. Polorr~grogr..Pnloeoclimrrt~,l.,Palueo~col.54 219-239. An.Z.. I.iu. T. S.. Kan, X.. Sun. J., Wac~g.I., Kaowanyi, Z. Y.. and Wei. M. (1987). Lorsspalrosol sequcnccs and cllronology at Laurian Mau localities. In '.AsprcLs of Loess Researell" (I'. S. Liu, e d ) , pp. 192-203. China Ocean Prcss, Bcijine. Andrziefi. P.,Rrll<m.11.. ahd Wcstcrcamp. D. (1976). Chronomrlric rt,wiitigaphie compar&r dcs cdificrr volcaniques ct fcirmalions redilllentaircs de la Martinique (Antillus franqaiscs). Bell. Bur Rech. Cuol .Win. (Fr.) 4, 334-346. Anonymous (1979). Mapenclostratigrapllic polarity units. A supplcmcntar? chapter Of thc International Subcummission on Stratigraphic Clasfificatian, Inlrrnulional Stratiz~aphic Guide. Crology 7, 578-583.

Berger, W. H.. Vincent, t..., md Thientein. 11. R. (1981). The drepsca record: Major strps. In -Tenmrric Ocean f-r,olutioa.:'SEPM Socisly fur Sedimenlary Grok,gy Spec. Puhl., Nu. 37, pp. 489-504 Bcrgercn. W. A,, Burckle, L. H., Cita. M. B., Conkr, H. R. 8.. Funnell, B. M.; Garmer, s.. Hays. J.D.,Kennett.J.P., 0pdyke.K. n..Past<~urct.L.,Shackleton. N.J.,andTakayanagi, Y. (1980). Tuwards a Quatcmary time scale. pmzr Res. 13,277-302. Bcrggrcn, W. A.. Aubry, M. P.. and Hamilton. N. (I983a). Neogcnc magnetohioslratibraphy of deep5ealirjllingprc~~r~%ite516(RioGrandc Rise. South Atlantic). I n "Initial Repons of thc Dccp Sca Drilling Pro.ject" (P. F. Barker. R. L. Carlsun, D. A. Johnson ?I OI., eds.). Vol. 72. pp. 675-70. U.S. Govt Printing Officc. Washington. DC. Bcrggrcn. W. A.;Hamiltun. N.. Johnson. D.A., Pujol, C.. Weiss, W., Cepek.P., and tiomhrrs, A. M.. Jr. (lY83b). Magnstobiostratipaphy 01 deep sea drilling project l e t 172. rite! 515-518. Rio Grandr risc (South Atlantic). In ..Initial Reports of the Deep Sea Drilling Projcct" (P. b. Barker, R. L. Carlson. D. A. Jahnroll er ul., eds.). Vol. 72, pp. 939-948, U.S. Govt. Printinp Oiiite. WashingLon; D T Berggrm, W. A , Kmt. D. V.. and Flynn, J. J. (1985a). Jurassic to Paleogene. Part 2. Palrogcnc yeochronolog aid chranostratigraphy. In "The Chronolosy of the Geological Rccord" (N. J. Snelling. ed.), pp. 141-195. Blackwell. Oxfixrd. Berggrcn. W. A., Kent, D. V , a n d van Couvcring, J. A. (1985h).'l'he Neogene. Part 2: Neogrnc geochronology and ChronMtratigraphy. Irr .'meChronology of the Geological Record.' (N. J. Snelling, ed.), pp. 211-260. Blncku,eIl, Oxford. Bergten. W. A., Kent. D. V.. Ohradovich. J. D., and Swlsher. C. C . ( 1 ~ ~ 2Toqard ). n revised Paleogema gcochronology. In "Eocene-Oligocene Climatic and Biodc Evolution'. (D. K.Prothero and W. A. Berggrrn. rda.), pp. 29-45, Princeton Univ. Prcss. Princeton, NJ. Bcrggren. W. A.. Kent, D. V., Su,isher, C. C.. and Aubry. M. P. 11995). A revised Cenoloic

-

-.

-,.

and Stratigraphic Correlation," (W. A. Bcrggrcn. D. V. Kent, M. Aubry. and J. Harden~ bol, cds.), SEPM Spec. Puhl Nu. 54, pp. 129-212. Beske-Diehl, S.. and Shivc. P. N. (1978). The rock mapelism uf the Mississippian Madison Limeslone nurth-central Wyoming. GeoPhy.~. 1. R. Arlron Soc. 55, 351-362. Biw. D. A , and Montuari. A. S . (19R8). Magnctic stratieraphv of the Masrirnano section . . across the Eocene~olieocenrhr~undary.Inr Sehcomm. Prrlu,,genr Strrrrigr, E/O M m . , Ancona. Spec. Puhl. 11: Vol. 4, pp. 111-117. Bingham. C. (1974). A n antipodally symmcnic diatnhution on a sphere. Ann. Srar 2.12Ol 122i. -. Rlackctr, P. M. S. (1952). A nrgalivr experiment rclating magnetism and the Earth's mtation. Pt~iios.Trunr R. Sor. I.ondon. .Ser. A 245, 309-370, Blackell, P. M. S. (1956). "LcctureS on Rock Magnetism." Weizmiln Scicncc Press. Jeruralen~. Blakcly, K.L. (194). tieolnagnetic reversals and ~Tustalspreading rarcs during the Miocene. J, Crophys. R a . 79, 299-2985. Blakemore, R. P. (197.5). Ma~ncticbacteria. Scirnre190,377-379. Blakrmore, R. P.. Short, K. A,. Bazylinski, D. A,.Rosenblatl. C., and Frankel, R. B. (1985). Microaeroblc conditions are required fnt magnetite formation within Aqumpirillunr n2~rgnelolurrirum.Geo,nicmhiol. J. 4 53-71, Bleil, IJ. (1985). Thc magnetostratigraphy of northwest Pacific sediments. Deep St.* Drilling Projcct Lcg 86. In "Initial Reports of the Deep Sea Drilling Prrljrd" (G.R. Hcath, L.H.B~rkleel~L, eds.). Vol.86,pp. 441-4SR.IJ.S. Gout. P~intingOficc,Washington. DC.

Blcil. LJ. (1989). Magnctostratigraphy of Ncogene and Quaremar). ~ r d i m nseries t from thc Norwegian Sea: Ocean D~illinpProgram Leg 10.1, 1989. In "F'rocecdinp of the Occan Drilling Program,Scientific Raults" ( 0 . Eldholm. I. Theide, E. Taylrlr el oL. cds.), Vol. 10.1, pp. 829-850. ODP. Crlllrge Station, TX. Bloemendal. J. (1983). Paleocnuironmental implications of the magnctic charactrristics of wdimcnts fromDSDP Sitc 514. Sauthcasl Argentine Basin. In "lnitial Reports of lhr Deep Sea Drilling Project" ( W . J. Ludwig, V. A. Krilsheninniktrv el a/., ede.). Vol. 71. pp. 1097-1108, U.S. Govt. Printing Office. Washington. DC. Blormendal, J., and deMenocal, P. B. (1989). Evidence for n change in the periodicity of tropical climate cycles at 2.4 Myr from whale-coic magnetic susccplibility mcaruremcnts. .Vatare (Lotldnn] 342, 897-W0. Bloemendal, J.. Oldfield. F., and Thompson, R. (1Y79). Magnetic rnrasuremcnt5 used to asscss wdimcnt influx at Llyn Goddionduon. ,hru!lrmre(Lonilon) S O , 50-53. Bloemcndal, J., I.mb. B.. and King, J. W. (1988). Paleocnvirgnzphy3,61-87. Bloemendal. J.. King, J . W., Tauxc, I.., and Vnlel, J. P. (1989) Rock magnctic stratigraphy of L.cg 108 (F.astern tropical Atlantic) sites 65% 659. 661, and 665. In "Proceedings r l l the Ocean Drilling Program, ScielttificResults" (W. Ruddirnan, M. Sarthcin ri ul,, eds.), VoI. 108, pp. 415428. ODP, Cr,lle$e Stalilm. TX. Bloemendal, J., King. J . W., Hall, E. R., and Doh, S. 1. (1992). Rock mapetism of Lnle Neagene and Plcislocenc deep-sea scdimznts:Kcl,rtinnship to sediment source, diagenctic processes, and scdiment litholugy. I . Genphys. Kcs. 97, 4361-4374 BIOCmmdal, J.. King, J. W.. Hunt. A,. deMenocal. P.B., and ilayahida. A. (1993). Orign of thc sedimenrary magnctic record a1 ocean drilling program sites on the Owen Kidge. Western Arabian .%a. J. Gur>phph~.s RI,?. 98, 4199-4219. Bloxham, J. and Gubbinr. D . (1985). Thc secular variadon of the Earth'$ magnetic field. Nnrure (London) 317,777-781. Bloxham, J., end Guhbins, D. (1987). Thermal core-mantlc inleractions, hiairire (London) 325,511-513. Donhornmet, N, and Babkinr. J. (1967) Sut la prgsence d'aimentation inverse dans 1% Chainc des Puys C.R. Heb. Seon(!er Acod Sri.. Ser 8 2&1,92-94. Bonhummet, N., and Zahringer. J. (1969). Paleomaenrlism and p
284

Bibliography

Champion. D. E.,Lanphere, M. A.. and Kuntz. M A. (1988). Evidcnce for a new geomagnetic reversal from lava nowr in Idaho: Discussion of shon polarity rcvcrsnls in the Drun~les and late Maluyama polarity chrons J Geophys. Rt,s. 93,11667-11680. Chnng, S. R., andKinchuink, J. L. (1989). Magnetofossils,the magnetization of sediments, and the e\201utionof magnetite biominerali~ation.Anmi Rev. Earth Planer Sci. 17, 169-195. Chang, S. R., Allen, C. R.. and Kirchvinck. 1. L. (1987). Magnetic stratigraphy and a t e q for block rotation of sedimenrary rocks wjthin the San Andrras fault zone. Mecca f i l l s southeastern California. Quut Res. 27, 30-40. Channell. 1.E. T.. and Dobson. J. P. (1989). Magnetic stratigraphy and magnctic m i n e ~ a l o g ~ at the Cretaceous-Tertiary boundary reclir,n. Rraggs, Alabama. Polneogeogr. Pulntoc/i. marol., Pulaea~ol.69,267-277. Channell. J . E. T.. and Erba. E. (1992). Early Cretacerrus polarity chrons CMU to CMll recorded in northern llalian land srctions near Brcscia. Earth Planel. Sci. I.er!, 108, 161-179. Clannell. 1.E. T.. and Grandesso,P. (1987). A revised corrclation ofMesozoic polarity chrons and calpit~nellidxones. Earth Plonef. Sci. Len 115. 222-240. Chanuell. J . E. T.. and Hawthorne, T. (1990). Progressive disalluticm uf titanomagnctites at ODP Site 653 (Tyrrhsnian Sea). Earth Plru~rt.Sci Left. 96, 469-480. Channell. J. E . 7 ,and McCahe. C. (1994). Comparison of magnetic hysteresis pammclcrs of unrcmagnctizcd and rcmagnetized limestones. 1.GrrJphys. Re>. 99,4610-4623. Channell,]. E. T., and Mediz2a.F. (1981). Upper Cretaceous and Palcogcnc magnctic stratigralrum thc Vcnctian (Southern) A l p . Earrh Ploner. Sci. Lrrr phy and hio~trilti~raphy 55,419-432. Channell. J. E. T.. and Torii, M. (1990). Twcr rvrntr rccardcd in thc BrunhcS Chron at Hole 650A (ODP Leg 107, Tyrrhcnian Sca): Gcomagtleticphenomena'! InProceedings of rhe Occan Drilling Pmgram. Scicntitic Resultr" [K. A. Kastenr. 1. Mu& rr a/., rd~.)~VelI. 1U7, pp. 347-359 College Station, TX. Channell: I. E. T.. Lowrie. W., and Mediiza, F. (1979). Middle and Early Cretaceous magtretic slraligraphy frwn thr: CZ8mon sccrion. northert~Italy. Earth Plnne~.Scl. Len. 42, 133166. Channell. I. E. T., Freeman. R., Heller, F., and Lowic, W.(1Y82a). 'Timing of diagenetic hematite growth in rcd pelagic limestones from Gubbio (Italy). Eanh Planet. Sci. Lrri. 58,189-201. Channell, I. E. T.. Opg. J. G . . and Lowrie, W. (1982b). Gcomngnctic polaritgin thc carly Cretaceous and Jurasi.ic. Philos. Trans. R Soe.London, S t r A 306,137-146. Channrll, I. E. T., Lowric, W.. PiaUi. P.. and Venturi, F. (1984). Jurassic magnetic stratigraphy from Umbrian (Italian) land scdion. Eorth Plunrr Sci. Lrtt. 68,309-325. Cllannell. I. E. T., Bralower. T. 1.. and Frandcssn! P. (1987). Biostratigraphic corrclation of Mesw~aicpularily chrons (CM1 t o CM23) at Capriolo and Xausa (Southern Alps. Italy). Eorlh PlnneL Sci. Lptl. 85, 203-321, Channcll. J. E. 7. Rio. D., and Thunzll. R. C. (1988). Miocenemliocene boundary magnerostratigraphy at Capo Sparlivenlo. Cahhria. Italy. Guoli~gy16,1096-109. Channel. 1. E. T.. Massari, F I Rcnetti, A., and Pczzoni, K. (1990a). Magncmstratigraphy and hiostmtigraphy of Calloviam-Oxiardianlimestones from the Trcnto plateau, (Monti Lcssini, northern Italy). Pnlueopeogr,. Palueoclintalol., PulaeoecoL 79,289-303. Channell. J. E, T., Turii. 41.. and Hawthorne, T. (19YOh). Magnetostratigraphy 01 sediments recorered at %teahS0;hSl. 6 2 , and654. (Leg 107 in thr Tyrrhmiansea). In "Procrcdings 111lhr Oucan Drilling P~oeram,Scicntlfic Kcnults" (K. A. Kasfens, J . hlusclc et nl, cds.1. VI~I.107. pp. 335-346. ODP. College Station, TX.

Channcll. I. E. T.. Erba. E., and Lini, A. (1993) MagneU,stratigraphic cahbratiorr of the Tale Valanginan carbon isotope event in pelagic limestones frrjm northcm lraly and Switzrland. 6arth Planet. Sci Lert. 118, 145-166. Channell. J.E. 'l'.,Poli.M. S.. Rio. D., Spruvieri. H.,andVilla, G. (1994). Ma~nericstratigraph!' and biostrntigraphy of Pliocene "argille uzurre" (Northern Apennines. Italy). PnlaFr)geogr. Palor~,climutoI..Pnloeoecol. ll0,81-'102. Cbanncll. J. E. T.. Erba. E.. Nakani~hi.M., and l'arnaki. K. (1995~).A Latc Jurilhsic-Early Cretacrnus timescalc and rluranic magnctic anurnilly block modcls. Iw "Gauchrrmology. Timrrcales, and Stratigraphic Correlarion: (W. A. Bersgren, U. V. Krnl, M. Aubq. and I. Hardenbol, edr.), SEPM Spcc. Puhl. No. 51, pp. 51-64, Channrll. I. E. T., Cccca,F..andErba. E. (lYY5b). Correlations of Hautcrivian and Barrenllan (Early Cretaceous) stage boundaries to p o l ~ ~ i lCy~ T O I ~ SEarrh . Plnnet. Sci 1.ufl. 134, 125-140. Chavc, A . D. (1984). Lowcr Palcuame-Upper Cretaccour ma@etortratinraphy, Sitc 525.527. 528 and 529. Deep ScB Drilling Project Leg 74. In "Initial Reports of thc Dcup Sril Drilling Pnrject" (T. C. Moore. Tr., P. D. Rabinowitr rr ol., edr.). Vol. 74, pp. 525-531. U.S. T.
Clewmt, 8.M. (1992) Evidcncc for dipular Rcld~during thc Cobb Mounvain gcomagnctic polarity reversals. ~Virnrre(London) 3513, 405407. Clcment. B. M., and T+ailwood. E. A. (1991). Magnetostrahgwphy of redimcnts from sites 701 and 702. 7rl "Proceedings of lhr Occan Drilling P r t a ~ a m .Scientific Rcsults" (P. F. Cienielrki, Y. Kristofferson el nl.: eds.). Vol. 114,pp. 359-365. ODP. Cullcge Stadon, TX. Clcment. B. M., and Kent, D. V. 11984). A detviledrccard of the Lnwcr Jaramill" polarity lransiliun from asoulhem hemisphere dccp seasediment cure. J . Geoph?,~.RPS.89, IMY1058. Clement. B. M., and Kcnr. D. V. (1987). Sharr polarity interval? within the Matuyama: T~ansitionfield records from hydraulic piston cored ardimcnts from the Nrlrth Atlantic. EarlA Plonel. Sci 7.eft 91, 253-261. Clement. B. M.. and Martinson, D. F. (1992). A quantitativecomp8risun o f t a o palromagnetic rccords of the Cohh Muunrain subchran from North Atlantic deep-sss scdimcnts. J. Grr,phy. Re.?. 97, 1735-1752. Clemenl. R. M.. and Robu~son.R. (1986) Thc magnetostratigraphy of Lcg 94 sediments. I n "Initial Reports of the Drcp Sra Drilling Project" (W. Ruddiman el 01.. edr.), Vol. 94, pp. 635-650. 11.5 Govt. Printing Office, Wasington. DC. Clement, B. M., and Stirrude. L. (1995). Inner corc anisotropy. annrnalics in thr rime-averagcd paleumagnetic field. and polarity Transition paths. Eurfh Planer Sn. Lerr DO,75-85, Clemcnt. 8. M.. Kent, D. V., and Opdyke. N. D. [19821. Brunhes-hlatuyama transition in three deep sea ~edimcntcores. Phi1u.s Trons.K . Soc. London, Srr A .W6,113~119. Clement. B. M.. Hall. F. J.. and Jdrrilrd. K. U. (1989). The magnctostratigraphy al Ocran Drilling Program Leg 105 rcdimcnts. In 'Proceedings of 'The Ocean Drilling Program, Scientific Results" (S. P. Srivasta\,a, M. Arlhur. R. Clcment, eds.). Vol. 105, pp, 583 595. College Station. TX. Clement. B. M.. Rhndda. P., Smith, P..and Sicrra. L. (1995). Rrcuning transitional grrlmagnelic ficld gcametries: evidence from scdimems and lavas. timphys. Rrc Lett. 22,31713174. Clydc, W. C.. Stumatakrrs, J . . and Gingerich; P. D. (1994). Chrrmalogy of the Wasalchiiln latl&mamrnal agc (early Eocene): Magnctostratigraphic results from the McCullough Praks Section. Northern Bighorn Basin. Wyoming. J Gcol. 102,367-377, Coccioni. R.. Erba, E.. and Prcmoli-Silva. I. (1992). Rarrcrniatr-Aptiim calcareous plankton bioshatigraphy from the Gorse Cerhara scction (Marche, cmlral Italy) and implications Tclr plankton crolution. Crrrocuous Rcs 13,517-537. Cue. K. S., and Liddicoat. 1.C. (1994). Overprinting o1 a natural magnetic rrmancncc in lakc sedinlenrs hy a ruhrcqucnt higl~.inlensity field. I\intrrr? (London) 367, 5741). Coe. R. S., GrommC, S.. and Mankinm; E. A. (1978). Gemagnrlic palcointensities frixn excursion sequences in lavas cm Oahu. Hawaii. J. Grophve RPS.89, 1059-106l). Colcman, D. S., and Bralower, '1'. J. (1993). New U-Ph lircun agc constraints an the Garly Cretacenus time scalc. EVS, Tmns, Atn. Gcrjplzy. Union 73, 556. Collinnm. D. W. (1965). The remanent magnetism andmagnetic prupertirs of rcd sediments. Ceopltyr. R. Asfrm. SOL.10. 105-126. Collinson. D. W. (1974) Thr rolc ofpigmcnt and sprcularilr in thc rcmaneltt magnetism ,r red sandslonrs. C;PopI~yxI . R. Arlron. Sur. 38, 253-264. Collinstm. D. W. (1983). "Method? in Rock-maflctism and Paleonlagmrtiam: Techniques and Instrumentation." Chapman & Hall. London. Cnndomcs. M.. Moreud,P., C:amus,G.. and Duthon, L. (1982). Chronological and geochemical study of luvi~\from thc Chaine des Puys, Mnbsif Central. France: Evidence lor crusral conlaminiltion. b n r r i b . Minerul. Petrol 81,796-303

Constable, C. 0.(1985). Eastern Australian geomagnetic field intensity over the past 14.000 ?.r. tirophyr. .I. R Asfron. Sor 8% 121-130. Conslahlr, C. G. (1992). Link bctwecn geomagnetic reversal paths and sccular varialiun 01 the Geld over the past 5 My+. NOIUYS (London) 358, 230-233. Constable. C. G., and McElhinny. M. W. (1985). Holoccnc grumagnetic secular variation records from north-eastern Australian lake sediments. Geophy.".J. R. Asrron. Soc. 81, 103-120. 1c measureConstahlc, C. G., and PaAer. R. (1991). Deconvolution ollong-core paleomagnit' mentr, spline therapy for thr linear problem. Cieophss. .I Inr 104,453-468. Cunalable. C. G., and Tauxe, L. (19K7). Pnlaeointensity in the pelagic rcalm: Marine sediment data comoared with archaeomagnetic and lake sediment records. tieophyr. J, R. Aarua. so<, 9Q,43-59. Cope. J. C. W.. Getty, T. A,. Hownrth. M. K., Morton. N..and 'lotrcn.;. H. S . (1980a). A eorrclotion of J~rassicnxks in the British Isler. Part I: Introduction and Lowcr Jurasilc. Geol Snc London, Spec. Rep. 14, 73. Cope. 1. C. W., Dufl. K. L.. Parsons. C. F., Tonens. H. S., Wirnhlcdon, W. A . and WriDt. 1. K. (1980b). A currelation of Jurassic rocks in the British Isles. Part 2: Middle and Upper Jurassic. Geol, Soc. Lrmrlon. Spec. Rep. 15, 109. Courtillot. V.. and Besse. J. (1987). Magnetic field reversals, polar wander, and core mantle coupling. Srimcr 237, 1140-1147. Cuw, A. (1968). Length of geumagnetic polarity intcr\,al.;. 3. Gut,ph;s. Res. 73,3247-3260. C o x . A. (1969). Geomagnetic reverrills Science 163, 237-245. Cox, A,. and Dnlrymplr. G . B. (1967). Statistical andysis of geomagnctic reversal data and the precision of po,lassium-argon dating. J. Gerphys. Res. 72, 2603-2614. Cox. A., Doell, R. R., and Dalrymple. ti. B. (196%). Geomagnetic polarity epochs and Plcistr~cmegeochanomctry. Nature (Londorr) 1% 1049-1051. Cox. A.. Docll, R . R., and Da1r)mplc. G. B. (1963b). Geomagnetic polarity epoch: Sierra Kevada 11. Science 142,382-385. Cox, A,, Dcrell. R. R., and Dnlrymplr, 0.B. (1964). Kcvzrsals ul the Earth's lllagnctic ficld. Science 144, 1537-1543. Cior. A., Doell. R. R., and Dalryulple, G, 6. (1965). Quaternary paleomagnctic stratigraphy In "The Quaternary of the IJnired States'' (H. F. Wright. Jr.. and D. ti. Frcy cds.), pp. 817-830. University Press, Princeton, N.J. Cox, A., Hopkins. O.M., and Dalrymplc. F. B. (1966). Geomagnetic polarity epochs: Pribilot Ldands, Alaska. CeoI Sr,c Am. Bull 77, 883-911). Creer, K. M. (1974). Geomagnetic variations for thc interval 7000-25.000 yr BP as recarded in a corc of ssmmznt for station 1474 of the Black Sea cruise of "Atlantic 11." Earrh Pluner. Sci. Lcn. 23, 34-42. Crerr. K. M.. and Tucholka. P. (1982). Construction of type curvcs of geoma~neticsecular variation for dating lakc sediments from east central North America. Carl. J Et~rtlfSci. 19,1106-1115. Crccr: K. M., Irving, E.:and Ruomrn, S. K. (1954). Thc direction of the geomagnctic field in rcmote epochs in Great Britain. J. enma mag^. Geoel?cW. 6, 161-th8. Creer. K. M.. Mitchell. I. G.. and Valencio, D. A. (1971). Evidcnce for a nc>rmalgeumagnctic field polarity event at 263 i 5 My B. P,within the Late Paleozoic rcvcml inter1,al. Nolure (London) 233,87-89. Crecr. K. M., Thompson. R.. Molyneilun. L.. and Mackrreth: I.'. J. H . (1972). Gcomapnrlic secular variation recorded in thc stable magmelic remancncc n l recent sediments. firth Planet. Sci. 1 . d ~14, 115-127.

288 Crew. K.M.. Atiders<>n.T. W.. and Lewis. C. F. M, (1976). i.atc Q ~ gr',magnctic ~ slratigraphy rccorded in Lake t r i c sedimenlb. Enrrh Pllrnrr. Sn. l.~rr.31, 37-47. K. M.. H ~ g g.I' . E., Readman. P. W., and Reynaud. C. (lY80a). Palromagnetic variation curvcs extending back in 13;400 years BP rscorded by sediment,^ depositc. i n Lac dc Joux. Suzit2erland. 3. Get,phvs. 48, 138-147. Creer, K. M.. Rcadman, P. W., and Jacohs. A. M. (1YXUbj. Paleomagnctic and pleuntolo$cal deting o f a section ar Gioa Tauro. Tlaly: Idcntiticatiotl of lhr Rlakc evcnt. Eairri)i pklnpr Sci. 1.rrr. 50, 284-300. C ~ e r r K. . M.. Readman, P. W.. and Papmiriinopoulos, 5. (19111). Gcornagnctic ecuini tirln in Grcccc through the last 6000 yrars obtained fromlakr sadimcni sludirs. (:roplij,y. . I R. Asrron. Soc. 66, 193-219. CT~LT. K. M.. 'rucllolka, P.. and Barton, C. E.. cds. (1983a). "Grom~gnctisrnof Dakrrl cai:.\ and Rcccnt Sedimcnrs." Cl\rvirr, Amsrerdam. Crccr, K. M., Valcncio. D . A., Sinita, A . M.. Tucholka. P.. and Vilas. J . F. A. (lYX3b). Cicomagtletic secular variations U-14000 yr BP as recorded hy lake sediments from Argentina. Grop1iy.s. I R. Aslrnn. Sor 74, 199-121. Cunninghan~.K. J.. Fdn. M. R., and Rakie-El Beid. K. (1994). 4lagnetostruligi~phicdatillS of an Upper Miocmt: c h a l l o ~ m a r i n eand conlinentai scdimcntaiy rucceisi<m in noitlleastern blorucuu. firfh ~ ~ a r t cSci, r Leo. 127, 77-93. Dalr).mpl*. G . B.. Ccrx. A.. Docll. R. R.. and Gr'rmmt.. F. 5. (1967). Pliocene geomagnetic polarity epcxhs. Enrrh Plnrrrr. Sci. Lcrr. 2, 16.3-173. Davies. T.A . and Gwslinc. O. S. (1976). Oceanic srdirncnts and sedimcnlilry proccsscs. In "Chcmic~lOccanogtaphy" ( J . P.Riley and R. Chcstcr. cds.), pp. 1-80. Academic Prchs. Ucw Yr~rh. Day. R.. Fuller, M..and Schmidt. V, A. (1977). ITystcrcsis properties of ri~imumil~nrtitcs: Grain-ai~rand cumpositional dependence. Phyr F;urrh Pioner /,tien 13, 260-267. Dearins. J . A , and Floycr, K. A. (1982). The milgnrlic susceptibility of srdimm~inymatcrial trappsd in Luugh Eicagh. Nortl~einIrelandandits crasionsl sigt~ificancr.I.imnn1. Ore(>ringr. 17, '109-975. Dearing, J. A . Morton. R. I.. Price. T. W.. and k ~ a t e rI., D. L. (1986). Tracing rnuvumcnts or li~proilhy magncticmcasurcments: Twusarrstudics. Pf,?s. Ennh Pl<~nui. Inter 4 2 , 9 3 I@. De Boer. P. 1.. (1982). Cyclicity and slclrage i,f organic matter in tlnddlr Cieraccous prlagic sediments. In "Cyclic and Event Stralilicalinn" (G. Einszle and A. Srilachcr. cds.), pp. 46-475. Springcr~Verlag.Berlin. D t i n a A,. l'aum. L., Mona:han, M.. and Drakc. K. [lY90) "'hri-",4r fige uaiihratian of the lilho and palcomagnctic stratigmphies rll the Usorom Fm.. Kenya. J. G?rr>l.98, 567-587. D r k k r n . M.J. (19.98). Magnetic propcrtirs rlfnalwral pyrrhotite. Par1 1: Behiwirrnr of illitial ,urcrptihility and saturation-ma:nctl~arinn rrlatcd rock-magnetic paramrlers in a grainsize drpcndcnt framework. Plrjs Eunh PIon?r. Inter, 52. 376-393. Dekkers. hl. I. (1'189). Magncticpropertiesofnalural pyrrhotirc. I I H i g h and low-tempcraturc behaviour Jrr and 'IKM Hs a function 01 grain rizc, Phy.7. Earlh P1rrrrr.r l,~rpr 57, 266-283. Dekkers. M. J.. h'ldttri. J . T..,Fillion. ti.. and Rouhrltr. P. (1989). Grain Eire deprndrncc uf the mawetic hehirvir,r of p y h o l i t c during ils lirw-trmpcraturc transition a1 34 K. GPOphy.~,RCS.L e ~ r16 855-858. d c Menocal. P. B.. Ruddiman. Mi. F..and Kent. D. V. (19110). Depth o l poat-drposiii~nal rcman'cnce acquisiliun in dccp-rcn sediments: A case stud? or thc Krnnhrs-Motuyamild rcvcrral and oxyprn i.loli,pic stage 19.1. Enrill Plonei. Sci Id". 99, 1-13. Dcnhnm. C. R. (1974). Counler-cic>ckwiscmotion of paleuma@elic directions 24.UUU years ago at Mono Lake. Caliiurniil. I . f7eorricgn. Uroeivclr 26, 487-JYX.

289 I

~

~ ~ (1976).~Blake polarity episadr in two cores from the Greulrr ~ n l i l l e souter Denham, C. R. ridge. Carrh P1unr.t Sci. Len. 2 9 , 4 2 2 4 3 . ha^, C. R. (1981). Numerical corrclatUm of recent paletmagnetic rrcmds in two bkr Tahoc curer. E ~ ~ r Pluwci lh Sci. Lerr. 5 4 4 8 - 5 2 . nrnham. C. R.. adcox.A. (1971). kvidmce that the Paschamp Pvl.*riry Event did ncll llccur 133l)0-W400 yean spo. Enrtlr Pl<mrI.Sci. Lei1 13,181-190. ~ ~ pJ. J..~and ~1ngfam. l 8. ~ L ,(1982).High-rrsalutionstratlgr*phywithstrontium isotopes. Scirrtce 227, 438-Y4U. ~ i ~ hJ l F., . and shiv%.p . N. (1979). Paleomagnctic studies of the early Permian Logclsidr fomati,m of northern Colorado. G r o p h y . J. R. Asrru~.To(:. 56, 278-282. ~ i ~ hJ ,l F., , and shivs, p. N, (1981 j. Paleomagncfic results from the Ldte Carbonifurl)u.i/ ~~~l~ pcrniian cavprr formation: Irnplicarions for northern Appalachian trclonics. Enrrh Planel Sri. Leri. 54, 281-201. ~ ) i \ f ~J.;~and ~ Opdyke. ~ ~ , N ~ D. . (IT10). Paleomagnctismof the hlaringnuin and Shcpody formatir,ns, K~~ ~ ~ ~A ~ a~m u r~i a nmagnetic ~ i stratigraphy. ~ k Can. : J runh Sci. 27, 803-810. D ~ Y V. J., ~ alld ~ o p~d y k~ e . ~~D. . (1Wlilj. , Magncticpalarlty stratigraphy in the UPPcrmost ~ i ~ ~b i a ui c ~chunk ~ ~ formation, i ~ Puitrvillc. ~ i ~~cnnsylvania.Ueolos? 19%127-130, . Magnetic polaritvrtratigraphy and carboniferous l)ivenere. v.J., O p d y k e . ~ r, ~(1991b). palenpi,le poritiunr [roll, tllc ~ogginsSectian. Cumlrcrland Bwsin. Nova S o t i a . J. Geoph)3, R+.V %, I D L I - ~ ~ M . ~ c ~ l ~ i l ~ ~ d -E. ~ (1980). r u w nMagnetic . hlocking lempdraNrCS of single M. A,. dCrrnaingrain? during cooling. I . Gt.uphyl. Rrs. 85, 2615-2637. ~ ~ R. , E. j ( 1~ ~79~ ) .Counterclock~~iar ~ . precessinn id lhc geomasncric field vrctm and westward drift of thz nnndipolc Bcld I Gc,ophyr. Hfs.84. 637-6114, ~ ~ l ~G . B. y (1966). ~ ~ Gromaguetic l ~ , polarity epochs: A nCw pulilnly ~ ~ ~ R,1 K,: 1 . cvcllt and the age of thc Rrunhes-Matu!'ama hirundary. Sciedce 152,1060-106l. ~ ~ ~ R. 1 R.. 1 .D ~ I ~G. 6.. ~ and ~ ICUX, ~ A.. (1966). Grr~mugnericpolarity c p c ~ h ~Sierm : ycvndaData. 3. J. Ger,l,hy. Rcs. 71, 571-541. ~ ~ S. hJ.. King, . J. \V., and Lrined. M. !198X). A rock-magnetic studl of Giant Pisllln cureLL.IJ-GPC from thc c c n t r ~ lNorth Pacific and i l h palcocenagraphic ri~ificallcc. Palrr~.eonogr~pli!' 3, 89-111. ~ ~D. M , (19x8). ~ ~a]com*gnelics ~ l of Kingcrikr ~ ~ Old Red, Sandsfone and rslmed rocks. southern N ~ In~plicationc ~ ~for prc-Carbonifcruils ~ : reparatiott of Bdllicu and British trrranes. 1'ecronryhy.sicr 148. 11-27, ~ ~ D. , J.~(1972). l ~ ti^ ~ , mineralogy o%unhri~trB and heatcdred%edirnentsbycocrcivit? spectrum analysis. Geophyr. .I. R. ilsl~on.Soc. 27.37-55. D~,,~~,P. D. j. (1979). O n the usc of Zijdervrld vector diagrams in multicomplmm~Paleonlagnrtic studies, Plr,ye hxrih Plartrt. Inler 20, 12-24 ~ ~D. J. (1983). ~ Viscrlm l magnetization ~ ~ of, (1.04-100 pm milgnrLi1es. G ~ V P ~ JY. R. Asrron. Si,c 27, 37 -55. D,,~~,,~,D.J . ( i y ~ h )~. ~ y s t e r c spr,,prrtias ii of n~agnetitcand theirdepclldcncr on paflxlc rift: of the pscud~l.ringle.domain rcmanrnce madels..l. Guoyi~ys.RPS. 91,9569-YS84. A ~ ~ , ,D.l I.~( ~~y y,o j ucvclopments . in rock milgnrtisn1. &P. Pwy Pt~?s.53,707-7911. ~ ~ , , [;. ~ R., i ~and~ H .O ~ Y C T.~ .P. (1990). UIPb zircrm and haddclrrile ages for the Palisades Slat=: lmplicatillns lor the age of the and ~ ; ~ t t 01~thirnurthrartern ~ h ~ ~ Unitcd ~ TriassiclJurassic houndarj. Geoiogv 18, 795 798. ~ ~ i ~ ~ A, , M. ~ ,( 1~q ~~4 )k~. i ~, ~thc~lower i mantle: n g Dctcrmination oflstcrill heterogcnsil? in p vclncity up to degree and ordcr 6. I C i r r ~ ~ ~Rr9 h ~ s391 . 5929 5452-

"c&,,,,,

25'0

Bihliography

Dzieaonski, A. M.. and Woodhause, J . H. (1987). Global images of the eanh's interior. Science 236,37-48. Eardley, A. I., Shuey. R. T.?Gvosdetsky, V.. Wash. W. P..Picard. M. D., Grey, D. c.. and Kukh. G. J. (1973) I.akc cycles in the B~~nnevillc Basin. IJtah. Ccol SOC. Am. Blc[l, U4,211-216. Eddy, J. A. (1976). The Maundcr minimum. Science 192,1189-1x2. Egbert. G. (1992). Sampling bias in VGP longitudw. Ceophyr. Res. Lett. 19,2353-2356, Einarsson, T.. and Sigurgeirsson, .r. (1955). Rnck magnetism in Iceland. Naiurr. (Lrjndon) 175, 892. Ellis. B. 1.. Levi, S..md Ycats, R. S. (1993). Magnetic stratigraplry ni rhc Morales formation: Late Nengcncclockwisfro1ation and compresston in the Cuyama Rssin, California Coag Ranger. Tedonics U , 1170-1179. Plsaxsar. W. M. (1946). Induction *ifects in lerrestrial magnetism. 1. Theory. Plfyr Rcv. 69,106-116. Elstun, D. P.. and Perucker. M. (1979). Detrital magnetization in red beds ui the Macnkopi formation. J . Geophys. Res. 84, 1653-1665. Ernhlcton, B. J . J.. and McDonncU, K. L. (19817. Magnetostratigraphy in the Sydncy Basin. southeastern Australiil. In "Global ReconstruCtions and the Gcomagnelic Ficld During thc Paleozoic" (M. W. McElhinny et ul., edr.); Adv. Earth Planet. Sci.. Vol. 10, pp. 1-10 Center for Academic Publ.. Tokyo. Emiliani. C . (1955). Plriaocenc temperalure%.J. Ceol. 63,538-578. Ernesto. M., Pacca, 1. G.. Hiodo, F. Y., and Kardy, A. J. R. (1990). Palaeomagnctism of the Mesozoic Sierra Geral formation southern Brazil. Phyr. Eurrh Plrrncr Inter M, 153-175. Eusley, R. A.. and Vcrosub. K. L. (1982). A magnetostratigraphic sludy uf rhc sediments oi the Ridge Baain. southern California and its tectonic and srdimmtulogic implications. Earth P/unrr Sci Len 59. 192-207. Evanoff, E., Prnthero. D. R., and Landcr. K. H. (1992). Eocene-Oligoccnc climatic change in Norlh mer rim! 'The White River Formation near Dnuglaa, east central Wyoming. 1tz "Eocene-Oligocene Climatic and Riotic Evolution" (D. R. Prothrra and W. A. Berggrm. eds.), pp. 116-130. Princeton IJniv. Prcss. Princeton. NI. Evans, M. E.. and Heller. F. (1994). Magnetic enhancement and pillcoclimat'e: Study or il loessl~dleusolcouplet across the Locss Plateau of China. Geophys. J. Inr. 117.257-ZM. Evans. M. E.. Wang, Y.. Rutter. N.: and Ding. Z. (1991). Preliminary rnngnetic stratigraphy of thr red clay underlying the locss sequence at Baoji, China. Ceophps. Rt-5. Lett IR(R); 14W-1412. Evrrndm. J. F., Savage,D. E.. Curtis, G. H.. and James. G. T. (19M). Potassium-argotldatrs and thc Ccnozoic mammalian chronology of North Amcnca. Atn. J Sci. Wi2,145-198. Eyrr. I.R,and Shaw. 1. (1994). Magnctic enhancement of Chinese lnrsr-thc role of yFe>O,. C i e o p h , ~J.~Int. 117, 265-271 Fang, W., Van dcr Voo. R.. and Liang, R. Z. (19hY). Reconnaissance magnetostratigraphy o l the Precambrian-Cambrian boundary scction at Meichueun. Southwest China. Ctrod, Geul Iher 12, 205-222. Farr, M. R.. Sprowl. D . R., and Iohnron, 1. (1993). Identification and inilial correlation of magnrtic TcvcSl in the Lower to hliddlc Ordovician of northern Arkansas. In "Applicatirlns of Paleomagnebsm LO Srdimcntary Geology" (D. M.Aissilr,ui. V . F. McNeiU. and N . F. Hurley, eds.). SEPM Society for S c d i m c n t q Geology Sprc. Publ. No. 49. pp. 83-93, Fisher, R. A. (1953). Dispnsion on a sphere. P m c R. Artron. SOc. A217, 295-305. Plynn. J . J. (1986). O~rrclarionand geocllronology illmiddle Eocene strata from the western U.S. Poluwgeogr., Palneoclunnfol, Palrrrurml. 55,335-406.

Flynn; J. 1.. MacFaddcn, B. J.. and McKcnna, M. C. (1984) Land mammal ages, faunal heterochrancitv. and t r m v ~ ~ rrsolution ~al in Cena~oicterrestrial sequences. I. Geol. 92, 687-705. Foland. K. A . Gilbert, L. A,. Sebring, C . A,. and Jiang-Fmg, C. (1986). "'Ari?PAr ages lor plutonr o l the Montcrcgian Hills. Quebec: F,t,idmca for a single episodc of Crrlaceous msgmatism. Geoi. Sot Am. BUN 97, 966-974, Fontcr T. H., and Hcllrr. F. (1994). Lorrs deposits from the Tujik deprc>sinn (central Asia): Magnetic properties and palcoulimate. Eerrh Plnnrr. Sci. Len 128, 501-512. Foster. I. H.. and Symons. D. '1'. A. (1979). Detinins a paleornagnctic polarity pattern in the Monteregian inhusives. Con. I Eurlh Sci. 16. 1716-1725. Foster, J. H. (1966) A paleomagnrtic spinner magnelc,meter using a Run gate gradiomctcr. Ewlh Plancr. Sci. Lrzr 1, 463-466. Fostcr. 1. H.. and Opdykr. N. D. (1970). llpper Mioecne to Rccznt magnctiu rtrati$raph!* in drrp-sea sedimcnts .I. Geuphys. Rer. 75,44654473. Frankel. R B. (1987). Anilerobzr pumping iron. Nat1:anrrP (1.rtndon) 330. 2U8. Frccman, R. (1986). Magnetic mineralogy uf pelagic limcsrunes. Gropttys. J. K. Artron. SOC. 85, 455-452. Fricdman, R.. Gee. J.. Tauxe, L., Downing. K.. and Lindsay. E. (1992). The nlagnctaslrilligI'a"hy of thc Chitanvilla and lowcr Vihowa formations of the Dera Ghazi Khan area. Pakistan. Sedimenr r701. XI, 253-2hR Friend. P. F., Johnr,n. N. M.. and McRnr, L. E. (198Y), 'l'imr.lrvrl plots and accumulatioll rates of scdimrnl sequences. Ceol Mag. 126,491-49%. Fry. G. J.. Bottjcr. D. 1.. and Lund, S. P. (1985). Magnctmlratigrlrph).l~of displace Upper Cretaceous strala in southcrn C:alifornia. Gcolog~z13, WX-651. Fuller. M, V.. and Knbayashi, K. (1967). Identification of magnetic phascs in crrlain rocks by lour-rcmperaturr analysis. In '.Methods in Palcornsgnetism" ( 0 . W. Collinstm. K. M. Creer, and S. K. Runcorn. cds.). pp. 529-534, Elscvirr. New York. Fuller, M. D.. Williams. I., and Hoffman. K. A. (1979). Palromagnetic records uf grtrma~netic field revcrswls and the morphology of the transition lirlds. Rev. Gwphyr. Cpoce Phys. 17, 179-203. Fullun, R. J., Irving, E.,and Wheadan, P. M. (1992). Stratigraphy and palcumagnetism of Btunhes and Matuj,ama, Quatcmary deposits at Mcrritt, Uritish Columbia. Con.3. Earth Sci. 2, 76-92. Fundor, S . (1989). Quarcrnirry geology ofthc ice-free arcasand udjacentshelvcs of Greenland. I n ..Ouaternary- G~.olc,g of Canada and Greenland" (R. 1. Fulton, cd.), Vul. I. pp, 743-792. Geol. Surv. Can. Giuher-Puertas, C. (1953). Varadon sceular del campo gcomagnetico. Obsew. Elro Merlzo. 11, 4-16. Galhrun, B. (1985J.Mnpneto~tratigraphy of the Berriasian Stralotypesection (Bcrrias, Fran~e). Eurlh Plmrer Sri. 1.etl. 74, 130-136. Galbrun. B. (1989). H&ullats magncu,slriltigraphiqucs h la limite Rognacicn-Vicmllien: Precis i o n ~de la limilr Cretaco-Trrliar dans le Basin D'Aix-cn-Yrovcnener. Cnh. RcseNe Geol. Hnrcr~-Prr,vrnrc,1Jigne 1, 34-37. Gdhrun, B . , Gabilly, J., and Kaspins, L. (1988a). Magnetostratigraphy o i the Toarcian stratotype sccti<,ns at Thouars and Airvault (Ucux-Srvres. Frilncr). Eurrh Plo,zet. Sci. I.ett 87,451-462. Galbrun, B., Baudin. F, Cumas-Rengilo. M. J..Faucault, A . Fuurcade, E., Goy. A,.MoUlercade, R.. and Rugst. C . (IY88b). Re~ulliltsmagnetostratigraphiqucs preliminaire* sur Ir Toarcian dc la Sierra Palomcrr (Cahine iberiquc. Espagoe). Bull. Soc. CwL I% 8,193-198.

Galhrun, H., Baudin. F.. Faurcade. I?., and Riuas, P. (1990). Magnetoslraliqilphy of the Toarcian Ammr~nillcnKosso limesronr at h n a l l e ~Spain. . G ~ o p h y sRrs. . LRI.17,24412444. Galbrun. B . . Crasquin-Soleau. S., and Jauqucy. J. (IYYZ'). MagnAtostratigraphic dcs sCdimrnts triasiques des Sites. 279 el 760, ODP Leg 122. Platcau de Wombat, Nord Ouest dc 12Australie.Mar Ccol. 107,293-198. Galbrun, B.. Feist. M.. Colombo. F.. Rocchia. R., and Tamharcdn,Y. (1993) Magnetorlrarigra. phy and hiostratigraphy of CretaRous-Tertiary CnntinmVal dcporits. Ager Basinpnrvin~ of Levida. Spain. PaL~eo~copr.. P~rlaeocli,nuruI.,Palrrenuml. 102,41-52. Gallet. Y., Besse, J.. Krystyn. J.. Marcow, J., and Thermianl, H. (1992). .Mngnetostratigraphy of the Late Triassic Bolucekreri Tepe .;eclii,n (\outhwrstcm 'l~urkey):Irnpli~ cations for changes in magaetic reversal irrqu~ncy.PIzys. Enrrh Planet Inr<,r 73. Ki-

108. Gallet, Y.,Bessc. I..KrysLyn, L.. Thrumiimt. H.. andblsrcoux, 1.11993).Ma@ietostratigraphy of the K a v u Trpe section isouthwrsern Turkey): A magnctic polarity timc scale lor 111s Norian. Etmh Plunet. S r i [.en 117,441-156. Gallet. Y.,Besse. J..Kryslyn, L.. Thdvmianl. H.. and Marcoux. J. i14Y4). Magnctostratigraphy ofthr Mayrrlingaaction (Austria) andE~enkoluMczarliklilrkcy scctions: Improvemn~nt of the Carnian (Lale Trii~\sic)magnetic polarity timc scalc. Earth Plo'(urrci.Sci. Luii. 125, 173-191. Galusha, 7.. Joln~son.N. M., Lindsay. E. H.. Opdvke. N. D.. and Tcdford. R . H. ilY84). . . Biostratigraphy and magnetostr~tiqilphy.lati ~licoencrocks. 111Kanch Arizona. GPO/, Soc Am. Ball 95,71&722. Gauss. C. F, 11838) Allgen~einen e u r i r deh E~dmagnrtismus.Lcipzig. Gee. J., Kootwijk. C. T., and Smith, G . M. (1991). Wdgnctostratigraphy of Palcoccnc and I:ppcr Crctaecous SedimentSfromBroken Ridge, eastern Indian Occan. In "Proceedings of thc Occan Drilling Program. Scientific Reullr" (J. Wcisrci. J. Picrce. E. Taylor. J. All el aL, cds.). Sol. 121, pp. 359-375. ODP, Cr,llrgr Station; TX. Cilkrt. W. (16W). "Dc Magncte." (P. F Mattely. Iran%.)1893: Davcr. Ncw York. 1 5 8 . Giouuntrli, F. (1979). A comparison of the nla~nettzatii,nnf detriral and chcmical sudimcnts hjrn Lakc Zurich. U~ophys.RFS LCII. 6, 233-23.5. Glas~.B. P., Kmt. D. V.. Schncider, D. V.. and Taune. L. (1991). Ivory Coast microrcklitc strewn field: Duscription and relation tu the Iaramillr>peomagnctic cvcnr. Earth Planer .%i. Len 107, 182-196, G~,rdrll.H. F.. and Warkins, N. D. (1Y68). The paleomapetic stratigraphy of the Suuthcm Ocean: 200 Wcrt to lb0' East longitude. Drep-Su. Ris. 15. 89-112. Gorrc. W. 5.. and Fulicr. M. (19761. Magnetometers sing RF-driven aquids andlhrir applicatinns in ruck rnagnctism and palcamag~l?tism.Rev. Geuphyx. Spaor P1ty.v. 14, 59-608, Owe. W. A.. and Hclsicy. C. E. (19721. Paleonlagnetism and rock magnetism of the P~rmiim Cutler and Elephant Canyon Formations in Utah. J. Ccophys. Krr. 77.1534-1548. Cough. D. I. and Opdykc. N. D. (1Yb3). 'lhc palcomagnetism of the Lupata Volcanica of Mazambiquc. Geophy~?.J. K. A.?rrf~n..Tor. 7, 457-48, Gradstein; F. M.. Aglrrhrrg, F. P., Ogg, 1. G., Hsrdcnhol, J.. Van Vccn, P., Thierry. J.. and Humg, 2. (1994). A Mesnraic time %ale. 3. Geophys. RPS.99, 24.051-24.074, Gradstein, F. M.. Aylerherg, F. P.. Ogg, I. G.. Hardenhul, J., van Vcun, P.. 'Thicrry. J.. and Huang. Z. (1995). A Triassic, Jurassic. and Cretaceous lime~calr.In "Gcachronology. Time Scalcs and Stratiprophic Correlation" (W. A. Berggm, D. V. Kent. M. Auhry. and J. Hardcnhol, eds.). SEPM Spec. Publ. No. 54, pp. 95-128. Graham. J. W. (1949). The stability and significanm of magnetism in sedimenlary rr,cks. .I Geopl~ys.Krs. 54,131-167.

Gtomm6. C.S.. and Hay. R. I,. (1963). Magnetization of basalt of bed 1, Olduvai Gorge. ."intr~rfl( L m d r m ) 200,560-561 GrommC, C. S.. and Hay. R. L. (1971). Geomagnetic polarity epochs: Agc and duration of the Olduvai normal polarity event. b;arrh Plancr. Sci. Ldr. 18, 179-185. FTL)UL I. J . . d e Jongc. R. F. C.. I.angerris, C. G., rcn Katc, W. G. H. %., and Smit. J. (1 989). Ma~netomatigraphyof the Cretaceous-'I'crtiar). boundary at Agost. Spain. Earth Plrmel. Sci. Lcfr. 94,385-397, Groussct. F. F... Labeyrie, L., Sinko, J. A., Crzmer, M.. Bond, G., Dupraf. J., Cortijo. E., and Huon, S. (1993). Pattcrns uf icr-rahed detritus in the glacial North Atlantic (40-55'N). Pnl?oceunornrphy 8, 175-192. Gubbins, D. (1987a). lbmmal core-mantle interactions and lime avcrnged paleomagnctic ficld. I. Gruphys. Res. 91,34133420. Gubbins, D. (1987b). Mechanism for geomagnetic reversals. .Nclttre (Lanrlo?~)326,167-169. Guhhins. D. (1994). Geon~agncticpolarity reversals: A connection with scculnr variation ond core-manllr interaction'? Kcv. Geopl~ys.32,61L83. Gubbins, D., and Coe. R.S. (1993). Longiludinally motincd geomagnetic rcvcrsed paths from non-dipolar transitionficlds. Nature (Lundon) 362, 51-53. Gubbins, D.. and Kelly, P. (IYY3). Pcrristrnl parterns in thc pcomagnetic field during the last 2.5 M yr. Nurure (Londo,~)365, 829-812. Gurcvich. Y. L.. and Slautsitay~,1. P. ($985). A paleomagnctic rcclion in the Upper Parmian and Triavsic deposits on Novaya Zcmlyu. Inr. Ceol. KC". 27,168-177. Guyodo, Y.. and Valet, J. P. (1996). Relative variation in gcomagneticintensity from sedimentary rrcnrds: The past 200 thuusimd yzarr. Earlh I'lanat Sci. LNIPIS, in press. Haae. ". M.. and Heller., F. (1991I. . . Iiltr Permian to Earlv Triassic magnctostraligraphy. Earth Planer Sci Lcrr 107, 42-54. Haeclbere.T..Shackleton.K.J..Pisiar. Y.. andSl~iphaardScientificPatty (1992). Dewlopmcnt c. of cumposite depth section? frlr Sites 844 through 854. In "Proceedings of the Occan DrillingProgram, Initial Repura" (L. A. Maycr, N. G.Pisias, T.K. Janrcek el a/., cds.), Val. 138, pp. 79-85. ODP. Ccrllrgr Station. 'TX. Haggcny. S. E. (197ha). Oxidatir~nol opaque mincrals in hasalts. In "Oxide Minerals" P. Kumblr, III, ed.), pp. 1-IWI. hifineralogical Socicty r!l America, Washington. DC. Haggcrty, S. E. (1976b). Oxidittiun 01 opaque mincrals in lrrrestrial igncous rocks. Irr "Chide ,Mincralr." (D. Rumble, TIT. d.). pp. 101300. Mincrdogical Socicty of America, Washington. DC. Hailwood, E. A . (1979). Palcumilgnelism of Late .Mesozoic to Holoccne sediments from thr Bay of Biscay and Ruckdl Plateau, drillcd on TPOD Leg 48. In Tnitial Reports of the Deep Sea DrillinpPn~jccr"(L. Montadcnt, D. R r l k t s d o / , cds.). Vol. 48. pp. 305-333. I7.S. Govt Printing Office. Washington. DC. Hailmod. E. A,. and Clement. B. M. (1991). Magnetostratigraphy of Sites 703 and 7111. Meteor Risc, southeastern South Atlantic. In .'Proceedings n1The Ocean Drilling Prnjrcl. Scientific Resuls" (P. F. Cicsiclski, Y. Krisroffersen et ol., rds.), Vol. 114. pp. 367-386. ODP, Collegr Station. TX. Hall. C. M., and FmeL J , W. (1993). Laser ""i\rFVAr age from ash D of ODP Site 758: Dating the Brunhes-Matuyama reversal and oxygcn iaotope stage 19.1. EOS, Trans Anr, G r o p b s . L'nion, 74, 110. Hall. F. R., Bloemendal, J., King, J. W., Arthur, M. A , and Aksu. A . F. (19898). Middlc to Isle Quaternary sedimenl flmes in the Labrador S a , ODP Lcg 105, Site 64h: A synthesis of rock-magnetic. oxyscn isotopic, carbonate and planktonic foraminilrral data. In "Procredingr of the Occan Drilling Program, Scientific Reaults" (S. P. Srivastava. M. Arthur, B. Clement crnL, cda.), Vol. 105, pp. 653-68R. ODP. Collogc Slation. TX. k

294

Bibliography

Hall, F. K.. Bus&. W.H.. andKing,]. W (1989b).Therelationship betwemvilriaiic,nrin rockmagnetic properties and wain s i x of sediments from ODP hole 645C. In "Procxcdings of the Ocean Drilling Pmgram. Scirntiflc Results'YS. P. Srivastava.M. Anhur. R. Clement ~r111.eii\.), V~11.105: pp. 837-84. ODP. College Station. TX. Hallam. A. (1975). "Jurassic Environments." Cambridse Univ. Press. Cdrnhridgc. UK. Halls. H. C. (1976). ,\ least nquares method to find a rcmanencc direction fmm converging remagncti,.ation circles. Gropkys. J. R. Asrron. Soc 45,297-304. Halls, H. C. (1978). 'The use of converging remagnetization circlr;s in p~lrnmagnctism.P/rY.~. Eztrih Ptuner Infer 16, 1-11 Halvorsun, E.. 1.unandnwski. M.. and Jclcnska, M. (1989). Paleolnagnetisnl of the i!ppcr Carboniferous Stzegom and Karkonose granites and the Kudowv Grimitnid from the Sudrt Mountains Poland. P1ry.s. Zartli Planer. lnrer. 55, 5 4 4 Hamilton. N.(1990). Mcsozoic magnctostratigraphy of Maud Rise, Antarctic, 1990. Itr "Pruceedings of the Ocean Drilling Project, Scientific Results" P. F. Barker; 1. P. Kannctt uf nl., eds.), Vol 113. pp. 52jJhO. ODP. Cnllegr Statiun, TX.

Project Lep 7 2 In "Initial Reports of the Deep Sea Drilling Proircl" (P. F. Barkcr. R. I.. Carlsun. D. A. Johnson et 01, eds.). Vol. 72. DO. . . 723-730. L:.S Govt. Print~ne Oificc, Washington, DC. Handschumacher. D. W.. Sacet.. W. W.. Hildr. T W. C.. and Bracev. PTS, D. R. 11988). ~ Crcracrnus trctc,nic rvt,lutirm 'II the Psofic plate and cxtcnsiun of the geomagncric polarity rcvcnal timcscalc with implications for thc origin of thc Jurassic "Quiet Zone." T ~ ~ c ~ o n o a h ~155. ~ s i c365-3W. r Hanna, R.; and Verosub. K. L. (1989). A review 01 lacuswine pillri~magnrticrecords Tram wcstcrn North America: fl-40.flDO vearr RP. Phvs. Farlh Planet. lnrer 56. 76-95. Haq. B. U.. Worsley,'T. R.. Burcklc. L. H.. Douglas. R. ti.. Keigwin, L. D.. Jr.. Opdyke, h'. D.. Savin, S. M , Somrner. M. .4..11. Vincent: E.. and Woodruff, F. (1980). The lilts Miocene marine cnrhrm irtrtupe shill and the rynchrunrily trl wmr phytoplanktonic biostratigraphic datums. Geology 8,427432, Haq. B. 11.. Hardznbol. J., and Van; P. R, (1987). Chronology of the fluctuatio~~ of sea leveb since tho Triassic (250 milli<myears ago la presenl). Sciencr235, 1156-1167. Hadand. W. B., Cox. A. V., Llcwcllyn. P. G., F'ickt~m.C. A. 0.. Smith. A. G.. and Waltm. R. (1982) "A Geologic Time Scale.'' Cambridgc Univ. Press. Cambridge, UK. Harland. W. B., Armstrong, R. L., Cox, A. V., Craig. L. E., Smnitll. A. G.. and Smith. D. G . (1YW). "A Cicalogic Timc Scale 1VRO." Camhndge Univ Press. Cambridge. UK. Harrison, C. G. A,, and Funnell. B. M. (1984). Relatiunship of pale~rmagneticrcvcrsals and minopaleontology in two late Cenozoic corcs from thc Pacific Occan. rVuture (London) 204, 566. HarLI. P., Tauxc. L., and Constable. C. (1993). Early Oligocene geoniagnetic field behavior from Dccp Sce Drilling Prajcct Sitr 522. J Gur~phys.Rr.s. 9U, 19,M9-19,665. Hartstra. K.L. (1982). A comparative study of the ARM and I,, of %,me natural magnetites of MD and PSD pain sire. Gropilys. I . R. Astron. Soc 71,497-518, Hawshida, A,. and Blormendal, J. (19911. bfagnetostratigraphy of ODP Lcg 117 sediments fmm the Owen Ridge and Omtn M a w , Wrstrrn Arabian Sea. In "Proceedings of the Ocran Drilling Rojcct. Scientific Resulls" (W. L. P ~ r l l N. , Niltruma ur ul., eds.). Vol. 117, p. 161. ODP. College Station, 'lX. Hays. I. D.. and Opdyke. N.D. (1967). Antarctic radiolaria. magnctic rcvcrealr. and climatic change. Science 1% 1001-1011.

.

.

Hibllagraphy

295

Hays. J. D.. Saitrk T.. Opdyke, N. D., and Burclilc, L. H. (1969). Plioccns and Pleistocene scdimenls of the equatorial Pacific Their paleoma~netic,biostratigraphic and climatic rrcrn'd. CuoL Sue. Am. Bull. 80, 1481-1514. Hays, J . D.. ln~brie.J., and Shncklcton, N. J. (1976). Variationsin the earth's nrhit: Pacemaker of the ice agcs. kicnce 194, 1121-1132. lieath, G. K.. Rca? D. H.. and Levi. S. (1985). Paleoma~nctismand accumulation ratrs of sediments at sites 576aod 578. Deep Sea Drilling Rojcct Leg 86. western North Atlmlic. In "Initial Reports of thc D r r p Sea Drilling P ~ o j e c r(H. R. Heath. L. H. Rurcklr el a/,, cds.), Val. 86. pp. 459-502. U.S. Govt. Printing Office. Washinpton, DC. Hedlcy, I. G. (1971). l'hc weak ferrornagnctism uf grothite (u FeOOH). I. Grr,ph?.~.37, 404-42fl. Hsider, F.; Dunlop. D. I . , and Sugiura, N. (1987). Magnelic properties of hydrothermally recrystalli,rd magnetite cr).slals. Science 236, 1287-12W. Ilsiniaer. C . . and Heller. F. i19761. A hirh . I. . tempcraturt. vector m a p e t o m ~ t e r Gcophys. R. ~ a r i nSoc. . 44,281-'287. Hrirtzler. 1. K.. Dicknm. 0 . 0..Hcm0n.E. M., PiTtman, W. C.,,nI. and LePicllon. X.(1968). Marirre magnetic anomalies. geomagnetic field rcvcrssl and morinus of the occan floor and contincnrs. J Gut,phys. Res. 73, 2119-2136. Hcllrr, F.(IY78). Kockmugnetic rtudics of IJpper Jurassic limeslrmes frolllsoutllern(icrmany. I . Geophyr 44. 525-543. Hqllrr, F. (1980). Sclf rcvrrnl of natural rrmanent magnctiradon in the OlhpLaschamp Iavas. Nanrrc (London) 284, 334-335. Hcllcr, F., and Channell, 1. E. T. (1979). Palrnmvgtletism of 11pper Cretaceom limcstlrnra irum the Munstcr Rasin. Germany. I. Grophy~.46, 413-427. Htller, F, and Evans, kf. E. (1995). Loess magnetism. Rrv. G~(1phys.33, 211-290. Hellrr., F.. and. Liu.. T. dating 01 loess deposits in China. ~S. (1982). Maencrost~atimaphical - . Niziurc ( L o t i d n ~.NO, ) 161-163. Hrlirr. F.. and Liu. T. 11984). of Chinese loess dcporils. Geu~hys.J R. Anron. , , Mae~retism .. Suc. 77. 125-141. Ileller. F., and Petersen. N. (1982). Self-mvcrsal explanation for the LaschampiOlhy gcomagnelic ficld rxcursion. P h y ~ Earth . Plnrvi. 1,ti 30, 358-372. Heller, F., Iawrir, W.. and Channell. J. E. 'V. (1984) Late Mioccnc magnetic stratigrqhy at Dccp Sea D~illirlgProject hole 521A, In "Initial Reports of the Dcep Sca Drilling Proiscl" (K. eds.), i a l . VoI.73.p~. , 637-644.U.S. Ciovt.Prinling , , J.Hsu,J.L. L B B r ~ ~ ~ l ~ c e Office. Washington, DC. Hcllcr. F.. Carracedo. J. C,. and Solcr. V. (1986). Reversed magnctiralitln inpyroclasti~.* from the 1985 eruptiotl of Ncradc dcl Ruiz, Cnlumbia, ~Va:arr'r~ (/,ondon) 324.241242. Hcller. F., Lowrie. W.. Li. H., and Wang, J. (1988). Magnetostratigrh ofthe Pcrmi-Triassic houndary section at Shangsi. Errnh Plonut. Sci. Len 88,348-356. Hcllcr, F.. Liu. X. M..Liu, T. S.. and Xu. T. C . (1VY1). Magnetic susceptibihty ui loess in China. Earlh Plart~I.Sci. Len. 103, 301-310. Hcllcr, F., Shen, C. D.. Bcrr. J.. Liu; X, M., Liu, 1'. S.. Brongcc. A,. Sutcr. M., and Bonani, C+. (1993). Quantitative estimatesand palncoSlim~licimplications of pcdogrnicfirromagnctic mineral formation in Chinesc locss. Eorrh Plunel. Sci. Lcn 114, 385-390. Hclslcy, C. E. (19651. Palcumagnetic rcrultvfrom the LuwerPcm~ianDunkard series of West Virginia. J. Geop$tys. Re2 71). 413-424. Helslcy, C. E. (1969) Magneticreversal stratigraphy of the Lower Triassic Moenkopc formalion of western Colorado. G o 1 Soc.Am. BUN SO. 2431-2450. Firlsley. C. E., and stcincr, M. B. (1969) Evidence for long intervals of normal polarity during the Cretaceous period. Errrih Plnner. Sci. Left. 5, 325-332. ~

lkebc. N., Chiji, M., Tsuchi: R., Momrumi, Y., and Kawata, T. (tV81). IT, "Proceedings of thc IGCP-I14 International Workshop in Pacifio Neogene Bii,stratinaphy." Kos. i-iv. pp, l-lrn. Imbric, J., Hays. J. D.. Maninmn, D. G.. Mclntyre. A.. Mix, A. C.. Morlcy, J. J.. Rrias. N. G., Prcll. W. 1.. and Shackleton, N. J. (1984). The urhital thcorv of Pleivtocenp ..- - ..climate: Support from a revised chronolr>poi lhc marine a'% Record, &A TO AS1 xrr, Sex C 126,269-205. Imlay, R. W. (1961). Late Jurassic Ammonites lrnm the Weslrm Sirrr'a Ncvada Calilomia. Geol Surv. Prof Pop. ( U S ) 374D. Dl-030. Irving. E. 11964). "Palcomagoetism and Its Application t o Geological and Gruphysical Pr<,h Icms." Wiley, New Ynrk and Lundirn. Inin& E. (1966,. Paleomagnctism of somc Carboniferous rocks fr0m New South Walcs and its relation to geological events. J . Geoplrys. Res. 71, 6025-6051 Irving. E.,and Couillard. G. W. (1973). Cretaceous narmal polarity interval, rkrurc (Lunrlon) Ph'hp. St;.244.10-11. Irtfinp, E., and Major. A. (1964). Post depositional detritvl remancnl rnagnctization in ri synthetic sediment. Srdimentnl<~g.y3, 135-113. Irving. k.,and Mongcr. J . W. H. 11987). Preliminary palcomagnetic rzsulls from ihc Pwmian Asitka group. British Colunlbia. Can. J. Earrtz Sci. 24, 1490-1497. Irving, E., and Park. J.K. (1972). Havpins and sUperinlrrr.i~ls.Can.J Earth Sci 9.1318-1324. Irving, E., and Parry, L. G. (1963). The magnetism of some Pernlianrocks irom New Suuth Waloe G ~ o p h y sJ.. K. AsiroIt. Soc. 7, 395411. Irving. E.. and Pullaiah, ti. (iY76). Reversals ul lhr geumagnctic field. maQnetoytratigraphy. and relative magnitude of p a l ~ ~ n r c u l avariation r in the Phaaeromic. Eurth S<;iRPV. U , 35-64, Trving. F...and StronsD. F. (1985) PBleonlagnetiSm olrrckc from Burin Pcnlnsula. Newfound~Y. land: Hypotllesibof Late Paleo~oicdisplacrmcnt of Aeadia criticized. J. G C L I P ~RLY. YU, 1949-1962, Irctl, G. A. and Ohradovich, J. D. (1991). Dating of the Miltuyarna-Brunhcs boundary based onU'Ar13'Ar ages of the Bishops Tull and thc Ccrro San Luis rhyolite. GmL Svc. A m A ~ S W . P,OX. n, AIM. I ~ e t l G. , A.. Ohradovich, J. 0.. and Mahnerl. H. 11. (198Rj. Thc Bishop ash hcd (Middlc Plcistoccnc) and some older iPlincrnc and Plcistoccnc) chcmicatly and mineralogically similar ash beds in Calilr,rnia, h'cvada, and Utah. GeoL Surv. Bult. (L!S.j 1675, 1-37. Jackson, ht. (I 990). Dia~cnc tic sources of Stable remaneoce is rrmagn.nrtirrdPalrtvoic natunic carboomcs: A rock nragnetic study. J . Grr,phys. Rr.5 95, 2753-2761, Jackson, M.. Rochette. P.. Fillion. G.. Ranerjec. S.. and Marvin. J. (1993). Rock magnetirn~ oiremagnelirrdPaleo7c~iccarbonates: Low-temperature hehavioralldsuscepribiiity charoclrri~tics.I. Gpophyx. RF?. YY, 6217 6225. Jau,hs. J. A. (1994). "Reversals of the Earth's Magnetic F i r l l " Cumhridgr Unir PTCEE, Cambridge, UK. Johnson. E , A.. Murphy, T.. ilndTurrrnson. 0 . W . (19481. Prehistory of theEarth3 magnetic field. Terr Mugn. Arrnos. Elcnr 53.349-372. Johnson, G . D.. Johnson. N. M.. Opdyke. N. D.. and Tahirkheli. R. A. K. (1979). Magnctic rcvcrsnl stratigraphy and sedimentary tectonic hisl~ayai the uppersiwalik Group castcrn Salt Range and southweslem Kashmir. In "Geadynamics of Pakistan-' (A. Farah and K. A. DeJong. eda.). pp. 149-165. Gmlo~icalSun,cy of Pakistan. Uuetta. Johnson. G.D.,Opdyke,U.D..Tandon, S. K.,utdNanda.A. C. 11983').Themapeticpfrlarity slraligraphy of thc Siwdik Gmup at Haritalyangar (India) add a new last ilppearancc ~~

~~

datum for Ramapithecus and Sivapithecusin Asia. Palaeogeogr, Palaeoclimorol., Pulaeu ecnl 44,223-249. Johnson, H. P.. Lowrie, W., m d Kent. D. V. (1975). Stability of anhysteretic rrmanent magnetization in fine and coarse magoetile and maghcmite patticlcs. Grophys. J . R. Arrron. Sor. 41, 1-10. Johnson. N. M.. Opdgke. N. D., and Lindsay, E.H. (1975). Magnctic polarity stratigraphy of Pliocem: Pliestocenc terrestrial deposits and vertebrate faunas San Pedro Valley, Arizona. Geul. Soc Am. Bull 86,5-12. Johnson, N. M., Opdykc. X, D.. Johson. D. G.: Lindsay, E. H.. and Tahirkhrli, R. A. K. (19U2). Magnctic polarity stratigraphy and ages of Siutalik Group rocks ol the Potwar Plateau, Pakistan. Pulaeo~cogr,Paluer,elimnro/., Polarorrol 37, 17-42. Johnsun. N. M., Oficer, C. B.: Opdpkc. N. D.. Waodard. G. D., Zeitler. P. K,, and Lindsay. E. H. (1983) Rates o l Cenozoic fectcmism in the VaUecit<,-Fish Creek Basin, western Imperial Valley, California. Geology 11,664-667. J0hnson.N. M.. Stiu.J.,Tauxe.L..Cerveny.P.F., andTahirke1i.R. A. K. (1985).Palcamagndic chronology, Ruvial proceases and tectonjc implications of the Sirvnlik deposits "car Chinji Village, Pakistan. J. (icr~l.93, 27-40. Johnson. N. M., Sheikh, K. A , Dawson-Saundm. E., and McRae, L. E. [19KX). 'The usc o l ma~ncticreversaltime lines in stratigraphic analysis: Acasc study in measuring rwiahility in scdirnmntatiot~ratcs. In "New Perspcctivrs in Barin Analysis" (K. L. Kleinspchn and C. Paula. eds.), pp. 189-200. Springcr-Verlag. New York. Johnson. R. G. (1982). Brunhcs-Matuyama magnetic reversal dated at 790.OW yt B. P. by marine-aslronomical correlations. J. Qoat Re.%.17, 135-147. Jordah T. E., Rutty, P. M., McRae. L. E.. Bccr, J. A,,Tabbutt, K.; and Damanti, J . F. (19W). Magnetic polarity stratigraphy of the Miuliscene Rio Azul section, Precordillera thrust bclt, Sdn Juan Province. Argentina. J. Geol 98, 519-539. Juirez, M. T.; Osete. M.L., McICnder. O., Langereir C. G.. and Zijderveld, J . U.A. (19%). Oxfardian magnctostraligfaphy offhcAquil6n andTosos seclicms (IberianKange. Spain) and evidence of a pre-Oligocene overprint. Phys. Enrlh Planer Inlur. 85. 195-211. Juirez. M.T., Osete,M. L., Melender. F.;and Lowric, W. (1995). Oxlordian magoefostraligrzphy in the Iberian Range. Grophys. Ras. Lull. 22,288!L2892. Kappclman. J.. Simons, E. I.., und S\vishcr, C. C.,111 (1992). New age determinariuns lor the Eoccnr-Oligocene. boundary stdimcnts in the Fayum Dcprrssiun. Xonhcm Eg)?pt. J. GeoL 100,647 -668. Kaflin. R. (1990a). Magnctic diagenesk in marina sediments from the Oregon continental 95,4405-4419. margin. J. Groph,vs, Karlin, K. (1990b). Magnetic mineral diagcnrsis in suboxic scdimmts at Bettis site W-N, RE Pacific Ocean. J. G ~ o p hRes. ~ ~ 95,4421-4436. . Karlin, R., irnd Levi, S. (1983). Diagcncsis of magnetic minerals in recent hcmipelagic sedimenls, Ara!arurc(London) 303, 327-330. Katlin, R., and Levi, S. (1985). Gcochrmical and sedimcntological control on the magnetic properties of hcmipelagic scdimenls. J Geophys. Re.) 90.10.373-10.3'32. Karlin, R.. Lyle. M.. and Hcsth, G.R. (1987). Authigcnic magnctitc fcrrmalian in suboxic marine scdimcnu. ~VnCr~re (London) 6,4YO-493. Kanai. N. (1984). Pdeomagnetic study of Lake Biaa sediments. In "Lake Biwa" (S. Horie, ed.), pp. 399-416. Dr. W. Junk Publ., Dordrechl, The Ncthrrlands. Keating, B. H., andHcrrero-Rcrveta,E. (1984). Magnetostratigraphy of Cretaccaus andEarly Ccno~rlicsediments of Deep Sea Drilling Project, Site 530, Angola Basin In InitiaJ Rcpopurlr of the Decp Sra Drilling Project" (W. W. Hay, J. C. Slbuct el 01.. eds.). Vol. 75, pp. 1211-1218. U.S. Govt. Printing Office. Washington. DC.

Keating, 8. H.. Hclslq, C. E.: andPassagno, E. A. (1975). IateCrctaceous rcverral sequence. G r o l o ~ y2,75-79. Keigwin. L. D.. Jr. (1979). Latc Ccnozoic stable isotope stratigraphy and paleoceanography of Deep Sea Drilling Project sites from the east equatorial and central North Pacific Ocean. Earrh Planel. Tci. i.ett. 4Si 361-3R2. Keller. M. H., Tahirkhrli, R. A . K.. Mirra. M. A.. lohnsr,n, F. D.. Jnhnson, N. M., and Opdykc. N. D. (1977). Magnetic polarity stratigraphy nl lhe upper Siralik deposits, Pabbi Hills, Pakistan. Garth Plnner. Sri Lea. 34 187-201. Kellogg, '1.B., Dupirsy, J . C.. and Shackleton, N. I. (1978). Planktonic loraminilrral and oxygcn isotupic slraligraphy and paleoclimatolo~yoi Norwegian Sea deep-sea cores. Boreal 7, 61 -73. Kennedy, W. J., Cohban, W. A , and Sooll: G. R. (1992). Ammttnire correlation oI the uppermost Campanian of western Eun~pr.the U.S. Gull Coast. Allantic seaboard and Western lntcrior, thc numerical age of rhc basc of the Maastrichdan. Get,/. Mug. 129, 497-500, Kennctt. I. P.. and Watkins. N. D. (1974). Larc Mioccnc-Early Pliocene paleomagnetic stratigraphy, paleoclimarology, and biastratigraphy in New Zcaland. GmL Suc Am. Bt~ll.85, 1385-1398. Kent, D. V. (1973). Post-depositional rcmancnt magnctization in deep sea sediment. ,\'Diere (London) 24632-34. Kent, D. V. (1979). Palcomagnctism ofthc Dcvonian Onondaga limrsrcme revisited. I, Geephys. Res. 3576-3588. Kent, D. V. (1982). Apparcnt mrrclstion of palcomagnctic intcnsily and climatic records in deep-sea Eedimcnts, Nanlre (London) 299,538-539. Kent, D. W,, and Gradstein, F. M.(1985). Crctaccous and Jurassic geochranology, Grol. So?. Am, Bull. 91, 1419-1427. Kent D. V.. and Lowlic. W. (1974). Origin of magnctic instability in sediment cores from thc central Nonh l'acific. J. Geopkys. Rex 79,2987-3WO. Kent, D. V., and Opdykc, N. D. (1977). Palcamagnctic ficld intenrily variation recordcd in a Brunhes epoch dccp sca scdiment corc. Nature [l.ondon) 2h6, 156-159. K,.nt, U. \ and Opd,L; \ . J I l l h i ! \I!tlt~;c,mp<~ntm~ rnngnelwalnlni frunt the hl311ch Chunk t.>rm;lll.,n C I I tn; ;rntrill 4ppal.i.'nlan\:ind lnor tuctonl. ~m,,l~cat.oor J C;n,pnl~ .. Res. YO, 5371-5384. Kent. D. V.. and Spariasu, D. J. (1983). High resolution magnetostratigraphy of Caribhcan Plio-Pleistocene scdimcnts, Paleogeogr., Paleoclinzar~,L,Palrvrcol. 42, 47-64. Kent. D. V.. Oiscn. P. E., and Witte. W. K. ( 1 9 5 ) . Late Triassic-Early Jurassic gcomagnctic polarity scqucncc from drill cores in the Newark rift basin, eastern North America. I . Geopltvs. Res. 100, 14,965-1499R. Kent. I, T. (1982). TheFishrr-Binghamdistributiunon asphere.J,R. Stnt Soc.. Se-l R44,71-80. Khan, M. I.,Kml. D. V., and Miller. K. G. (1985). Magnetostratigraphy of Oliogocenr ro Plcistocmr wdimenls of DSDP Leg 82. In "Initial Rcpons of thc Deep Sea Drilling Project" 82, (H. Bougault, S. C. Cande et anL, eds.), Vul. 82. pp. 385-392. U.S. Govt. Prinling Office. Washington! DC. Khan, M. J., Opdyke. Y. D.. and Tahirkhcli. R. A. K. (1988). Magnetic stratigraphy ol the Siwalik Group. Bhitanni, Manvat, and Khasor Ranger. notthwsstern Pakistan and the liming of Nsogene tectonics of the Trans Indns. J. Geophy.~.Rru 93, 11773-11790. Khramov, A. N. (1958). "Palaeomagnetism and Stratigraphic Correlation." Gostoptechjzdat. Leningrad (English translation hy A. J. h j k i n r . Published by Geaphys. Dept., A.N.U.. Canberra. 1960). Khramov, A. N. (1974). "Palromagnelism ill1 the Paleozoic." NEDRA, Moscow (in Russian).

.

Khramov. A. N.. and Rodionov, R. (1981). The geomagnetic field during Pilleu~uictime. In "Global Reconstruction and the Gcumagnetic Ficld During the Paleozoic'' (M. McElhinny n o / , eds.). Ad.. Earth Planet. Sci., pp. W-116.Ccntcr for Academic Publ..Tokyo. King, J. W., and Channeil, J. E. T. (1991). Szdimcniaq magnetism, mviro~in~entalmagnetism. and magnztisuatipaphy. In "(IS. National Kcpart to lnfemati<)ndUnion of Geodcsy and Geophysics." Rev. Gcophys., Suppl., pp. 358-3713, King. J. W , Bmerjce, S. K.. Marvin, J., and Ordenlir. O . (1VXZ). A comparisr?n of dilcerent magneticmethodsfor dctcrmining the relative grainsize of nlagnetite in naturalmincralr. some results lor lake sedinients. Eailrrh Planer Sci. Lett 59, 404-419. King. J . W., Bancrjcc, S. K., and Marvin. J. (1983). A new rock mawetic appruach lrl < r l r c f i n ~ sediments for geomagnetic intensity studies: Application to paleointensity for thc last 4W0 yea-. I. G r ~ p h y sRm. . BX, 5911-5q21. Kirschvink. J. L. (1978).Thc Precamhria~l-Cambrianboundary prohlcm: Ma~netostratigr.raphy of the Arnadeus basin, ccntral Australia. Geol. Mug. 115(2), 13Y.l5(l. Kirsch5ink. J. L. (1980). The least squares lincs and plane imalgsis ol palsomaanetic data. G ~ o.n h. y rI.. R. Asrro,,n. Suu. 62, 699-718. Kirschvink. J . L.. and 1.owcnstam. H. A . (1979). Mineralization and magnctization of chitun teeth: Paleomaenctic, scdimcntoloric. of oreanic maenctitc. - and hioloiu implications Eurrh PltmeL Sci Lcrt 44, I 9 3 204. Kirschvink,I. L.. and Ruzanov. A. Y. (1984). Magnetostratigraphy of lower Camhrian slradta from the Siberian Plathrrm: A pillrr~ma~netic pole and a polarity timc-scalc. Geol. Mag. U1, 189-203. Kirschvink. J. L.. Magaritr. M., Ripperdan. R. L.. Zhnravler. A. Y., nnd R ~ i ~ m o A. v . Y.

-

~ i d a yl(4). 69-91. Kliefield. R,. andChanne11. I. E.T, 11981) Widcs~rcadrcmaenctirationr,l Helvetic limeltones. .'I. Geophy.~. Rrs. 86, 1888-1k0. Klirgord. K. D., Hrmsl;r. S. P., Mudie, J. D., arld Parker. R. L. (1975). An analysis of nearbottom magnctic anomalies: Seil floclr spreading and the magneti~edlayer. Gaoptry.r,J. R. Anrun. Soc. 43. 387-424. Kobayashi, K., and Komura. M. (1972). Iron sulfides in the sediment corcs from thr S*a 111 Japan and thcir gcuphysieal implicallorrs. Eunh Pl~~ner. Sir. Lerr. 16, 2W-208. KocnigrhrrHer. J. G. (1938). Katurnl residual mapnelhm r r l eruplive mcks, Parts I and 11. Tcrr Magn. Armt,.~. E l ~ c t r43, 114-127. 299-320. Kur;lcheva, M. (1983). Archromagnctic data for Bulgaria and south cartcrn Yugr,sla\ia. In '.GrU~LTOPSin southwestern Iceland. Genphvf. Re.?. LEII.7 . 337-340. Krocpnick. R. B., Denistm, R. F.. and Dahl. U. A. (1988). The Cenozoic sea warcr "'SriWSr curvc. data rcvlcw and implications for correlation of marine strata. Polroce~nography 3,743-756.

302

Bibliography

Krumsick. K., and Hahn, G. G. (1989). Magnetostratigraphy near the Cretaceous: Tertiary boundary at Aix-en-Provencr (southern France). Cah. R m r v e Genl Hnure-Provenc*. Dj~rle1, 13-50. Kukla, G.. and An,Z. S. (11)89j.Loesssmatigraphy in centralChina. Palaeogeogr, Palart,clima. mL. Polneovrul. 72, 2203-2225. Kwkla. 0.. Hcllcr. F.,Liu, X. M.. Xu. T. C., Liu, 'T.S.. and An. Z . S. (1988). Plristocenc climates in China dated by magnetic susceptibility. Geology 16, Rll-816. LaBrecquc. J. L.: Kml, D. V., and Cande, S. C. (1977). Revised magnctic polarity time-scale for the Latc Crctaccous and Ccnozoic timc. Geologjr 5,330-335. LaBrwque. 1. L., Hsu, K. I.. Carmen. M. F.. 11..Karpoff. A . M., McKmac, J. A,, Percival, S. F., Jr.. Pctcrscn, N. P.. Piscotto, K. A., Schreihcr, F., Tauxc,L., Tuckcr. P., Weishrrt, K J.. and Wright. R. (1983). DSDP Leg 73: Contributions to Paleogene stratigraphy in nommclalure, chrooido~yand sedimentatinn rates. Puler,pruugr., Pulr~~~;limornl., Palroecot. 42, 91-125. Laj. C.. Nurdemann. D.. and Pomeau, Y. (IY7Y). Correlation function analysis of geomagnetic field rrvrr$alr. 1. Groph>'.v.Ru.7 84,45114515, Laj. C.. Mazaud, A. Wccks, R.. Fullcr. M.. and Herrero-Acrvcra. F,. (1991). Geomagnetic reversal paths. Narure (London) 351, 447. Lsj. C.. Mazuud. A. Weeks, R.. Fuller. M., and Henero-Bervero. E. (1992). Statls6cal assi.53men1 o f the preferred l<mpitudinalhand, ior rrcenl geomagnetic rcversal rccords. G w phys. Re$. L.ett, 19, ?W3-2006. Langereis. C. ti.. a i d Hilgen. F. J. (IYYI). The Rossello composite: A Mediterranean and global reference section for the Early to early Lalz Pliocene. Earth Phner. Sri. I,crt. 104,211-225. Lange~ris;C.G., Zachariassc. W. J..sndZijdcwcld. J. D. A. (1984). LateMiocene magmtobiostratigraphy of Crcfc. Mar. Micropalrontol 8, 265-281. Lallgereis C. G.. Sen. S.. Sumengen, M.; and Unay, E. (1989). Preliminary magnetustratigraphic results of rome Netlgenr mammal lr>calitirzfrom Anatolia (Turkey). In "NA'l'O Advanced Study Wurkshop in Eumpcan Neagcnc Mammal Chronology" (E. Lindsay, P. Mein, and V. Fahlbush, eds.), NATO Ser. A. pp. 495-505. NATO. Dordrecht. The Netherlands. Larson, E. F.. Walker. T. R., Patterson. P. E,, Hohlitt. R, P., and Koscnbaum, J. ti. (1982). . ' Paleomagnetbm of the Moenkopi formation. Colorado Plateau: Basis for long-term model of acquisition of chemical remanent mametization in red beds. J Gronhv.~.Rer " 87, 1081-tli6. Larson, H. L.. and Chasc, C 0.(1972). late Mcsozoicevolution of thc wcstcrn WciticOcean. Grol Soc, Am. Bull 83,36273644, Larron. R. L.. and Hilde. T. W. C. (1975). A revised time scale of magnetic reversals for Lhe Early Cretaceous and Lale lurassic. J. Ger>phy.>.Re.?.80, 2586-2594. Lanon. R. L.. and Olson, P. (1991). Mantle plumes contml magnetic revcrsal 6cqucncy. Eanlz Planer Sci Letr. 107,437-447. Laraon. R. L., and Pitman, W. C., 111 (1972). World wide correlation of Mesozoic magnetic anomalies and its implications. Geol Soc. Am. BUN 83, 3645-3662. Larson. R. L.. Fischer, A. G., Erba. E.. and Premoli-Silva, I. (1993). "Apticnrr-Alhic~re," A Workshop R e p ) n on Glohal Eumts and Rhythms of the Mid-Cretaccaus, 4-9 October 1992. Perugia, ltaly. La7arenk0, A. A, tlolikhovrkaya. N. S., and Semenor. V. V. (IYXi). An attempt at a detailed stratigraphic subdivision of the loess association of the Tashkan region. Inr. Grot. Rev. 23. 1335-13M1.

.

~

Ledbetter, M. (1983). Magnetostratigraphy of Middlc-Upper Miocene and Uppcr Eacene sections in hole 512. In "Initial Reports of thc Deep Sca Diving Pmjen" (W. J. Ludwig, V. A, Krashcninnikav era[, t.ds.), Vol. 71. U.S. Grwt. Printing Oflicc, Warhinglon.DC. Lelrman, B.. Laj. C.. Kissrl: C., Maraud, A,, Psterne, M., afid LabeNic, I.. (1996). Relative changes of thc gromagnrtic field intensity during lhe last 280 kycar from pislon cores in the Acores arca. Phys. Eanh PlrmeI. Imcr. 93, 269-284. L c r h c h o s J. F.. Evans. M. E.. and Raads~uard,H. (1979) Ma~nelostratigraphy.bioltratigrdphy and grLake geomagnetic excursion. J . Geophys. Rm. 84,261-271. Liddicoat. J. C., Opdykr, N. D.. and Smith. ti. 1. (1980). P~~t.nmagTI&cpolmity in a 930m core from Searles Vallcy California. Nulare (London) %6,22-25. Liencn, B. R., and Helsley, C. E. (1980). Ma~nrtorlraui_maphyof the Moenkopi Fo'ormation at Bean Ean. Utah. 1.Geuphys. Rex U5. 1474-1480. Lindsay, E. H. (IYYU). Scdimcnts. grtrmorphology, magnetosrratigmphp,and vertebratepaleontology in the San Pedro Vallcy. Arizona. J. Grol. 9 8 6U5-619. Lindsay, E. H. (1995). Copemys and thc BarstovianlHemingfordian boundary. J V ~ n c h r Puteonrol. (in press). Lindsag, E. EL. Opdyke. 3 . 0.. and Johnson, N. M. (1980). Plioccne dispersal of lhe horse Equrcs and late Cenozoic mammdian dispersal events, &rum (London) 287, 135-138. Lindsay, E. H.. Opdykr. N. D..atrdJohnson. N. M. (1991). Blancan-Hemphillian landmammal ages and late Cenozoic mammal dispersal cvents Annrd. Rev. Earrh Planer. Sci. U , 445-488. 1.inknvn.T. 1.11965). '.Some Kcsults r~tPaleomagneticSfudyof Arctic Ocean RoorScdimenus"

Canada, 1966). Liu, T. S.. An. Z. S.. Yuan. A. Y.,and Han. J. M. (1985). The lo~r-paleosolsequence in China and climatic history. Epi.vo,dus &21-28. Liu, X. M.,Liu, T. S., Xu. T. C., Liu. C , and Chmg. M. Y. (1988). The pcirnsry study oh magnctastratigraphy of a loess profile in Xifeng area Gansu province. Gcophys I. R. Adron. Soc. 92, 345-348.

307

Bibliography

Mackereth. F. J . H. (1971). On the variation in direction of the horkonial component of remanent rrlagnctiration in lakc sediments. E#r(lt PIanrI. Sci. Len 12,332-338. Madrid. V. 41.. Stuart. R. M., and Verosub. K. L. (1986). Magnetostratigraphy of the late Neogene Purisima formation. Santa C r w County. California. Eanh Platter. Sci. 1,err 79,431-440. Maenaka. K. (1983). Magnctostratigraphic study of thc Osaka Group, with special refcrencr to the existence ofpre and post-la ram ill<^ episodes in theLake Matuyamapolarity epoch. Mem. Hann:onn Univ. 14 1-65. Magnur. G.. and Opdyke, N. D. (1931). A paleomagnctic investigation nf the Mintum forma. tiun, Colorado: A study in tstahhhing the timing OF remancnce acquisition. Tt-cronophysics Il.181-189. Maher, R. A. (1986). Characterization of suils by mincrdl magnctic measurements. Phyr. tbrrh Pldner. /~nrer.42, 76-92. Maher, B. A. (1988). Magnetic prnprrties of sumesynthctic sub-micron magnetites. Ceophys. I, Inr 94, 83-96, Maher. B . A,, and Taylor, R. M. (1'188). Formation of ultraline-grainod magnetite in sails. Nature (London) 036, 368-370. Maher. B. A., and Thompson, R. (19911. Mineral magnetic record of Chinese locss and paleosols. Geology 19,3-6. Maher. B. A,. undThomps<>n.R. (1992).Paleoclimahc significanceof mincrdl magneticrec~rd of thc Chinese locss and palcosuls. Quar. Rn.. 37, 155-170. Msher, B. A,, Thompson, R. and Zhrru, L. P. (1994). Spatial and temporal reconstructians nf Earrh Planer -- channrs .. -" in the Asian nalacomunsaon: Anew mineral mametic aDDrrlach. Sci LER125,461171. Mahoney. J. I., Storey, M., Duncan, R., Spencer, K. I., and Pringle, M. (1993). Geochcmirtry and age of the Ontong-lava Plateau. In "Monograph nn the Mcsurtic Pacific" (M. S. Pringlo, W W Sager. W.V. Slifitrr.and S. Stcin. eds.). pp. 233-261. Am. Geophys. CJnion. Washington. DC. Mankinen, E. A,, and Dalrymplc. G. B. (199). Revised geomagnetic polarity timc scale for the interval 0-5 my.b.p. I. Grophys. Kes. 84,615-626. Mimkinen, E. A,, and FrommC, C. S. (1982). Paleomagnctic data from the Cosa Range. California and current status of the Cobh Mtruntain normal geomagne6c polarity cvent. Gcophyr Res. Left. 9,1239-1282Mankinen, E. A.. Donnelly. J. M., and Gromme. C. S. (1978) Ceomagnedc polarity event recordcd at 1.1 m.y.b.p, on CohhMountain ClearLakc volcanicficld, California. G ~ d o p 6, 653-656. Marshall. L. G . , Butler, R. F.. Drake. R. E.. Curtis. Ci, H.. and 'L'edford. R. H. (2979). Calibration of thc Grcat Amrrican lnrcrchaoge. Scierrcr 205, 272-279. Marshall, L. G..Butlrr. R. F., Urakr,R. E.. and Curtis, G.H. (1982). Grochronology nfTypc Uguiiln (Latc Cenozoic) land mammal age, Argentina. Scimce 216,986-989. Marshall. L. G.,Drake. R. E., Curtis, 0. H.. Butler, R. F..Flanagan. K. M., and Naescr. C. W. (1986). Grochronology of type Santacrucian (middle tertiary) land mammal age. Patagonia. Atgentina. I. Gerrl, 94,449-457, Martinson. D. G., Pisias. N. G.,Hays, I. D.. Imhric, J., Moore. T. C.. Jr,, and Shacklctan. N. 1.(1987). Age dating and thc nrhital theory oi ihe Lcc Ages: Development of a highresolution O lo 30U.UM-year chronc,stratigraphg. Quar. Re\. 27, 1-29, Minon. E. (1982). Latc JurassiclEarly Cretaceous magnctic stratigraphy from thc Sumeg section,Hungw. Enrth Planet. Sci. Lcrr. 57, 182-190. Mhtan, E., MSrton. P.. and HeUcr. F. (1980). Remanenf magnetization 01 a Plicnsbachian limcstone sequence at Bakcnycaernye (Hungary). Earth Planet Sci Len a,21R-226. ~~

~~

Mary,C,..Moreau, M., Orue-Etxebarria, X.,Estibaliz, A,. and Courtillot, V. (1991).Biosfratigrapby and magnetostratigraphy of the Cretaccodertiary Sopelam section (Basque country). Earth Planer. Sci. Len 106,133-150. Mary, C.. Iaccarino, S., Courtillot, V., Besse, I., and AissaauiD. M. (1993) Magnetostratigraphy of Pliocene sediments from the Stirone River. (P. V.alley). Cn9phys. J. Int. 11% 359-380. Matuyama, M. (1929). On the direction of rnignetizatic~nof basalts in Japan, Tyosen. and Manchuria. Proc. Imp. Acad (Tokyo) 5,203. Mauritsch, H.J., and Turner, P, (1975). The identification of magnetite in limestones using the low temperature transition. Earth Planer. Sci Lett. 24,414-418. Mazaud, A., Laj, C.. and Bender, M. (1994). A geomagnetic chronology lor antarctic ice accumulation. Geophys. Rer. Lett. 21, 337-340. McCabc, C., and Channell, J. E. T. (1994). Lats Paleozoicremaffletization in limesfones ofihe Craven Basin (northern Ebgland) and the rock magnetic fingerprint of remagnetization secondary carbonates. J Grr~phys.Kes. 99,4603-4612. McCabe. C., Van der Vun, R.,Peacor. O. R.,Scotesa. C. R.. andFreeman.h. (1983). Diagcnctic magnetite carries ancienl yet secondary remanence in somo Paleozoic sedimentary carbonates. L;LOIOB~ 11,221-223. McDougall, I., and Chamalaun. F. H. (1966). Geomagnetic polarity Xala of time. Nnlvrr (London) 2U,1415-1418. McDougall, I.. and Tarling. 5.H. (1963a). Dating of rcucrsak of thc Earlh's magnetic fields. Nature (London) 198,1012-1013. McDougall. I.. and Tarhg. D. H. (1963h). Dating of polarity zones in the Hawaiian Islands. Nature (Londrm) 200,54-56. McDougall. I.,and Tarling, D. H. (1964j.Dating geomagneticpolarity zunes. N a i u r ~(London) 202, 171-172. McDougd. I., and Wensink. 1. (19661. Paleornagnetism and geochranology of the PlioccnePleistocene lnvas in Iceland. Enrih Planet. Sri. Lett 1,232-236. McDougall, I.. Allsop, H. 1..and Charnalaun, F. H. (196h). Isotopic dating of the New Volcanic series of Victoria, Australia and gcomagnetic pularity epochs. J. Geophys. Res. 11,61074118. McDougalL I.,Saemundson, K.. Suhannesson, H., Watkinr, N. D.. and Kristjansson, L. (1977). Extension of the gcomagnetic polarity time scale to 6.5 my.: K-Ar dating, geologicdl and paleomagnrtic study of a 3500 M lava succession in wcstrm Iceland. Bull. Geol Soc. Am. 88, 1-15. McDoUgall, I., Kristjanson, L.,and Sarmundson, K (1984) Magnetostratigraphy and geochronology of northwc~tIcejand. J. Geophys. Rex 8% 7029-7064. McDougall. I., Brown. F. H., Cerling. T. E., and Hillhousc, J. W. ('IYYZ). A rzapplaisal of thc eeomaenetic polarity time scale to 4 Ma using data from TurkanaBasin, East Africa. Geophys. KCE.Lerr 19,2349-2352. McElhhy, M. W. (1964). Statistical sipificancc of the fold test in paleornagnetism. Geophys. I R Astron. Soc 8,338-340. McElhinnv. "Paleomaenetiam and Plale Tectonics." Camhrid~eUniv. Press. , M. W. (1973). . Cambridge. UK. McElhiinny, M. W,, and Opdykc,N. D. (1973) Tke rcrnappetizatinn h p t h c s i s dixountcd, a nalcomagzetic study of thc Trenton limestune (bcw York State). Geol Sor. Ant. Bull. s6.3697-3708. McElhinny. M. W., and Scnanayakr W. E. (1982). Variations in thc geomagnetic &pole. 1. The past 50,000 years. J. Geomngn. Geoelectr. 34, 39-51.

.

Molina-Gatza, R. S.. Van der Voo. R., anrt Geissman. J. W. (1989). Paleomagnctism r,f the Dewey Lake Formation northwest Texas: end of the Kiaman superchrrm in North America. J . Geophys. Res. 94, 17,8111-17.888. Molina-Gwra, R. S., Geissman, J. W..Van der Voo. R.,Lucas. S. 0..andHaydm, S. H. (1991). Palcomagnelismofthe MoenkopiandChinle F m in ccntralNcw Mcxico. Implicationsfor the North Amcrican apparent polar wander path and Triassic magnetostratigraphy. J. Cet,phys. Res. 96, 14239-14262. Molostor.sk!: E. A. (1992). Paleomagnetic stratigraphy of thc Permian rystcm. In. Geol Rcu. 34, IW1-1W7. Molyncux, L.. Thompson, R., Oldfield. F., and McCallan, M. E. (1972). Rapid measurement of the rcmanrnt magnetization of long cores of sediment. Nature (London) 237,42-41. Monechi, S., and Thierstcin. H. R. (1985). Late Cretaceous-Eocene nannofassil and magnetos t r a t i ~ p h i ecorrelations near Guhbio, Italy. Mar Micmnuleonrol. 9, 419-440. Monechi. S., ~ l c i lC.. ; and Backman, J. (1985). ~agnetohiochronolog~ of Late Cretaceous. Paleasene and Late Cenozoic . ~e~a$Lic - scdimenlarv sequences from the nurlhwerl Pacific. DSDP Leg 86, Site 577. In "Initial Rcparts of Deep Sea Drilling Project" (G. R. Heath. L. H. Burklr rr el., eds.), Vol. 86. pp. 787-797. U S . Govt. Printing Office, Washgton, DC. Montanari, A , Deino, A. L.. Drake. R. E.,Turrin, B. D., DePaolo, D. J., Odin. G.S.,Curtis, G. H., Alvarer. W., and Bice. D . (19881. Radioisotopic &ilting of the Eoccnc-Oligocene hound3t! cn Ih. pcl:~y; .eq.kn.cr elf the 1~3r~llca~rcrn Apcnninr.. I,, l h c tocinc. Oliruccne l30und.m ~nrhc \larchc-t Imt>ri:g il.s,tn I. I l d v..l 1.I . P ~ c ~ c I I I - SRI I( ~' AJ' . ~ L . and A. Montanari, eds.), Intl. 5ubcomm.Paleogcnr Stratigraphy Rcport. Eu~cnrJOlipocene Boundary Meet., Ancona, pp. 195-208. Morner, N A,, Lanser. J. P., and Hosprrs. 3. (1971). Late Wcichselian paleomagnetic rewrsal. Noture (London) 234, 173-174. Moskowitz. B. M.. Frankel. R. B., Flanders, P. 1.. Blakemore, R. P., and Schwartz, 8.B.

thr

'

.

.

the drtecbon of biogenic magnctitc. Earth Plrrnel. Sci. LelL UO,283-3W. Mothersill, J. S. (1979).7hepaleomagneticrecord of thc lata QuaternarysedimcntsofThundrr Bay. Con. J. Earth Sci. 16, 1016-1023. Mothersill. J. S. (1981). Late Quaternary paleomagnetic record of Godarich Basin, Lake Huron. Con. J ErrrIh Sci. 18.448-456. Mullender. T. A. T., viln Vehen. A. I, and Dekkcrs. M. J. r1993). Continuous drift correction and separate idcntficatiun of the fenimaenetic - and paramaenetic contributions in thcrmomagnetic runs. Geophys. J . Int 114,663-672. Mullins, C. E. (1977). Magdeticsuseeptibility of the soil and its sign~canccin sail science-a reviow. J. Soil Sci 28,233-246. Muttoni, G . , and Kent, D. V. (1994). Paleomagnetism of Latrsl Anisian (Middle Triasric) sections of the Prcva Limestune and the Uuchenstcin fomatirln, Soulhem Alps; It@. Eurlh Planer Sci. Lm U2,1-18. Channell, J. E. T.. Nicora. A , and Rettori, R. (1994). Magnotostratigraphy and Muttoni. G., biostratigraphy of an Anisian-Ladinian (Middle Triassic boundary section from Hydra, Greecc). Poleogeogr., PnlrrroclirnaroL, Palneoecol,, 111,249-262. Muttoni, G,.Kent, D. V., and Gaetani, M. (1995). Magnetostratigraphy of a candidate global stratotypesectionandpoint forthe baseof the AnisianfromChios (Greece). Implications for the Early to Middle Tdassic geomagnetic polarity sequence. Phys. Eonh Plonet Inter 92, 245-260.

Nabcl, P. E. A., and Valencio, D.k (19M). La magneto3tratigraphihia dcl ensenadense dc Is dudad de Buenos Aifes, su significadogeologicas.Asoc. Gcol Argenr. Rev. 36(1), 7 S 7 8 . Nagata, T. (1952). Reverse thermal-remanent magnetism. Nature (London) 169, 704. Nagata. 1'.(1961). "Rock Magnetism." Maruzcn. Tokyo. Nagata, T , Uyeda, S..and Ozima, M. (1957). Magnetic interaction belween ferromagnetic minerals contained in rocks. Philos. Mag., Suppl. Adv. Phys. 6,264-287. Nagata, T..Arai, Y.. and Momose. K. (1963). Secular variationof the geomagnetic lolal force during the last 5000 years. I (irophys. Rex 68,5277-5282. Nagata, T., Kohayashi, K.. and Fuller. M. (1964). Identification of magnetitc and hematite in rocks by magnetic obxwation at Ion, temperaturel Geophy~.Res. 69,2111-2120. Nagy. E. A,. and Valet, 1. P. (1993). New advances for palcomagnetic studies o f sediment carer using Uihannels. Ceophys, Res. Let!. 20, 671-674. Nnkanishi, M., Tamaki, K., and Kohayarhi, K. (1989). Mesozoic magnetic anomaly lineations and seafloor spreading history o f the northwestern Pacific, J . Geophys. Res. 9 4 1543715462. Nakanishi, M.; Tamaki, K., and Kobayashi, K. (1992). Magnetic anomaly lineations from Late Jurarsic to Early Cretaceous in the west-ccntral Pacitic Ocean. Gcophys. I. Inr 109,701-719. Napolcone, G.. and Ripepr, M. (1YW). Cyclic geomagnetic changes in Mid-Cretaceous rhythmites, Italy. Terra Novrr 1,437-442. Napoleone. G.,Premoli-Silva. F.. HeUer, F.. Cheli, P.. Corezri, S., and Fischer, A. G . (1983). Eocene maenetic stratiwa~hv . . . at Guhhio. Italv. and its implications fur Paleogenc gerv" chronology. Gml. Sor. Am. Bell. 94,181-141. Needham, J. (1962). "Science and Civilization in China." Vol. 4, Part 1. Cambridge Univ.. Cambridge. UK. Nee1.L. (1951) L'inversion dr L'aimentationpemanentc desroches, Ann. Cier,phys. 7,YO-102. ~, Negrini, R. M., Davis. J. 0.. and Verosub. K. L. (1984). Mono Lake geomagnetic excursion found at Summer Lakc, Oregon. C;eology U,643-646. Negrini. R. M., Verosub, K. I,., and Davis, J. 0. (1987). Long-term nongcucrntric axial dipole dircctions and a geomagnetic excursion from the middle Pleistoccnr, sediments of the Humboldt River canyon,Prrshing County. Nevada. J Cieophys. Res. 92,10,617-10,627. Negini, R. M.. Erbes. D. B., Roherls,A. P.,Vcrosub,K. L., Sarna-Wojcicki. A.M., audMcyer. C. E. (1994). Repeating waveform initiated hy a 18C-190 ka geomagnelic excursion in westcrn North America: Implicarions for field behavior during polarity transitions and subscquent secular variation. J. Gtophys. RPI 99,24,105-24.119. Ness, G.. Lcvi. S., and Couch, R. (1981). Marinc magnetic anomaly time scale for theCenozoic and Latr Cretaceous: A prkis, critique, and synthesis. Rev. Geuphyv. Space Phys. 18, 753-770. Neville. C., Opdyke, N. D., Lindsay, E. H., and Johnson.M. N. (IYIY). Magnetic stratigraphy of PLiuc~nedeposits of the G l e m Fcrry fc?rmation,Idaho, and its implicationsf01 North American mammalian hiostratigraphy. Am. 3. Sci. 279, 503-526. Yick. K..Xia, X..and Elmtat, R. D. (1991). Paleomagnetic and pehopaphic evidence lor carly magnetization> in succcssivr lerra rosa palcosols. Lowcr Pennsylvania Black Price Limestone. Arizona. J . Ceophys. Rex. 96,9873-9885. Ninkuvich,D..Updykr,N. D..Heven, B. C.. and Foster, J. H. (19%). Paleomagneticstratigraphy. rates of deposition and trphrachronologj in North Pacific deep-sea sediments. Eartlr Plunet. Sci. I.ett. 1,476-492. Nucchi. M., Parisi, G.. Monam. P., Moncchi, S., Mandile, .q.. Napoleone, G.. Ripepc. M., Orlando, M., Premoli-Silva, I., and Bricr, D. M. (19%). The EoCenc-Oligocene baundary in the IImbrian pelagic regression. In "Terminal Eucme Events; Developments in

Palaruntology and Stratigraphy" (C. Pomerol and I. Premoli-Silva, eds.), Vol. 9, pp. 25-40. Elsevier, Amsterdam. Obradavich, J. D. (1993). A Cretaceous lime scale. Spec Pup.-Geol Assoc. Con. 39,379-396. Obradovich. J. D.. and Cobban, W. A. (1975). A time-scale for the Late Cretaceous of the wcstcrn interior of North America. Spec. Pap.-Crol Airor. Cm. 13, 31-54. Ohradovich. J. D.. Sutter. J. F., and Kunk, M. J. (1986). Magnetic polarif). chron tic points for the Cretaceous and Early Tertiary. Terra Cognita 6, 140. Oda, H.. and Shibuya.H. (19%).Deconvolutionoflong-~orepaleoma@etic data of the Ocean Drilling Program by Akaike'sBayesian information Criterion Minimintion. J. Geophy~, Res. 101,2815-2834. Odin. G. S.. Curry. D. S.. Gale. N. H.; and Kennedy, W. J. (1982). The Ph.merozic time x a l e in 1981. In "Numerical Dating in Stratigraphy" (G. S. Odin, ed.), pp. 957-960. Wilcy. New York. Odin. G. S.. Montanari. A : Deino, A., Drake. R., Guise, P. G.. Krruzer, H., and Rex, D. C. (1991). Reliability of volcanic-sedimentary biolite ages amow thc Eocene-Oligocene boundary (Apennincu. Italy). Chem. Geol,. Is01 Get,.sci. Sect. 86,203-224. Ogg. J. G. (198R). Early Crctaccous and Tithonian magnetostratigraphy of thc Galicia Margin (ODP Leg 103). In "Proceedings of the Ocean Drilling Program, Scientific Results" (0. Boillot, E. L. Wintcrcr et al., eds.), Vol. 103, pp. 659-681. ODP, Collcgc Station, Tx. Ogg. 1.C.,and Luwie. W. (1986). Magnetosttatigraphy n l the Jurassic-Cretaceous boundary. G r ~ l r , ~4,y 547-550. Ogg. 1.G.. and Stcincr, M. B. (1991). Early Triassic magnetic pnlarily time scale-integratioo of magnetostratigraphy. a m o n i t e zonations and sequence stratigraphy from stratotyp rectluns (Canadian Arctic Archipelaga). Eanh Planet. Sci. L m lW,69-89. Ogg, J. G., Stcincr. M. 8..Oloriz. F.. andTavera. J.M. (1984). lurusicmagnctosUatigraphy. 1. Kimmeridgian-Tithonian ofsierra Gorda and Carcahuey.southern Spain.Earth Planel. sci. Lett 7 1 147-162. Ogg, J. G.. Srcincr:M. B., Company. M., and Tavera, I. M. (1988). Magnetostratigraphy across thc Bcrriasian-Valanginian stage boundary (Early Cretaceous), at Cehegin (Murcia Rovincc, southern Spain). Earth Planer. Sci. Lett. RI, 205-215. Ogg.J.G., Stcincr,M.B., Wiczorck,J..and Hoftman,M.(lYYla). Jurassicmagnetostratigraphy. 4. Early Callovian through Middle Oxfordian of the K~akowUplands (Poland). Eorrh Planer. Sci. I.elt. 104, 488-504. 0%.1.G., Kudama, K.. and WaUick. B. P. (1991b). Lower Cretaceous magnetoatraligraphy and palrolatitude off northwest Australia, ODP Site 765 and DSDP Site 261. Argo abyssal p l i n , and ODP Sitc 766. Gascoyne abyssal plain. In '.Proceedings of the Ocean Drilling Prujed, Scientific Results" (F. M. Gradstein, J. N. Ludden er aL, eds.), Vol. 123. pp. 523-554. ODP, Collcgc Station. TX. Ogg, J. G.. Hasenyager. R. W.. Wimbledon. W. A., Channrll, J. E. T..and Bralower, T. J. (1991~).Magnetnstratigrnphy of the Jurassic-Cretaceous boundary interval-Tethyan and English faunal realms. Cretaceous Kc$. l2.455-482. Oldheld. F.. Dearing, I. A., and Rattarbcc, A. W. (1983). New appn,aches to roccnt environmental change. Gmgr J. 149, 167-181. Olsen. P. E. (1980). The latest Triasic and Early JurasSic formations 01. the Nrward hasin ui easrcrn North America, (Newark Supergroup): Stratigraphy, structure and setlimmtalion. N.J. Acad. Sci Bull. 25,25-51. Olaan, P. (1983). Geomagnetic polarity reverrals in a turhulcnt core. Phyr Eanh Pinnet. Inter, 33,260-274. Olson; R. K.. and Poore. R. 2.(1990). Eocene-Oligocene sea-lcvcl changes on the New Jersey caaslsl plain linkcd to the deep-searecord. Grol SK. Am. Bull 102,331-339.

Omarzi, S. K.. Coe, R. S. and Barron. J. A. (1993). Magnetonratigraphy: A powerful tool for high resolution age-dating and correlation in the Miocene Monterey Formation of California: Resultsfrom Shell Beach Section, Pismo Basin.ln "Applications ofPaleomagnetism tu Sedimentary Geology" (D. J. Aissaoni, D.F. McNeill, and N. F. Hurlcy, eds.), SEPM Society for Sedimentary Geology Spec. Publ. No. 49. pp. 95-111. Onstou, T. C. (1980). Application of the Bingham distribution function in paleomagnctic studies. J. Geophys, Rev. 85, 1500-1510. Opdykc, N. D. (1968). The paleomagnetism of oceanic cores. In "History of the Earth's Crut" [R.A. Phimey. cd.). pp. 67-72. Princeton llniv. Prcss, Princeton, NJ. Opdvke. N. D. (1972). Paleumagnetism of deep sea cores. Rev. Geophys. Space Phys. 10, . . 213-249. 0pdyke.N.D. (1990). O).agagneticstratigraphy of Cenozoic terrestrialsediments and mammalian dispersal. I . GeoL 98, 621-637. 0pdyke.N.D.. andDiVenerc.V.1. (1994). Magneticpolanty stratigraphy oftheMauchChunk Formation Pennsylvania). EOS, Truns. An,. Geophys. L'nion, Spring Meet. 75, 130. Opdyke, N. D.. and Foster. J. H. (1970). The palcarnagnetism of cores from theNorth Pacific, geological invcstigatim of the North Pacific. Mem.-Geol Soc. Am. 126, 83-119. nndvke. . . Thc paleomagnctism of sedimenl cores from the Indian - - ~ .N. D.. and Glass. B. (1969). O e a n . naeP-Sca Res. 16,249-261. Oodvke. N. D.. and Henry, K W. (1969). A test o f the dipole hypothesis. Eu'urrh Planet Sci. Len 6, 139-151 Opdyke, N. D.. and Runcorn, S. K. (1956). New evidcnce for rcversal of thc geomagnetic field "car the Plio-Pleistocene boundary. Science U3, 1126-1127. Opdykc, N. D.. Claoue-Long. J . , Roherts. J., Irving, E. (1995). Preliminary Magnctostratigraphic results from the Carhaniferous uf Eastern Australia. EOS. Abstracts, Fall Meeting. 171. 0pdyke.N. D., Glass, B., Hays. J. P., andFoster, 1.(1966). Paleomagnetic study of Antarctica deep-sea corcs. Science 154,349-357. Opdyke, N. D., Ninkovich. D., Lowrie. W.. and Hayes, I. D. (1972). The palaeumagnetism of two Aegean decp-sea cores. Eurth Planef. .?ti Lcn 14, 145-149, Opdyke. N. D., Kent. D. V., and Luwrie, W. (1973). Details of magnctic polarity transilions =curded in s h i ~ hdeposition rate dcep sea corc. Eorrh Planet. Sci Lett 20, 315-324. Opdyke, N. D., Rurkle, L. H.. and Todd, A. (1974). The extension oi the magnetic time scale in sediments of thc central Pacific Ocean. Eorrh Planer. Sci. LetI. 22, 3W-306, Opdykc, N. D.. Lindsay, E. H., Johnson, N . M., and Downs, T.(1977). The paleomagnetirm and magncric polarity stratigraphy of the mammal-bearing section of Anra-Bonego State Park, California. Quar. Kes. 7, 316-326. Opdyke, N. D., Lindsay. E.. Johnson. G. D.. Johnson, N., Tahukheli, R. A. K.. and Muza, M. A. (1979). Magnetic pr,larity stratigraphy and vertebrate paleontology of the upper Siwalik suhgroup of northern Pakistan. Palaeogeugr., Palrrer,climalol., Palrrvoecol 27, 1-34. Opdyke, N. D., Johnson. N. M., Johnson. G., Lindsay, E., and Tahirkheli. R. A. K. (1982). Palcornagnetism of the middle Siwalik formalions of nr~rthernPakislan and ratatinn of the Salt Range DCcollcment. Palmgcogr. PalaeoriimaioL, Palueoecol 37, 1-15. Opdyke, N. D., Mein. P., Moissenet, E.. Pcrcz-Gonzalez, A , Lindsay, E.. and Pctko, M. (1989). The magnetic stratigraphy trf the late Miocene scdirnrnts of thc Cabriel basin, Spain. In “European Neogcnc Mammal Chmnology" (E. Lindsay el ol. (eds.). pp. 507514. Plenum. New York. Opdyke. N. D.? Khramov, A. W.. Gurevitch, E., losifidi, A. G., and Makarov, 1. A. (1993). A paleomagnetic study of the middle Carboniferous ofthe DonrlzBasin. Ukraine. EOS. Trans. Am. Geophys. U n i o ~Spring Meet., p. 118.

.

314

Bibliography

Opdyke. N. D., Mein, P., Lindsay, E., Perez-Gonazales, A,, Moissenet. E.. and Norton, V. L. (1996). Continental Deposits, Magnetostratigaphy and Vertebrate P a l e o n t ~ l n ~ , Eastcfn Spain. Palaeogeogr. Pala~oclimatol.,PalurocoL, submittcd for publication. O'Rrilly. W. (1984). "Rock and MineraI Magnetism:' Blackie. Glasguw and London. Orguira, M. 1.(1990). Palcomagnetism of late Crnanoic fossiliferous sedimsnts from Barranca de lus Lobes (Bucnos Aircs Province, Argentina), The masnetic ages of the South Am~ricanland-mammal agcs. Phys. Eanh Planer Inter 64 121-132. Orguira, M. I.,and Valencio. D. A. (1984). Estudio paleomagnetico de lossedimrntos asignado al crnroico tardio aflorantcs cn la barranca de los L(lhmr Provincia de Burnos ~ i ~ ACIU C o n ~ rGeol. . Argent. [N.S.] 4 162-173. O'Sullivan, P. E., Oldfield. R., and Battarbee. R. W. (1972). Prellminarg studies of Lough Nragh sediments. I. Stratigraphy.chronology and pollen anslysis.ln "Quaternary Planet E c o b g y ' ~ H J. . B. Birks and R. G. West, eds.). Blackwell. Oxford. Ozdernir, O., and Banerjee. S. K. (1982). A preliminary magmtic study of soil samples from wcst-ccntral MinnesOta. Earth Planel. Sci Lert 59, 393-403. bzdcmir, 0..Dunlop. D. J., and Moskowitz. 8 . M. (1993). The effect of oxidation on thz I'crwcy transition in magnetite. Ceophj.~.Rrr. I-ett.El, 1671-1674. Palmcr, D. F.. Hanyey. T. L., and Dodson, R. E. (1979). Palcomagneticand sedimcntolo@c~l studic5 at Lake Tahoe: California-Nevada. Earth PInner Sci. Luff.46,125-137. Palrncr. J. A . Perry, S. P. G., and Tarling. D. H. (1985). Caboniferous magnrtostmtigraphv. J. GeoL Sac. London 142,945-955. Parkcr. E . N. (1969). The occasional revertnl of the geomagnetic field. Asuophyj. J. 158, 815-827. Parry. L, Ci, 0980). Shape-related faclon in the magnetization of immobilized magnetite particles. Phys. Earth Plunel. Zn1t.r 22,144-154. Parry, L. G. (1982). Magnetization of immohilizcd particles dispersions with two dirtinct particle sizes. Phys. Earrh f ! ~ n e t .Inter 28, 230-241. Paync, M. E., Wolbhg, D. L., and Hunt, A. (1983). Magnetostratigraphy uf a corc from thc Raton Basin.Eew Mexim. implicalionv for synchroneity of Crelacrous-Teninry boundary events, New Mexico. Geology 5,41-44. Peck. J. A,, King. J. W., Colman, S. M., and Kravchjnsky, V. A . (1994). A rock-magnctV record from Lake Baikal, Siberia: Evidencc for Late Quaternary climatr change. Earth Planet Sci Lerl. U2,221-238. Peck. J. A.,King, J.W., Colman, S. M.,and Kravinsb. V. A. (1996). An 84-kyr paleornagnetic record from sediments nf Lake Baiknl, Siberia. J. Geophvs. Rex. 101. 11,365-11,385. Pcng, L.. and King, J. W. (1992). A Late Quaternary. eeomametic . - secular variation record &om Lake Waiau, Hawaii. ilnd the question of the Pacific non-dipole low. J. Geophys. Kes. Y7,44074424. Peuagno.E. A.. and Blome, C. D. (1990) Tmplications of newJurassic stratigaphic, geochrononometric and paleolatitudinal data from the wcstern Klamath terrane (Smith River and Roguc Valley subterranes). Ger,lr,g). 18, 665-668, P e l m m , N., Hcller. F.. and Lowrie. W. (1984). Magnetostratigraphy of the Cretaceous/ Tertiary, geological boundaiy. In "Initial Reports nf the Dccp Sca Drilli?g Projzct" (K. J. Hsu, 1.L. LaBrecque el a/., eds.), Vol. 73, 657-661. U.S. Govt. Printing Office. Washington; DC. Peter-, N., vvn Doknock, T., and Vali, H. (1986). Fnssil bacterial magactite in deep-sea sediments from the South Atlantic Ocean. Nolnr@(Lnndon) 320, 611-615. Pevzner, M. A,, Vangmgeium, E. A., Zazhigan. V. S.; and Liskun, I. F. (1983). Correlation orthe Upper Neogenr sediments of Central Asia and Europe on the basis of paleomagnetic and biostratigraphic data. Inr L'eol. Rev 25, 1075-1088.

.

.

Phillios.. 1. D. 119771. Time variation and asymmerry in the sVatistia of geomagnelic reversal . sequences. I . Geoph),~.Rrs. 82,835-843Picard, hl.D. (1%4).Palroma~ncticmrrelali~~nofunitswithinChugwaterrrriarsic)formation. . wcst-central Wyoming. An?. Asvoc. Pet. Geol Bull 48,269-291. Pitk, T., and Tauxz,L. (199%). Holocenc pnlecrintensities: Tellicr experiments on submarine hasaltic glasr from the East Pacific Rise. J. Geophys. Re$. 98, 17.946-11.949. Pick, T., and Tauxc. L. (1993h). Geomnenetic paleointcnritirs during the Crct&crtrusnormal superchron measured using suhmarine basaltic plssr. Narurc (London) 366, 238-242. . Piper. J. D. A. (1987). "Palacornagnetism and thc Conlinental Crust." W r y . Krw Yotk. Pitman, W. C.,TlI, (1978). Relatic~nshipbetwccncustir~yandstratigraphic rrqurnces of passive margins. Grol. Soc. Am. Rull (W, 1389-1403. Pitmao. W. C.,111. and Heirtzlcr, J. R. (1966). Magnctic anomalies over the Pacific-Antarctic c ~ 1164-1171. ridge. S c i ~ n 154, Pluhar: C. J.. Kirschvinck, J. L.. and AaamS, R. W. (1991). Stratigaplly and intrahasin correlation of thc Mojave River formation, central Mojave Dwrrt; California. Sari Rernarilino Co. NLIS. ASSOC. Q. 38, 31-42. Pluhur. C. J..Holt, 3. W.,and Kinchvink, 1. L. (1992). MagnctuatratigapI~yofPlio-Pleistl.,Mauritsch. H. I., Eder, G.. Grass. F., Rogl, f~.,Stradnrr, IL, and Surenian. R . (1986). The Crcraccous-Tertiary hcrundary in the Ciosau Rmin, Austria. 1Varr6rc (London) 322,793-799. Prcmoli-Elva, L (1977). Upper Crrtacmus-Paleocencmayneticsrrati~aphy at Gubhio, Eta$. IT. Biost~atigraphy.Ceol Zoc. Am. Bull. 88%371-374. Prrmoli-Silva, I.,Orlando. M.. Monechi, S.,Madilc, M., Napnlr<me.G., and Ripcpe, M. (1988). Calcareous plankton hiostraligraphy and magnctostratigrilphy at the Eoccne-Oligocene transition in thc Fuhhio area. Inr. Sccbrornm. Palengene Srrrrrgr. ( U U .Me&, Anacona. 1987, Spec. Puhl. TI, Vol. 6, pp. 137-160. Prcnlice. M. 1,and Mauhews. R. K. (1991). Tertiarj icr sheet dynamicr, the snow pun . 96, 6811-6827. hypothesis. J. G r o p h ~ rRes. PrL~ut,M.: and Camps" P. (1993). Abscncr r,f preferred longitude sccton for poles from vcrlcanic records of gmmagnctic polarity reversals. Aromre (London) 366, 53-56. PrCvot. M..Mankincn,E. A..and Gromrn6.S. (1985j.The SteCns Mountain (Oregon) geomagnetic polarity transition. 2.Picld inlensily vaIiations and discu.;ri<ma1 reversal modcls. . I Geoph.ys. Kes. W , 10.417-10,448. PTCV<,~, M.. Derder. M. E.. McWilliams.M., and Thompson, J. (1990). Intensity of thc Earth's magnetic ficld: Evidence far a Mesozoic dipole lea,. Forrh Plmer. Sci. Lrtt. 97, 129-139. Prince, R. A,. Hcath, G . R.. and Kominr. M. (1980). Palaeamagnetic studies of ccntral North Pacific sediment corss: Stratigraphy, scdimcntation rates, and thc origin of magnetic instability. G m l Sot. Am. Bull.. Part 2 91(8). 1789-1835. Pringle, M. S.; Staudigel, H.. Duncan. R. A., Christic, D. M.: ODP Leg143 and 144 Scietltlflc Staffs (1993). 4 U A r l w ~ages r of basement lavas at Resolution MIT. and Wodejebato

.

~

Guyots compared with magneto- and hin-stratig~aphicrcsultr from ODP Legs 143!144. EOS, Trans. Am. Gcophys. Lnion 73(43), 353. Pmthero, D. R. (1985). Chadronian (early Oligocme) magnetostratigraphy of eastern Wyoming: Implications for the age of the Eirenc-Olipoccnc boundary. J . Grol. 93,553-565. Pn~lhrro,D. R. (1991). Magnclic stratigraphy of Eoccnc-Oligocene mammal lncalitirs in southcm San Oicgo County. irr "Eocene Geolofiic Hi~toryoi lhc San Diegu Region" (P. L. Ahhot~and I. A. May. edr.), Pac Sect. SF.PM Sucictv for Sedimcntarv Geolow.. No. 68, pp. 125-130. ixlthcI.incnln Prothero. D. R.. and 12mmentrout.Jhl. r1985). , , Maenelostraiimanhiccorrcialion . , Creek Formatiurn, Washington: Implications for thc azc of thc Eoccnc-Olieoccnc boundarp. Geology 13,103-211. Profhero. D. R.. and Su,ishrr. C. C. (1492). Magnetorlrilligraphy and gaochrtmology cxf the terrestrial Eocene-Oligoccnc transition in North Amcrica. In "Eacenu-Oligocene Clirnalicand Biotic Evolution" (D. R.Prothera and W. Berggren, eds.),pp.46-73. Princeton Llniv. PKSS,Princeton, NJ. Prothero. D. R.. Denham. CL R.. and Farmer. H. G. (1983). Magnctostr~tigraphyof the While River Group and its ilr~plicationsfor Oligocene geochronology. P a l n e ~ ~ c o ~ y r , Palaeocli~narol.,Polaenrcol 42,151-166. Pullaiah, G.. Irvm$ E.. Buchan, K. L., and Dunlop, D. 1. (1975). Magnrliralir,n chilngcs caused hy burial a d uplift. Enrfh PlonN Sci. Lett. 28, 133-143. Purucker,M. E.. Ela1on.D. P.. and Shoemaker,E.M. (1980) Early acquisition ot Characteristic magncti~atirrnin red hedr of the Mornkopi Formarion [Triassic), Gray Mnuntuini. Arizona. J. Ci~ophys.Re8 95, 997-1012. Quade, J., Cerling.F, E.. and Bowman, J . K. (1989). Uc%~lopment of Asian mon~aonrcvealed by marked ecological shift during the latest Miocene in northern Pakistan. .Voii,re (Lundun) 342,163-166. Ouidellrur, X., and Vulrt; I-P. (1994). Palr<,magnrllc records of rrcursirms and rcvcrsels: Possiblc hiasc5 caused by magnetization artifacts. f'hyhys Lor$, Plarlrr 1ntc.1 82. 27-48, Radhakrishnamurty. C.. Likhite. S. D.. Amin. B. S., and Somayajulu. B.L. K. (1968) Mapnllic susceptibility srratigaphy in ocean sediment cores. Earth Plilnrt. Sci. 1,prr. 4, 4M-468. Rail. A. D. (1966). Boundaries of an arCd of "cry long magnctic anomali~sin tltc north east Pacific. J. Geophpr Kes. 71,2631-3636. Rapp, S. D., MacFadden, B. 1..and Schiebour. J. A. 11983). " Magnetic - oolilritv sliiltirrmh\, of the early Tertiary Black Peak* Formittir,n, Big Bend Kational Park, 'Texas.J. ljeol, 91,555-572. Raymo, M. E.. Kuddiman, W. E.. and Clement. B. M. (1986). Pliocene-Pleistocene paleOceanography of the North Atlantic at Deep Sea Drilling Pn>jrcl Site 60". I,, "lnitial Reports of the Deep Sea Drilling Prnjrcl" (U'. F. Ruddimsn, R. R. Kidd, E.'rhomns, etnL, rdr.). VoI. 94, pp. 895-901. U.S. Govt. Printing Officc. Washington, DC. Raymo, M. F.,Ruddiman. W. F.,Backman, J., Clcmcnt. B. M, and Martinson. D. G. (1989). Larc Plloccne variation in Northern Hemisphere ice sheers and North Atlantic deep water circulation. P~~llleucuanugnrph~ 4,413-446. Raynolds; R. G. II., and Juhnrcm, G. D. (1985). Ralcs rlf Ncugene drp<,ritiunal and dcfurmational pn,ccsres. northwcst Himalayan fore dccp margin, Pakistan. ix "lhc Chmnology of tho Geological Kccard" (3.J. S n c h g . cd.). pp. 297-311. Blackwell, Oxford. Kcevc. S. C.. and Helgley, C. E. (192). Magnetic reversal sequence in the upper pottian of the Chinle Formation. Montoya. New Mexico. Geul Suc, Am. Bull. 83,3795-3812 rrsults Reeve, S. C.: Leythaeuser. D.. Iirlslay. C. E., and Bay. K. W. (1971). Palr~~milgnelic from the 1;pprr Triassic o l Earl Greenland. .I. Oeophy.?. R r s 79, 3302-3307. , . L

L.

Rennc P. R.,Fulfard, M. M., and Rushy-Spera, C. (1941). High resolution 4nAri'9Archronostratigraphy 01 the L a ~ eCretaceous El Gallo Formation, Baja Cxlifornia Del Norrc, Meximr. Geopf~.vsRrr, Lrrr. 18, 4SS462. RennP. P. R.. Walter. K., Vcrosub, K., Sweittzer. M., and Aronson. 1. (1993). N c r data from Hadar (Ethiopia) support orhilally tuned time scale to 3.3 ,Ma Grt)phys. Res. Let. 20, 1067-1070. RennC. P. K.. Dcino, P. L., Walter, H. C.. Turnl~.8 . D.. Swisher. C. C, 111. Becker. '1.A., Cums,G. 11, Sharp, W. D.: and Abdur-Rahim.J. (1994). lnlercalibration of astronomical and radioisotopic time. C~olr,gy22,783-786. Kcynoldr. R. L.. Tuttle, M. L.,Rice. C. A,.Fishman.N. S.; Karachewski, J.A.. and Sherman, D. M. (19%). Magnetization and geochemistry of grrigirr.hearing Cretaceous strata. north %lopebasin. Alaska. Am. 1,Sci. 294,485-528. Ricck. H. J., SarnaWojcick. A.M.,Meyer, C, E..and Adam, D.P. (1992). MaLqdosPratigraphy and tephmchronolopy of an upper Pliocene to Holocene rccord in lakr srdimz~ltsat 'Yulclakr. northern California. Geol So<:. Am. 81~11104, . 409-428. Rio.D., Spruvie~i.R..and ChanncU. J. F,. T. (1940). Pliocene -early Plcistacenr chrunnrtratiarnohv ~,and the Tvrrhenian dcco-sea reuw.l lrirm sits 653. In "Pracccdines of Lhr Ocean Drilling Program. Scientific Kcsults" (K. A. Kastznr. L Muscle ur nL, edr.). 1'<11. 107. pp. 705-715 ODP. College Station. TX. Rlo, D., Sprnvirn. R.. and 'll~unell.K. (1991). Pliocene-lowcr Pleistocene chronostraligraphy: A re-eealualion of the Meditenancan typssrctions. &,I. Svc, Am. B ~ l 103,1Q4')-1058. l Ripperdan. R. L., and Kirschvink. J. L. (1992). Paleoma%nrtic results from the Camhnan Ordovician hclundarp section at Black Mountain. Getngina Basin, Westcm Clueensland: Australia. in "Global Perspectives on Ordovician Geology" (B.D. Wcbby and J. R. Laurie. eda.). , no. 93-103. Balkcma. Rottcrdam. Robcns. A. P.. and Turner, G. M. (1993). lliagenrlic lormation of ferrimsgnctic inm rulphide mincrais inrilnidlv dcoasited marinc rcdirncnts. Srwth Island.New Zcaland. Earth Pltmel. Srl. Latr 115:25j-2;3. Roberts, A. P.. Turner, G. M.. and VcUa. P. P. (199.1a). hlagnetostratigraphic chronology of hiostratigraphic andoceat~o$raphiccucntsin NCWZealand. lam Miocrneloearl~~Pliocene Ccol Snc. Am. Brrll 106,665-683. Roberts. A. P., Vrrwuh. K. L.. and Negrini. K. M. (19946). MiddleiLatc Pleirtucrne relative paiacaint?nsily of the geomagnetic field from lacusirine sediments, lakr Ch~waucan. westcm linitrd Slates. Cevphgr. J. I d . 118,lfll-110. Robinson. S. G. (1986). The Late Plcistoconc palaecrclimatic record of Nmth Allantic deepsca sedimrntr revealed by mincralmagnrtic mrilsurcmentr. Phps. Eorrh Pluner lrllrr. 42,ZZ-46. Robinson. S. G. (1990). Applications of whole-core magnetic surceplihility mearuremcnts of deep-sea sediments. h "Proceedings of the Ocean Drilling P~r,&ram. Scieluitic Result\" (R. A. Duncan, J. Backman. L. C. Peterson, er 01.. cds.). Val. 115, pp. 737-771. ODP, Collegc Station. XX. Robinson, S. G., Maslin, M. A. and McCave, 1. N. (1995). Magnetic susccpfihility variations in Upper Plcistoccnr derp~seasedirnentsof thc VE Atlantic: lmplicatiunh lor in rafting 10, 221-251). and palcwinuiation at the last glacial maximum. Pirleoc~anogmpl~y Roehe. A. (1950). Sur les caracrfrer magnetiques du systkme empiif de g
..

Rochette. P.,andFilli<m.G.1198V). Field and temprraturo hchavior of retnanencr in eynthetic gocthifc: Paleotnagnetic implications. Urophys. Rrs. Lett 16, 851-854. Rudda. P, McDougall. I.. Cassil. R. i\.. Falvry, D. A.. Todd. R., and Wilcoxon, J. A. (1985). Isotopic ages. magnetostmtigrapl~y,and biuslratigrrphy of thzEarly Pliuccne Suva Marl, Fiji. Brrll. Geol Soc. Am. 96, 529-538. Rudionov. V. P. (19661. Dipalc character ofthc ~eomagnrticficld in the late Cambrian alld O~dovinanin thc south of the Siberian Platform. G m l . G~nj?:.1,W-101. Roggcnthm, W. A. (1976). Magnctic stratigraph!, of the Paleocene. A comparison brtwcct~ Spain and Italy. . b l m ~Gcol. . Sac. Irnl 15, 7G82. Roggcnthm, W. A , and Napolconc, G . (1977). Upper Cretaceous-Palt~~cmemagncticstratigrilphy at Guhbio, Italy. 1V. Upper Maastrichtian -Paleocene magneticstratigraphy. Geol. Soc. Am. Brrll. 88, 378-382. Ruw1oy.D. B . ,and Lottes. A. L. (1988). Plate-kinematic reconrlruclir~nrof the Nonh Atlantic 155, 73-120. and Arctic late lurasaic to present. Ttcro,~uphy.~ic.s Roy, J. L.,and Morris. W. A. (1983). A review i~lpalromngneticresults from the carbonifcrau.; of North America: The concept of carbonilercrur gcumagnrtic ficld horizon !markers. Earth Planer. Sci Lerr 61167-181. Ruddiman, W. F.,Mchtyre. A,. and Raymo, M. (1986). Matuyama 41,000 ?car cycles: Korth Atlantic Occan and northern hemisphere ice shrt.t%.Enrth Planet Sci. Lerr. 80,117-129. Ruddiman, W. F.. Kaymo, M. E., Martinson. D. G., Clemont, B, hl., and Backmau, 1. (1989). Plcistoccne evolution: Northern hrmi\~hhrrcice shcct and north Atlaltic Ocean. Poleor~nnograph?4.343-412. Rummery. T . A. (1983). The use of magnrlF mrilsuremrnfs in intnprcting the tire histories of lakc drainage basins. Hydrobiologiu 103, 53-58. Rummcry. 'r. A,, Bloemendal. 1.. Daring, I.. Oldfield. F.. and Thompson. R. (1979). The pcrsistcncc of tire-hduced magnetic mides in soils and lake sedimcnts. ANn. C.'copt~~vll.s. 35, 103-1U7. Runcorn, S. K. (1959)?On the theory of the geomabmetic secular variation. Atlrr tieophys. 15,R7-92. Runct>rm.S. K. (1992). Polar paths in geomagnetic reversals. 1Vuf1rrr(I-ondotz) 356,654-656. Ruocn,, M. (1989). A 3 Ma pale,omagneticrecord of coaslal cnnlinrntal depositsin Argentina. Pularogengr., Pala~oclimalul..PulaeoccuL 72, '105-1 13. Rutten. M. G., and Wcnsink. H. jl9W). Paleomagnelic daling. glaciations ~ n dchronology o l lhe Plin-Plcistoccne in Iceland. Inr. Geol Congr, Rep. SPSS.,Nordm, Zlst, Vol. 4, pp. 62-70, Rutter, N., Zhongli, 0.. Evans. M. E.. and Yuchum, W. (1990). Magncturtratigraphg of the Baojo loess-palcosol scction in the North Central China Loess Plateau. Qnar, lnt 718, 97-1 02. Ryan, W. B. P., andFluud. J. D. (1972). Preliminary paleomagnetic measurements nn srdimcnts from the lonian (rite 125) and 'l'ymhcnian (site 132) basins of the Medilerranran Sea. IN "Initial R r p o r ~o ~ l Lhr D r ~ Sea p Drillinp Rojcct" ( W . B. F. Ryan. K. J. Hau rr oL, eds.). Vol. 13. pp. 599-601. TJ.S Govt Printing Officc, Washington, DC. Sadler, P. M. (1981). Sediment accumulalion rater and the complctcncss of stratigraphic sections. J . Ceol 89.569-584. Sak8.H.. and Kcnting, B. (1991). Paleomagnetism ofLeg119, Holrr 737A. 738C, 742A, 7458. and746A. In "Proccedines ScientificRrsults' ( J . Barran. - of the Ocean DrillinePromam. . . B. Larsrn el a/., cds.). Vol. 11Y, pp, 741-770. ODP, College Station, TX. Salloway, 1.C. (1983). Palcamagnctism of scdiments front Deep Sea Drilling Prujcn Leg 71. In "Initial Reports of the Dccp Sca Drilling Project" (W. I. Ludwig, A. Krarhrninnikov ct a!., eds.). Vol. 71. pp. 1073-1091. ITS. Govt. Printing Oilice, Washigton. DC.

Sayre. W. 0.(1981).Preliminaryrcportonthepal~amagnetirmafAptian and Albianlimestonc and lrachytes from the mid-Pacific mountains and Hcss Rise. In "Initial Reportr oi the Deep Sca Drilling Project" ( J . Thcide, T. L. Vallirr. and C. G. Adelseck. eds.). Vol. 02, pp. 983-993. U.S. Govt. Pnnting Office, Washington, DC. Schncidrr, D. A. (199.1). An estimate r,i late Pleistocene geomagnetic intensity variations from the Sulu sea sedimcnts. Eurrh Plnnpr Sci. Left. 120, 301-310. Schncidcr. D. A. (1995). Paleomagnetism ulsome Occan Drilling ProgramIrg 138,sediments: Detailing Miocene magnetostratiguphy. h "Proceedings of the Ocrw Drilling Program, Scientific Kesulls" (N. G. Piaias, L. A. Maycr. 1.R. Janeeek rr oL. eds.). Val. 138. pp. 59-72. Collepe Station, TX. Schncidcr. D. A., and Kcnt, D. V. (1988). Inclination anomalies fronl Indian Ocean sediments and the possibility of a standing "on-dipole ficld. .I. Geoptrys. Res. 93, 11,621-11,630. Schneider. D. A., and Kent. D. V. (1990a). Thc time average paleomagnetic field. Rev. Geophys. 28,71-96. Schncidcr,D. A,. and Kcnt, D. V. (IYYOb). Palromagnctism ~ f k 115 g scdimcnts: Tmplication ,fc,rhi.ogene magnetostratigraphy andpaleoiatitudc ofthe Reunion holrpot. In "Proceedings of thc Ocean Drilling Project. Scientific Rerults" (R. A. Duncan, J. Blackman. L. C. Pcterson el 01.. eds.), Vrll 115. pp. 717-736. ODP, Cullqe Station. TX. Schncpp, E.. and Hradctlky, K (1994). Comhined palcointenrity and a.4r/'9~r ngc spectrum data from volcanic rocks of the West Elfel ficld (Gemmay): Evidence for an early Rrunhes vr
-

%~

Shacklcton. N. I., and Updyke. N. D. (1977). Oxygen isotopc and palaeomagnctic cvidzncc for uarly Northern Hcmisphcre glaciatirm. Nature (I-ondon) 270,216-214. Shacklcton, K. J., Backman. J.. Zimmerman, H.. K m t . D . V.. Hall, M. A.. Robcrrs. D. G., Schnitkcr. D., Baldant. J. G..Desprairiccr.. A.,H
Snowball. I., and Thompson. R. (1988). 'l'hr occurrence oC greigitr in sediments from Loch Lomtmd. I. Qtrar. Sui 3, 121-125. Snowhall. I., and Thaintrmsition in oatural pyrrhotitr dcducrd from domain stmcntrc t?hsen.ations. .I. Geopkys. 42.351 359. Soriel. H . C. (1981). Domain slructurc of natural fine-grained pyrrhotitc in il rock matrix 26, 98-106. (diabase). Ph?r Earth PIunei 111t~r Spllrks, N. H. C.. Mann, S.. Bszylinski, D. A.. Lovley, D. K., Jimnarch. H. W., and Frankel. R. B. (1490). Structure and morphology of magnetite anaenrbically~producedby a marme magnctotactic bacterium and a dissimilarury iron-rcducing bactcrmm. Eilrih Plonrt Sci. Len 98, 1 4 2 2 . Speiss, V. (199U). Crnowic magnrtostratigraphy of Lcg 113 drill sitcr. Maud Risc. Weddell Sea. Antarctica. It1 .'Proceedings of the Ocean Drilling Program, Scientific Kcsults" 113. (P. F. Barkcr. 1. P. Kcnnrllet 0 1 , cds.), V o l 113. pp. 261-290. College Statilm. ?X. Spcll, T. L.. and McDougall, 1. (1992). Kcvisions to thc age of thc RrunhesIMatuyama hound^ arg and r h r Plzistoccne geomagnetic polarity timescale. Gznpkys. Rrs. I.utr. 19, 11821184. Spmrieri. R. (IY92). Moditerrancan Pliocene hiuchranology: A high resoluliun record b n v d onquantitativc plsnktonicforaminifera distribution. Kiv. Irul Pnleonn~l.Strar Y8,hl-100. Sprawl. D. R., and Ranrrjee. S. K. (198Y). 'The Iioloccnc paleosecular variation rccurd f r o b F.lk Lake. Minnesota. J. Gr<,{,l~ylhys.Ke.7 94. 9369-9188. Stamalakos, J., and Kodama, K. ( l Y Y l ) . ~effecrs r of grain-acalc deimmation on the Bloomcburg formation polc. I. Gcopltys. Rut.. 96, 1991. Strunvald, R. A.. Clark. D. L.. and Andrew. J . A. (1968). Milguetic stratigaphy and ixunal pattcrns in Arctic Occan scdimmnts. Earrk Pluner, Sri. Lett. 5.79-85. Stsincr. M. B. (IY78).Magneticpc~lnriiyduringthe Middle Jurassic as rrcnrdedin the Summcrvillr and Cuair formatlirns. Earth Plu,~rl.S c i L.w 38, 331-145. Stcincr.M. B. (19831. Detrital rrmanent magnztizationin henlatitc.1. Geophys. Rc.Y. 88,65236539. Srciner. M. (198R). Palcumapnztisin of the latc Pennsylvanian and Permian. .I. Ceophy.~.Rrs. 93,2201-221 5 Slemcr M . and Hclaley, C (1974) M a p c t ~ cpolarlt! grqumce of the Kaycnta ft,rlnatlon Gur~lugy2, 191-194. Stcbcr. M. B.. and l5elslcg; C. E,llY75a). Late Jursssic magnctic polarity sequence. torth Pluner. Sci. Let,. 27. 108-112. Stcinrr, M. B., and Heirlc?. C. E. (1975h). R e v n a ~ pattcrn l and apparcnt polar wander fur the latc Jurassic. C;rol Soc, Am. B~
322

Bibliography

Steiner. M.B..Ogg, J. G..Melendez.G.. and Scqurirus.L. (198518hj.Jur~sicmagnctostratigraphy. 2. Middlc-Latc Oxfordian of Aquilon. Iberian Cordillera, northern Spain. Earth Plawr. Sci I.elt. 76, 151-166. Stcincr, M. B..Ogg, J. G.. andSandrwdl, J. (1987). Jurassicmagnctostratigapl~y,3. BathonianBajocian of Carcahury Sierrd Harana and Campillu dc Arcnas (Subbectic Cordillera. southcrn Spainj. Farrh PlaneL Sti. 1 . m 82, 357-172. Stcincr,M. B., Ogg: J., Zhang, A.;andSum, S. (1989). Thc Late PrrmianEarlylriassicmagnetic p o l a r i ~timescale and plate mutinnr uf Soulh China. .I. Geophyr. Ra. 84, 7343-7363. Stcincr, M. B., Moralcs, M., and Shoemaker. E. M. (19931. Magneu~traligraphic,hiastrati. graphic, and iitholugie cr~rrrlationsin Triassic strata nl the western llnilrd Slates. In "Applications of Pal~eUmagnrtismto Scdimcntar). Geology" (D. M. Aissiloui. D. F. McNeill. and N. F. Hurlcy. cds.), SEPM Society for Sedimenvary Geolugy Sprc Publ. No. 49. pp. 31-57. Steiner. M. R.. Lucas. S. G.. and Shoemaker. E. M. (1Y94). Correlatiun and age of rhc Late Jurassic Morrison formation from magnetostratigraphy analysis. In "Mesozoic Systenls of the Rocky Mountain Region. USA, Rocky Mountain Scctian". (M. V. Caputo. J. A. Peterson. and K. L Franczyk, eds.): SEPM Spec. Publ., Ucnvcr, CO. pp. 315-330. Stephenson, A. (1971). Single domain grain distributions. 1.A method for the determination of single domain she distributions. Phys Earrh Planer. inter. 4, 353-360, 361--369. Stoher, J. C; and Th,mpam, R. (1977). Palaeornagnetic secular variariotl studies of Finnish lakr sediment and the aarricrr oi rrmanene. Earrtc Planer. Sci. Lrrr 37, 139-149. Stoher, J. C., and Thompson, R. (1979). Magnetic remanence acquisition in Finnish lake sediments. Geophg,r. 3. K Astron. Suc 57, 727-739. Stoh, J. F.. Lovley.D.R.:andHaggerly. S. E. (1990).Biogenicmagnetitea1drhemagnetizatio~l of scdimcnts. J. Geophyv. Rrs. 95,4355-4361. Stoncr, J.S. Channell, 1.E. 7,lIillaue-Marcel. C.. andMareschal, J. C. (1994). High resolution rock magnetic r t d y 01 a Late Pteistocene core from the Labrador Sea. Can. J. Earlh Sci. 31, 104-114. Stnnrr, I. S., Channell. J. E. T.. and Hillalre-Marcel. C. (1995a). Magnetic propertier of deepe a sediments olf southwest Greenland: Evidence for major differences between the last two deglaciations. Ger,lu,$y 23,241-244. Stimer. J. S.. ChanneU, J. E. T.. and HiUaire-Marcll, C. (1995bj. Late. Pleistocene relative geclmapnetic paleointensily from the deep Labrador Sea: Regional and global correlations. Eurrh Pluner Sci. Letr 134. Z7-252. Stoner, J. S., ChanneU, 1. E. T.,and Hiliaire-Marcel, C. (1996). The magnetic ~ignatureof rapidly deposited delrilal layers from the deep Labrador Sea: Relatiunship to Nortll Atlantic Heinrich layers. Puleurmnogr~rphy11,309-325. Suk,D.,Peacor,R., andVan der Voo,R. (1990). Replacement of pyrite framboidsbymagnetite in limestone and implications for palaeomagnetism. Na'arur~,(London) 345, 611-613. Suk. D.. Van der Voo, R.. and Peacor, D. (1993). Origin of magnetite for remagnetization of Early Paleozoic Illnestones of New York State. I . Gcophrs. . . Kes. %. 419-434. Son I\'. m a J ~ t h \ d !.\I~ iIU0.I) SCJIIIII~CCICSIIUII nl.c~.>..op\ 111.1 r.t;k mArn:llL cnrrl:r\ ~n rcniaJr>?l./?.itxl, l ' a l ~ o ~ o . r : ~ ~ ~1r.m ~ ~ n\l~,\,?urt. : ~ T : ~ J ( ; t , .s , ~.t ~ ~ > V v . ? ~ ~ 3 5 - 2 ~ 1 4 ~ Suttcr. J. F. (1988). innovative approachcs to thc datingofigncous eventsin theearly Mesr>zuic basins of thc castcrn United Sratcs. Geol SLID. Rull ((1.S.) 1776, IN-ZW. Swishcr. C. C.. 111, and Knox. R. W. 0. (1991). The age rrf ihr Palrchcm&
Swisher, C. C. 111, Grajales-Nishimura, b M.. Montanari, A,, Margolis, S. V.. Claeys, P., Alvarez. W.? Renni.. P. R., Cedillo-Pardo. E., Maurrsse, F. J. M. K., Curtis. F.,Smil, S.. and McWilliams. M. 0. (1992). Coeval '"ArPYAr ages of 65.0 million years ago from Chicxulub crater mclt rnck and Cretaceous-'l'ertiary boundary tektites. Science 257,954-958. Svmonds. D. T. A. (1990). -, ~, Early Permian pole: Evidence from L l x Pictou red beds. Prince Edward Island. Canada. Grology 18,234-237. Talling, P.J.,Burbank, D. W., Lawrt<m.T. F.,H o b b ~K. . S., and Lund. S. P. (1994). Magnetostratigraphic chronology of Cretaceous tn Eocene thrust belt cvnlution. cenwal Utah. U.S.A. I . Geol. 102, 181-196. Tamaki, K., mdLarwn,R. L. (1988). TheMrsomic tectonic h~storyofthe Magellanmicrnplate in the western central Pacific. J. Geophys. Res. 93, 2857-2874. Tarduno, J. A. (1990). Brief revorscd polarity interval during the Cretaceuus normal polarity superchmn. Geology 18,683-686. Tarduno, J. A.. Sliter, W. V..Bralower. T. J., McWilliams. M., Premoli-Silva, I.. and Ogg, I. G. (1989). M-sequence revcrsnls recordcd in DSDP scdiment cores from the aestcrn Mid-Pacific Mountains and Magcllan Rise. Geol Soc. Am. Rull. 101, 130&1316. Tarduno, J. A,, Sliter. M. V.. Kroenkc, L.. Lzckic, M.. Maper. H., Mahoney. J. J., Musgravo, R., Storcy, M.. and Wintrrer, E. L. (1991). Rapid formation rrf the OntongJava Plateau by Aptian mantle plume volcanism. Science 254, 399-303. Tanluno, J. A,, Luwrie, W.. Sliter, W. V., Bralowcr, T. J.. and Hcllrr, F. (1992). Reversed polarity eharaneristics magnetizations in the Albian Contessa section. Umbrian Appcnines. Italy: Implications'f~athe existence of a Mid-Cretaceous mixed polarity interval. I. Geoyhys. Rrr. 97, 241 -271. Tarling. D. H. (19x3). "Palcomapetism." Chapman & Hall. London. Tarling, D. H., and Mitchell. I. G. (1976). Revised Cenozoic polarity timc s a l e . Geology 4,133-136. Tauxc, L. (1993). Sldimentsry records o i relative palenintensity or the geomagnetic ficld: Theory and practice. Krv. Grophys. 31, 319.-354. Tauxr. L., and Rutler, R. (1987). >lagnrtostratigraphy: In pursuit of missing links. Kev. G W I P ~ ~25, S . 919-951). Tauxc, L.. and Clark, D. R. (I9R7). New paleomagnetic results from the Eureka Sound Group: Implications for the q e of the early Tertiary Arctic biota. Geol. Sac Ant. Bull w,739-747. Tauuc, L., and Kcnt. D. V, (1984). Propcrlies of a dctrital remancnce carried by hematitc from studv of modern river deoosits and laboratorv rrdeposition experiments. Gruphys. J. R. Adron. Fnc. 77,543-561. Tauxc. L., and Opdyke, N. D. (1982). A time framcwtnk based on magnetostmtigaphy for the Siwalik sediments of the Khaur artra.nOrthcrnPakistan. Policru~cogr,Palrrroclimntol., Palrreoecol 37,43-61. Tauxe. L.. and Shackleton. N. 1. (1994). Relative paleointensily records from the OntongJava plateau. Geophr. 1. Inr 117,769-782. Tauxe, L.. and Valet. J. P. (1989). Relative palcointensity oi the Earth's magnetic ficld born marine sedimentary cores A global perspective. Phjv. Earrh Planet. i n a r . 56, 59-60. Tawre, L..and Watson, G . S. (1994). The fold test: An cigen analysis approach. Enrrh Planer. sn. Len. 122,331-342. Tauxc. L.. and Wu, G. (1990). Normalized rmanencc in sediments from wcstcrn equatorial Pacific: Relative paleointensityofthe geemagnetic field. J. Ger>ph,vs.Re*. 95,12337-12350. Tauac, L.. Kent. D. V., and Opdyke, N. D. (1980). Magnetic components contrihutitlg to the NRM of Middle Sidwalik red bedr. Earrh Planet Sn'. Lelr 47, 279-284. ~~

Bibliography

Trwnscnd, H. A,. and Hailwor,d. E. (1985). Magnetostratigraphic correlalicms of Palacogene sediments in the Hampshire and London bssins suuthcrn U.K. J. Gt~ulSoc. Londorr 142,957-982, 'Trench, A,, McKermw. W. S., and Tnnr,ik, T. H. (1YY1). Ordoncian magnctmtratigraphy: A correlation of global data. J. Geol Sot., L~rndnn1.18, Y49-958. Trcnch. A,. McKcrrow. W. S., TOrSvik. T. H.. Li. 2. X.. and McCmken. S. R. (1993). Thc polarity of the Silurian mapnctic field: Indications from a glohal data compilation. I. T;uol. Soc, Lmrdon 150, 823-831. Tric. E., Laj. C.. .I&lianno. C., Valet; 1.-P.. Kirsel. C.. Mazaud. A,, and laccarino, S. (199la). High-resolntion record of the Olduvai trannitinn frilm Po valley [Ital!~) rcdimcnts: support fur dipolar transition gcomctN Phys. Earrh Pluner. Inrrr 65, 319-336. Tric, E., Laj, C.. Valet, J.P..Tucholka, P.. Patcmc, M.. and Guichard. F. (1991b). The Rlakc geomagnetic event: Transition gerlrnelry, dynamical charactcristics atid geomq~neticsignificancc. Lnrrt! Planct. Sci. Lerr. 102, 1-13. Tric. E., Valet. I. P., Tucholka, P.; Paterne. M.. Labeyrie, L.. Guichard, F.. Tauxe, L., and Fontuone. M. (1992). P*ler,intmsity of thc gcomagnetic held for the last 8O.Wfl years. J G~ophys.Res. 97.9337-9351, Tsuchi, R., Takayanagi, Y.. and Shibata. K. (1981). Neogene hiorvrnlsin ihr Japancsc islands. Irl "Neogcne of J a w : ILs Riostratigraph!, and Chronology" (1. Tsuchi. ed.), pp. 15-12. Kurofunc Printing Co., Shizuoka. Tucholka. P.. Fr,nlugue, M., Guichard. F., and Paterne. M. (1987). The Rlakc polarity epirodc in cores from the Mrdilrrrancan Sca. Earth Planet. Sci. Lrlr. 86, 320-326. Tucker,P. (198U). Stirred remanent magne6an5im: A laboratory analogue of post-depoitinnal realignrnml. I Ccophys. 48, 153- 157. Tuckcr,P. (1981).~aleoi~lensilics from scdimcnts: Normalization by laboratory rcdaposition. Forth P l a ~ e tSci. Lerr. 56, 398-404. Tucker, P.; and Tauuxe. L. (19WI. Me downhole varialinn of lhc ruck-magnctic properties of Leg 73 sedimmts.ln "Initial Repons of thc Deep Sea Drilling Projr~.l" (K. J. Hsu, J, L. LaBrecque eta/,, eds.), Val. 73, pp. 673-685.U.S. Govt. Printing Office. Washinglim, D r . Turncr, G. M. (1987). A 50W year geomagnetic paleusecular variation record from westzrn Can&. Ceophyr. J. R. Astron. Soc. 91, 1U3-121. Tuiner. G. M., and Thompson, R. (1979). Behaviour of the Earth's magnetic field as remrdcd in the sediments of Loch Lnmond. Eorrh Planet Sci. Lea 42, 412-426. Turner. G . M.. and ThompsOn, R. (1981). Lake sediment record of thc gcomagnetic secular variation in Britain during Holocene times. Geopkys. 3. R. Asrrtm. Soc 65,703-725. Turner, G. M.. andT'hompsi)n;R. (1982).Dctransformalionofthe British geomagnrticsccular variation record for Holocene timzs. Grophy.r J. R. Asrron. Soc. 70, 789-792. Turner, F. (1975). Depositional magnetization of the Carhonifrroua limestoncs from the o ~ y563-581. Craven Basin of northcrn England. S e d i m ~ l t l o ~ 22, Turner, P., Metcalfe, I., and Tarling. D. H. (1979). Palcomagnctic studies of some Dimantian limestoncs from the Craven Basin and a contrihutinn 111 Aspian magnctostratigraphy. Proc, Yurks. Grrrl SIK 42, 371-396. Turner. P.. Turner, A., Ramos. A., and Sopeaa, A. (1989). Palaromagnctism of PermoTriassic rocks in the Iberian Cordillera, Spain: Acquisition of secondary and characlcristic remanencc. J. Ceol Sor., L o n d o ~146.61-76. Turrio. B. D.. Donnelly-Nolan. 1. M., and Hearn. B. C., Jr. (1994). ''ArijYA~ages from thc rhyolite uf Alder Creek, California: Age of the CubhMuuntain normal polarity subchron revisited. Geolr,g? 22, 251-254. Ilrrutia-Fucugauchi. J.. Bohnel. H. M.. and Valencin, D. A. (1990). Magnetostratigraphy of a Middle Jurassic red hcd sequt.nce from southern Mexico. Phys. Earth Plcnvr. Irtrer 64,237-246. ~

~

327

Uyeda, S. (1958). Thcrmorcmanent magnetism as a medium of palrromagnctism with special reference to rcvcrsr thrrmoremanent mapletism. Jpn. I Groplzys. Z,1-123. Vail, P. R., and Hardenbol. 1.(1979) Seairvel changes duringthcl'crtiap. Occanr~r22,71-79. Vail, P. R.. Mitchum, R. M., Jr., and Thompson, S., 111 (1977). Glohzl cycles of relative ch-angesof sea levcl. In ;'Seismic Stratigraphy: Applications to Hydmcarhon Exploration" (C. E . Paytnn, cd.). SEPM (Society for Sedimentary Geology Spec. Publ. No. 45, pp. 83-9. Valencio. D. A. (1981a). Palcr>magnetismof Lower Ordovician and Uppcr Prr.Cambrian rocks from Argentina. In '.Global Recnnstruction and lhc Ceolnagnctic Field During rhc Paleozoic" (.M. McElltinny n a/., rds.), Adv. Earlh Planet. Sci., Yol. 10, pp. 71-75. Ccntrr for Acadcmic Publ.. Tokyo. Valcncia, D. A . (lY81h). Reversals and cxcursions of thc grornapnetic licld as defined by palcuma_enetic data from upper Paleozoic-lower Mrco~nicsediments and igneous rocks. IIZ'.Global Kcconstruction and the Geomagnetic Field During rhc Palevroic" (.M. McElhinny et al., cds.). Adv. Earth Planet. Sci., Val. 10. pp. 137-142. Ccmlcr for Academic Publ.. Tokyo. Valencio, 1). A., andorgeira. M 1. (1983).Lamegnaostratigrafia del ensenadense y bonnrense de la cludad de Buenos Aires. Parte 11. Asor. Geol Argent Ruv. 3R(I). 24-33. Valencio. U.A.. Vdas. 1. F., and Mmdia. J. E. (1977). Paleomagnctism of a sequcncc of red beds of the Middle and Upprr rections of PagdmoGroup (Argenlina) and the corrchrion of Upprr Paleozoic-Lawcr h<esoroic rocks. GuOphys. J. R. Asnrn. Soc. 62,2749. Valencio, D. A . Vila. J. F., andPacca, 1. ti. (1983). The significance<,Iihe palaeomagnetism of Jurasxic-Cretaceous rocks from South America: Pre-drift movements. hairpins and magnet~,rlraugraphy.Gaopky. J, R. A s t m n Sr,c 73. 135-151. Valet. J. P.. and Mcynadicr, I.. (1993). Gcomagnrtic field jntcnrity and rzversals during the pan four million ycars. .Vu'itiurc (Lonriou) 366,134~238. Valet. J. P., Tiluxe, L.. and Clement. B. M. (19RY). Equatorial and mid-latitude records of thc laslgeomagnetic reversal from thc Atlantic Ocean. Eanh Planer. Sri. Lerr 94,371-384. Valet, J. P.. Tucholka. P.. Cuurliliot. V., and hfrynadier. L. (1992). Paleomagnctic constraints on thr. geometry of the grnmagneticficld durins reversals. .Vatare (Londotz)356,400-407. Valet. J. P.. Mrynadier. L.Rassinot. F. C.. and Gamier, F. (1994). Rela1i~'cpalrointensity across thr last geomagnetic reversal from sediments of the Atlantic, Indian and Pacific Occanr. Geophyr. R ~ JLerr. . 21,485-488. Vall. H., Funlei, 0..Amaranlidis. ti., and Pelenen. N. (1987) Magnetotactic baclrtia and thcir magnetofossils in sedimeots. Enrtk Plro!ri. Sri. Lett 86,389-4(W. Vandamma, D.. Courtillul. V.. Besse. J,. and Montigny, R. (1991). Paleornagnetism and age dctcrminations of the Dercan Traps (India): Results of a iiaupur-Bomhay traverse. and review of earlicr antk. Rev. L'eophys. 29, 159 -140. Vandmherg, 1.. and Wc,nders. A. A. H. (1980). Paleornapnetism of latc M n o ~ o i cpekaec limestones from the southern Alps. I. Geophys. Rm. 85,3623-3637Van der Voo. K. (1990). Pttancrorr,ic paleomagnctic poles from Europe and North Amcrica and comperisrms with continental reconstructiuns. Rev. Cenphys. 28, 167-206. Van der Voo, R. (1993). "Pnleomagnetism of thc Allantic. I'crhys and Iapetus Oceans." Camhtidgc IJniv. Press, Cambridge. UK. Verosuh, K. L. (l977aj. Depositional and post-depositional promses in the mngnctizulion of sediments. Rev C)mphys. Space Phys. IS(?], 129-143. Verosuh, K. L. (1977b). 'The absencc of lhr Mono Lakc gerrmagnetic excursion from the palromagnetic record af Clear Lake. California. Earrh Planer. Sci. Len. 36,219-230. \ferosuh, K. L, and Banerjee. S. K. (3977). Geomagnetic cxcursions and thcir palromagnctic record. RPL,Geophyx 15. 145-155.

Veroub, K. L., Davis. J. 0.. and Valnstro. S. (1980) A paleumagnetic rccord from e r a m i d Lake, Pleiada, and its implications for proposed geomagnetic mcurhi~mr.Earth P l n ~ t ~ t . Sci. Len d9, 141-148. Verosub. K. I,., Mrhringer, P. J., Jr., and Wiltrntraw, P. (1986). Holocene secular variation in western Norrh America: Paleomapetic recurd from F*h Lakc, Harncy Countv, Orcgcm. . I Gco,>lrys, Res. 9 1 3604-3623. Vrrosub. K. L.: Haggart. I. W.. and Ward. P. D. (19R9). Magnetostratigraphy of Upper Crrtaceousstrata oftheSacramento Valley. California. Guol SK A I I ~Bull 101,511 -533. Vincent. P., and Berger. W. H. (1985). Carbon dioxide and polar coaling in the Miocmo: Thc Montere? hypothesis. 1f1 "c' Carbon Cycle and Atmospheric CO,: Yatural Variatirtns Archean to Present" iE. T.Sundquist and W. S. BreLkrr, eds.). Geophyi. Monogr.. \rol. 32. pp. 455-468. Am. C.rt,phys. IInion, Washington, DC. Vincent. F.. Killingslry, J. and Rcrger. W. H. (19145) Miocene oxygen a i ~ dcarbon iallopu strarigraphg of thc tropical Indian Ocean. MPPJ. GIoI. Sot: Am. 163. 103-13fi Vine, F. J . (1966). Spreading of thc ocean floor: Nca c*idmce. Science 154. 1405-1415. Vine. F. J.. and Matthews. D. H. (1963). Magncric annmalirs over oceanic ridper. h'arurr (London) 199, 947-949, Vine, F. J.. and Wilson. 1. T. (1965). Magnetic anumxlirr over a young ocean rid8r off Vancouver Island. Science 150,485-459. Vogl, P. R.. Anderson, C. N., and Braccy. 1).R. (1971). M?n,znio magnetic ant~malirs,seafloor spreading, and geomagnetic rcvcrsals in the soulhwralcrn Uorth Allanfic. J. Geuphy~.Red. 76. 4792-4823. Walker, T. R.. Larson. E. E.. and Hoblitt. R . P. (1981). Katurr and migin o l hematite in the Moenkopi Fornlation (Iriarsic). Colrrrado plalrau: A contrihutii,n I', the origin of magnetism in rcd bcds. J . Geophg~.Rer. Rh, 317-334. Walter, R. C. (1YY4). Agc of I.ucy and Firs1 Family: Single cryclal '"AIP"AT duling or lhr Ucncn Dora and lower KadaHadilr mrmhrr\nfthe IIadar F
Walter. R. C.. and Anlnsnn, T. L. (1993) Age and source <,Ithe Sidi lIak<,ma tuff: Hadar Fc,rrnad<m,ELhiupia. .I l l u m Evrrl 25,229-240. Walter, R. C., Maneea,P. C., Hay. R. L., Drdke, R. L., and Curtis, C..H. (199Y). Laser fusion "'ArPS'Ar dating Red 1, Olduvai G n r ~ rTanzania. , N~~trrre (I.nndalon) 354, 145-149. Walt<m,A. H. (TO%). Magnrlnslratignilphy o l t h c Lowm and Middle members of the Ucvils Fravcyard Formation (Middle Ec,rrnr) Tnms-Pecr,s. Tmar. I,: "Eclcme-Olipt)ccne ClimslicandRi~~ticEvnlulion" (D. R. Pnxhcro m d W. B r g g m . rdr.), pp. 75-86,Princetrln IJniv. Press. Princeton. NJ. Wtllkina, N. D. (1968). ShorC period Eeomagnelic polarity evmts in deep b r a sedimentary cores. E&nh Plunrr. Sci Lert 4, 341-349. WaUdns, K.D. (1972). Review n l the development of the peomagnetic polarity timescale and . Sor. Am. Bull 83. 551-574. discussion oiprospects for its fine d e h i t i m ~Grol. Watkins. N. D., and Walker. G. P.L (1977). Magnetostratigraphy of eastern Iceland. An,. J. Sci 277,5513-584. Watkins. N. D.. McDougall, 1.. and Kristjansson. L. (1~75).A dctailcd palcamagnctic survcy of the type location for the Gilsa gcomagnctic polarity cvenf, t n r t h Planer. Scl. Lett, 27.436-444. Watkins. N . D., McDougdI. I., and Kristjansson, 1.. (1477). LJppcr Mirlcene and Plirlcrnr ~comagneticsccular vanation in the Rrrg~rfjnrdarA ~ e an l wcylrm Tli.limd. (;u~~phy,s, I . R. Artron. Soc. 49, 609-632. Watson. G . S . (l45ha). Analpis oldispenion con &sphere..Won. ,Not. R. Asiron. Sr,r Geuy,hy.s.. Stcppi. 7, 153-159.

i

i

Watron,G.S.(1956b). A test forrandomnessof dimctions.Morr. Not. R.A.~tmn.Suc Geophys., Suppl. 7 , 160-161. Watson, O. S.. and Enkin, R. J. (1993). Thc fold teat in palcomagllctism as a parameter crtimation problem. Uenphyr. R e . Left. 20,2135-2137. Wraver, P. P. E., and Clemml. B. M. (1986). Synchroneity uf Plioecnr plankton f<~raminifzial datums in Lhe North Atlantic. .War Micropoleonrul. 10,295-307. Webb, S. D.. and Opdykr, N.D. (1995). Global climatic a u e n c e on Cenozoic land mammal faunas. In "Effccls of Past Global Changes ou Lifc, Sludies in Gcophyria" Spec. Puhl. ,Varl. Arad Scr pp. 247-274. Wccks, R., Laj. C., Endignoux, L., Fuller, M., Roberrs,A.. Mangannc;R.,Rtanchard. F.. and Gorre, W. S. (1993). Improvements in lang-core measurement techniques: Applications in palaeomagnrlism and palaeaceanography. tirophys. I . Int. 114, 651-662. Weeks, R. J,, Laj. C.; Endignoux. L., Mazaud, A,, Lahcyrir. L.. Robcrts. A . P.. Kisscl, C.. and Rlancliard. E. (1995). Normalised NRM intensity dunnp the last ?40,DO0 years in &on cores from Cmtral North Atlantic Ocean: geomagnrlic field intensity arenvironmcntal ~5gnal'fPlrjr. Eurrh Pianel. Inrer. 87, 213-229. Wei. W. (1995). Rcvised age calibration points fur the geomagnetic polarity lime scalc. C e o p h y . Rm. Lett 22, 957-960. Weinrich. N.. and Thcycr. F. (1Y85). Palromagnctism of Deep Sca Drilling Project Leg 85 sedimmls: Neogcnr magnetOstrntigraphy and tectonic history of the Central Equatorial Pacific. In '-Initial Reports of Ihc Deep Sea Drilling Project" (I.. Mayor, and F. Theyer. eds.), Vrll. 85, pp. X49-901. U.S.Govt. Printing Oificr; Washington, DC. Weissen, H.. McKemic. 1.A . and Chamnell. J. E. 'L. (1985). Natural v.ariations in tha carbon cyclc during theearly Cretaceous. la .'The Carhon Cycle and AlmosphedcCO,: Natural Variations Archcan 10 Present" (E. T. Sundquist and W. S. Rrorcker, cds.), Oeophys. Monogr. Vol. 32, pp. 531-545, Am. Geophys. Union. Washingtoil; D L Wmsink. H . (1964). Palcc,magnetic strari~raphyof younger basalti and intercalated PlioPlcirtocene tillitcs In Iceland. Geoi. Rlmdxh. 54, 364-384. Westgatc, J . A . Sternper, B. A.,and PCw6. J. L. (1990). A 3 MY recardof Pliocene-Pleistocene loess in interior Alaxka. Geolox~,18(9), 858-861 Wrsrphal, M.. and Durand. J. P. (IYYO). Magnetostrilligaphic der a&ies continental HuritBlacustrcr du Cretacc supCtieu danr synclinal do I'Arc (region d'Aien Pravence. France). Bull Sor Geol fr. 8(4, Patt 1). 609-620. Whistler. D. P., and Burhank, D. W. (1992). hlioccnr hiostrarigrsphy and hiuchronolopy the Dove Spring Formation, Mojave Desctt. California, and charactcri~ationof the Clarcndunian mammal age (latc Miocene) in California. Ge,L Sot. Am. Bull. 104, 644-658. Whitelaw. M. 3. (1989). h.fagiletic polarity stratigraphy and mammaljan fauna of thc l a t e 97, Pliocene (Early Mnluyama) section at Watrslord (Vict(nia), AurWalia. 3. 624-631. Whitelaw. M. J. (1940). Magnctic polarity s t r a t i ~ ~ a p hofythePliocene Hnmiltou a!tdFor\yth's Rank Local sections. Hamilton (Victotia) Australia. I . Geol. 99. 3310-315. Whitclaw. M. J. (1491a). Magnetic polarity atratigraphy el Plioccnc and PlrisIocenl fossil vertebrate localities in sautheastcm Auslralia, Geol. Soc. Am. Ruli. 103, 1493-1503. Whilelaw. M. J. (1991h). Magnetic polatily stratigraphy of thc Fisherman's Cliff and Bone Rulch vcrrrhrate - ~ . ~fossil ~ iaunas from the Murray Basin New Suuth Wales. Aluilrulia. 6arth Plmpr. S c i Lctr 104,417 -423. Whitdaw, M.1. (1992). MaL~eticpohritystratigraphy ofthrccPli~~nescctionvandintcrmcm for the ages of vcnrbratc fossil sites "car Racchus Marsh Victoria, Australia. dust. J.

Whitelaw. M. J. (iYY3). Ageconnraints on the Duck Ponds and Limcburner's Point mnmmalian faunas hared n n mapctic polarity stratigraphy "I the Crelang arca (Victoria). Australia. Quar. R e . 39, 120-124. Wiedmmayer.F.(1980). Die Ammonitcn der niediterrancanProvin~im Plicmbachian und in, untercn Toarcian a u f p n d nltuer Ilnter~chungenirn Ciencrosos-Beckm (Lombardiachc Alpcn). Uenkschr. Srhweiz. !Vafurlorsch.Ges. 93, 1 260. Wixnall, P. 8 , and Hallam, A. (19931. tireixbachian iearlicst Triassic) Dalacoenvironmcnral ~~~.~~~ changes in the Salt Range,Pakistan and southeast China and their bearing on rhe Parmo'Triassic mass extinctio~l.Prrlcteogrogr.. Polnpuclimu~ol.,P o l a ~ o ~ c o102, r 215-237. Wilson. U. S, (1993). Contirmation of lhr astrunomical calihation of thc magnetic polarity timewalc from scaAloor spreading rater. "lunrrc (Loqdon) 364,788-190. Wllson.R. L. (I961). Pxlacomagnctisni in Northern Irelilnd.Pt. l.'Lhe thermal drmagnctizatiirn of nurural mapnelic movrmcnlr oiro&s. G c o p h y ~.I. Roy. Asfron. Soc 5,45-69. Wilson, R . L. (1962). The patelrt~magnrtismof hakcd contact rnckr and rercrsals of the ranh's magnctic tield. Geophys. J. R. A . ~ r r r xSoc 7. 194-ZU?. Wilson, R. I.. (1972). Paleomasnetic differences hrtwccn normal and revenrd field source.. and the pnrhlem of far-sldcd and r~ght~hvndrd prllc positions. C;raph,vi. J. R. Astron. Sac. 28.295-304. Wilson. R.I.,,and Lomax. R. (1972). Mapeticrcmancnce related la slow n,tatianof'ierromagnetic material in alternating magnetic lields. G~oplrys.J. K. Asrron. Soc. 30, 295-3U4. X'ing. S. L.. Bown, T. M., and Ohradovich. I. D. (1991). Early Eoccne biotic and sliamtic change in intcrior western No~thAmcrica. Geology 19,1189-1192. Wittc, W. K.. and Kmt.D. V. (1989). Amiddlr C a r ~ a n f o e a r l yXorian (-225 Ma)paleopole from scdimentsof the ?&ward Basin Pennsylvania. Feol Sot. A m BUN 101,1175-1181. Witte, W'. K.. Kmt, U. V.. and Olsen. P. E. (1YYL 1. Magn~t<,straligraphyand palsomagnetic poles from theZatc Triassic-earlies Jurusic strata of the Neu,ard basin. Grol Sor Ant. AldL 103, 1648-1662. Wollin, G., kicson, D. ?.,Ryan. W. R.F.. and Foster. I. IT. (1971). Magnetism of the Earth add climatic changcs. Earth Plrrnel. Sci. I.etr 12, 175-183. Wollin, G..Ryan. W . R . F.,Ericson, D. B.. and Foitrr.1. H. cIY77). Palaeodimate,palacomag nerism and the eccentricity of thc carth's orbit. (irephtphvs. Ker L e n 44, 267. Wuod. H. E., thaneiy, R. W., Clary, J., Calbcrt. E. H., Jrpson, G. I..,Rccsile. I . B.. Jr., and Stock. C. (1941). Noniendature afid correlation of the Korth Amrricm continental Tertiary. Geol Sor Am. Hull. 52. 1-48. Wcoodburnc. M. D. (19871. Rinciplcs, classifjcations and rrcommendations. In 'Cenclrilic Mammals of North America" (M. 0. Woodburne. ed.). pp. 9-17. liniv, of C~lifomia Press. Berkeley. Woodruff. F.,and Savin, S. M. (1991). Mid Miocene isotopic stratigraphy in the deep-sea: High resolution correlarions. Paleoclimatic cyclcs in sediment prrservatiun. Pa1eocenrro~raph.~6, 755-806;. Wright, S. D.. and Miller. K. C . (1992). Miaccncstableisorope stratigraphy, site747,Kergurlrn Plateau PIDL O c ~ a n1)rill Pwprnm UO, Sci. Results, part B.. 855-866. Wu. F., Van dcr Voo. R., and Limg. Q. 2. (1989) Reconnaissance rnagnctostratigraphy aC the Pre Cambrian-Camhrian boundary sec5on at Mcishucun. south wesl China. C u d . Geol lher 4 205-222. WyMe, P. J.: Irving, E., and Oxadetz. K. G. (1983). Palromagnctism of the Esayutr Fr~mation (Prmianj of notthem Ellesrnere Island: Possible clue for the solutinn of thc Nares Strait 100,241-256. dilemma. Tuc~~~noplz.v,~ics W F n e , P. T., Irving. E.. and Osudevz. K. G . (1988). Paleomapelinm 01 Cretaccousvokmic rocks Of the Sverdrup Rasin-may?netostratigaphy. paleolatitudcs. and rotations. Can. J. Eurrh Sci. 25. 1220-1239, .

A

Yama~aki,T., and loka, N. (1994). Lung-lam Eecular variation of the geomagnetic licld during the last 200 kyr recorded in sedimcnt corcs hum ihc western cquarorial Pacific. torllt Plotler. Sci. Lt.rt. 128, 527-544. Yamazaki. T..and Karsura:T. (1900) Ma&!nelicgrailr size and viscnusremandnt magnztization 01 pelagic clay. J. (>ur,phy,).Res. 95. 4373-4382, Yamazaki. 'I.., Katsura. I.. and Marnmo. K. (1991). O~igiilof stablc rcmancnt milgnrti~atiim of siliccour re'diments in thc central equiltitrialPacific. Eizrfl,Y l a n ~ LSci. Lun 105,Rl-93. Yang. Z..M<,rrau,M . ~ t i .Buchcr, , H.. r)r,mmerguts, J.-L.. and'l'rouiller. A. (1 996) Hettangiao and Sinemuilan magnctortvaligvilphyfrom Paris Basin. J. Ctop1iv.s. R e r 101,8015 8042. Y~skaua,K..Xu'akaiima,l..Kaw~i,N.,Torii.M.,Natsuhara.N.,sndH~rrie.S. (1973).Paleomagnriism of a corcfrom lake Biwu (I).J. Gcon,~gn.Georluccr 25,447-474. Yur.L.. Xishida. J., Sun, W.. Mia?. J.. Wang. Y . . and Sasajima. S. i198.1). The magnclosrratigrapliic ~ t u d gof Chinevc lorhs section from Shnan Xian and Xiieng. I,, " l b c Kcccnt Rsncarch oi L o u s i n China" (S. Sasajima end W. Yonyen, eds.), pp. 49-62. Kyoto Lniversiry, KyntoiX~nlhwestUilir. Zacllariassc, W. J.; Zijderveld, I. D. A,, Langcreis, C ti., Hilgen, F. J.. and Vcrhallm. P. J. J. M. (1989). Early Late Plioccnu hiochronolog? ?.ad suriacc water temperature variatirms in the Mediterranean. !Mar lirrt,pnleunloI, 14, 339-355. Zachariilsr: W. J., Gudjonsson. L.. Hilgn. F. I., Langereis. C . C., Luurm*, L. .I.. Vrrhallen. P. J. 1.M.. and Zijdcrvcld. S. D. A. (19901. Late Gauss ro carly Matuyamu invasiwls of h'coglohoquadnnn ai1antir.u in ihr Mcdirerranean and assuciatrd records of climate changc. P ~ l e n c e a r r ~ , ~5,~23g-252. ~~~hy Zhcng, H., An, Z., and Shaw. J. (1992). Ncw contfihutions tn Chinese PliO.Pleistoccnc magneloatratigraphy. Phys Eardl Plnn~r.Inter 70, 146-133. Zhou, L. P.. Oldticld. F.. Wintle. A. G., Robinson. S. ti.. and Wanp, 1. T. (1B0). Partly pedogcnic origin of magnetic variations in Chinese locs3. Narure (Londnn)346,737- 739. Zhu. K. X.. Thou, L. P..Laj. C.. Mazaud. A . and Ding. n 1.. (1994). The Blake geomagnetic polarity episode recorded in Chinesc lams. Grophy,$.Re5. Lett 21,647-7W. Zijdrrvrld, I. D. A. (1967). AC drmapelifi~tionof rocks: Analysis of rcrults. In "Methods inPaleomagnctism" ( 0 . W.Collinsrln. K. M. Crrer,andS. K. Kuncom,cdr.),pp.?54-286. Elsevier. Amsterdam. Zijderveld, J . U. A., 7,achariassr. W. J.. Verhallen. P. J. J . M.,and Hilgen. T. J. il986). Tllc agc of the Mir~cmr-Plioceneboundary. N ~ M Srrmgr / 16, ltil)-181. Zijdcn~cld,1.D. A . Lansereis. C. G., Hilgcn. F. J.. Vcrhallan, P. J. J. M., and Zachariarn. W. 5. (1991) Idlegrated n~agnetostratigra~hy nntlhiortratigraphy ofthe Lppcr Pliocenelower Pleistocene from thc .Wonto S i n p irnd Crolune areas in snuthcrn Calahria (Ilaiy). Earth Plnrrcl Sci. Len. 107. 697-714.

lnd ex

335

Index

Bcninslan slralorypr scudon. 193 RerriasianiValangi~tia~~ boundary. 203 Big Ir,sl Excursion. 246 Binghilm distrihurion. 60 Bi
Clarcndooisn. 115-146 and Bentovian fos5r;ils. 145 Clarcndunian-Hcmphillinn. 146 Clarks rurk hiisin. 159 Climate-canln~llcd'lilhnpaphio uariahility. 273, 276 Ciocliwise looplnp, ot secular %arldtlon 239 CM25. 175 Cobb Clountaln subchron Ui-YO 24'1 Cochltl Y6 Crrsrute force (HcJ. . , 28 256 rclcrcivitp <,l rrmiinencc. 28. 51 speclril. 56 Cocrcivityllrmprratura. 72 Collisiun ill India and Eurasia. 160 Conlponcnr ill magnrlira~i~rn. 54 Cancentrati<m of thc mngnctic minrriil. 252 paramctcr i k ) . Ki ot scry tin< erilincd [SP) magneti~c,

253 CaCOq cycles. liM Calahria. 102 Camhriiln ape. 238 Cilmhm-Ordor-icisn houndary, 228 Campanian Mnasrrlcrian houndury, 204 CarnianiLadinian hrrundiiry. 212 CarnianiEorian hr,undary. 208 Cenozoic ii"0 recurd. 136 CENl'Y4. 176 Ccnrml Africa. 155 Changes in climate. 267 Charrpxingian. 216 Chapadmalal. 155 Chartian, 110 Chemical drmapnctizution. 54

Cl~rmOslratigraphy,l l h China. 154 Chinese loess. 151. 245 Chnm 26n. 132 Chron ~~omenclaturc. 78 Chronogram tcchniqilc. 95 Chronngrsms. 170 C:lrclc 01 YS'C cunfidence~(uo;),6i1 Cinnon section. 189

Conglomcrnrc IcxL. 47 Cnnjugntc rnagnrlic ani>rnslirs. 117 Continental i u d h e d s 46 Continuaus thcrmal drmi~gnelixi~lii,n oi IKM. 70 corc expansion. 251 R C 9-181. 104 Conelatiun of Late Jurassic-Crcraccow Stage. 102

of polarity chrons. 1Y3 with the nxygcn iroropc curve. 263 Covariance of 6'"U ualuc%,'I40 Cletaaous long normal supcrchron. 24. 18?. 186 megnnic stratigraphy, 1x2 qllict ronc, 2311 C'rcraccunr-Terliarv ( K T boundarv. 114. IS'), 1x2 Critical anelc. ti3 Cmm-carelalion lechr&;quz,241 C:ryngcnic magnrlometer, 49 Crypl
Declmatlon, Y, 16 Dcconualul~~m, 269 Deep Sea Drllllng Pro~rcr(DSDP), 126 Deep sca sediments. 88 Dcglaciation of Greenland, 262 Dclta

Duratlon of polarll? chmnc, 230 Dvnamo process, 1 Y

~

,?r

record f a r thc Oligxene from DSDP Sitc 522, 142 record. 140 isotope stages 1 to IX. 142 "0. 130 isotope event Mi7. 164 stratigraphy, 136 timc scale, 105 Demagnctbalian spectra, 55 Deseadan lend mammal age, 156 Detrital rcm.anmt magnetization (DRMJ. 35,45 Dentzric oxidatiun. 30 Dewey Lakc Furnation. 218 Diagcnetic alteration. 42 groqth of rnaghzmite, 43 Diamagncrism. 26 Dinosaur fauna. 159 Dipolar transilion fields. 21 Dipolc hyp~lhrsis,15, 17 momenl, 11 Uolomilization. 101 Donetz hasin, 222 Drop in sea level. 162 usnp Leg

68, 99 72. 130

Duchcsntdn, 156 and Chadton~an.156 mterchange, 163 Dunkard Furmat~on.219 \uhchron. 219

Early and late Rarslovian. 151 Aptiam, 171-172 Hcmingrordian. 164 EarlyiMiddlr Triassic, ?14 Emene boundary. 136 Ehro hasin. 154 Eccrnlricity cvcles. 109 Flapbar 164 Flle\mrre and Axcl Heghe~elalands, 214 End of the PCKS. 210 Fnvironmental changes. 251. 268 E(~cnei0ligoccnuboundary. 136. 138. 140 Eolian accun~ulationrate. 263 derrital rnapnelite. 44 flux, 204 grain lize. 264 Fquill duration of arnmonile zones. 177 Eq:4rrrrs, 156. 104 at Hagcrmann, 145 Ericu cxcwsion. 243 Eureka Sound Formation, 160 Evidence for rcvcrsal of Lhe field; 5 Evolutionary fint appearances. 145 Exsolution lamellar. 31

faunal diachoneity, 160 dispersal. 160 interchangc 160. 166 migratiun, 144. IM Fayurn, 155 Ferrimagneiic graiasi~r-srnsiliveratios. 260 Ferrirnagoctism. 28 Fcrromagnrlism, 27-28 Field reversal. 3. 4 sampling. 85 Finitc rotali~mpolss,117 Pin1 rlrder migratory cvcnls, 163 Fihher statistics, 58-60 Flooding of the Mediterranean, 146

Formation of unconformities. 162 Frequency-dcpmdent magnetic susccptihility (kfl,253. 255 Future prospects For rock magdetic stratigraphy, 269

Gamma proccsr. 24 GausslMatngarna boundary. 155-156, 161 Geomagnetic equator. 12 excursions, 242 held, 14 intensit!, !!nrs, 23 paleointensily, 276 polarity reversal. 19 polarity time scalc (GPTSI. 1. 74. 113, 115 Gd%asubclvon, 96 Glac~al eplsodcs. 140 eustasv. 161. lh3

Global event. 2.45 (;lohigempsis ku~ieri.P O Gl~~l~t~erinelloider hlowi, 171 Goethitc, 33 Gothenhcrg excursion, 243 Grain s i x rprctrnm of fcrrirna,p.netic grains, 2 h3 Grain \ire, 252 Granulometric paramrrerr, 252 GRAPE. IVY Great Vallcy Group, 173 Greigitc, 34 Gricshachian, 215 Guadillupian or Wordian, 218 Guhbio (Bottacione) srctiott. 182 Gubbio/Comcssa hections, 132-133

Hautc-Savoie, I54 Hauteri\,ianmarrcmian boundary, 193 Hawaiian block model, 175 Iinsalions, 169

Heinrich evenrs, 257 Hrlveticlimestones, 38 Ftematire. 31. 62 Hemingfordian-Rarst~pvian boundary. 1% Hemipclagic sedimmts, 44 Hcmphillian. 146, 155 HcmphilliilnlBlancan boundary. 164 Hctcmchroneily. 159 Hcttangian, 176 Hcttangian-Oxfordian, 176 High rr5oludon magnrlic ~tratigraphv~ 276 vtrdligaphic correlation, 276 High-amplitude sccular vari.xlion. 249 High-frequency field, 253 Hipparion, 146 fauna. I51 HIRM, 253 Hoffman-Day rnzlhod. 56 Holc 68YB. 133 6WR. 133 702R, 133 Homo rrecnrs. 155 bluayquerian, 155 Hydraulic pist<>=corer (HPC). 99, 1117. 126 Hgperzoncs. 205 Hysteresis loop. 27.65

Icc-rartcd detritus (IRD). 257 Ice shcct insrahility, 259 Icc volume change, 136, 140 Jllawara reversal. 215 llrnrnite. 3U Immigrant rara, 161 Inclination, 10 India. 146 Indian Ouead. 130 Induced rnapnctizaticm. 27 Initial susccptihility, 27 Illncr core, 23 Intcrralion of ~hcmmtralieraohvand

IRM acquiritinn. 62

-

ISEA reverse polarity zone, 189 Isopuric charts. 13 laothermal rcrnanrnt magnctiration [IRM). -76, 254

Isotopic stagc 19 10s lcthmur of Panama, 1%

Japilnese. Hawaiian, and Pheonix Msequrnee lincalions, 168 Jaramillr~,95 crcck. 7 Junialil Fm, 225 Jurassic anomalies. 169 ma~ncticstratigraphs 196

Lena Rnver. 2ZX ~ e n g t h sof magnettc intcrvdb. 230 Lcncatrd m a p r t l c anomal~cs113, 168 Llanrlrn-C arddoc. 225 Lock in zone, 39 Loess deposit.;. 46,265 LOW-irsqucncyticld. 253 I .rie-Fuller tcst, 66

hlO-lvE9.I64 hl-srauence

houndary. 193

K-Ar ages. 107 K:'l' Boundary magnctic slratigraphy. 160. 182 Karna and Reunion chrons. 95 k,. 252 k , , f C 255 k ,,.,,,SIRhl. 263 karur Tcpc magnctuxmes. 208 Karanian stagc, 218 and 'Parrarian lime. 21 h Kenc and Gradstein timescule. 1711 Kinman magnetic iatcrp-a1, 215 Ki-eridgian-Oxfordian boundary. 175 Kinmeridpian-Tithonian mapnltic stratigaphy. 196 Kimmtrid@aniTithonian boundary. 196 Krakow L:pland, 2lXl

,rdmcnt [magnct~cproperhes), 45 265 Tahoe, 244 Windernierc. 113 Land mammal ages (I.F/IAs). 115 Lilschamp tncunion, 244 Late Arikareen. I64 Focenc, 156 Midconr carbon shift. 142, 164 Least squares fitting, 56 Leg 142, 152

Magnetic anumalieh I l l . 115 cnhancemcnl. 265.267 cpnchs, 578 aquatar. 10 excursi~ms.20.242 grain size (from ARMlk). 264 mineral Hur, 269 mineralngg. 253 mincrala. 28-24 parameters. 251 polarity stratigraphy. 1 plrle. LO properties of m ~ ~ i sedimems, nr 37 stratigaphy. 75 susccptihility (k).250~251 snisatropy. 25 Maenelite. 29 Magnetization components. 50 Magnetoh~ochzmochrorlol~y, 143 MaC"etahi~~cl1rnnologp, 81. 132, 136. 143 Magnctometers. 4 Magnelostrat~graph~c cla~aiticat!un.75 polarity clarrification, 75 rcvrrsal h<,ri~ons, 75 units, 75, 78 section, 119 studira of tcrrrstrial sequences. 144 t ) ~ szcrion, c 81,203 units. 75 Magneto?ons, 76

338 Maklica Limestonr Formation, I91 Mammal dispersal: 160 faunas. 159. 161 Mammalian hiochronology, 145 Mammoth subchron. 95 Mam?nrrrhus in North America, 145 Mantle plumes. 132 Marine records of secular variation. 278 ~cdimenrs,253 Mariposa Formation, 175 Mas~ifCentrall,F~rancc.244 Massimano section. 136 Ma\ter secular banatlon curve for Brxtaln 234,219 Matuyania. 3 Miluch Chunk Formation, 222 Maud Rise ( O W Leg 113). 133 Maximum angular drviation (MAD), 56 Mayrrling scctinn (Austria). 212 McLaughlin cyclcs. 207 Mclt event. 261 Me,vzo~cplattorm l~mr%lonc, 38 MIL, 140 161 M17 140 161

Middlellatc Eocene. 136 Midland Valley of Scotland, 222 Migration of hydn>carhons.38 Migrationary first appearances, 145, 163 Migrato~ycvcnts. 163-164 Milankovivh cycles, 140 cycl~r\tratigaphy,207.208 Miocene magnctir stratigraphy. 126 rnagnelohiochronolo~, 130 Mincenc-Pliocene boundary, 102, 105, 146 MNI. 154 Moenk~rpFarmallon, 212 Mono Lake excuwon, 234 243 Monlehrrmosan, 155 Monte harhone F'~mafron.I02 Monterey event. 142 Monn trans~tron.70

~ndex

Morr~sonFormal~rm,2W Morrowan paleosols, 220 MP29 154 MP30. I54 MP3WMX1 hnundary, 154 Mu metal shield* 54 Mururoa Atoll. 101 Muschclkak Formatlnn of Spam, 214

Namurian. 220 Nannou~nidcriris, I72 XannoIl
Ocean Drllllng Pronnrn ( 0 0 P ) . 126 Ochoan 21% Octupole. 24 transltlon held geometrlcs 19 ODP Hole9 845A and 8458,13U Lee Slle 710. 170 011 and 012, 140, 161 Olduval chron. 95 Gorge, 7 Olyocene, 156 Ohgocen-Msocene hounilary, 130. 154 Oma arca. 155

Orbilal cycirs ulrccrntricity. obliquity, and precession. 107 Ordovician. 225 Orientation, 49 Orleanianihrtaracian boundary. 152 Orthogonal projections. 54 Oxfordian, 176 Oxfordian-A~tian time scales, 175 Oxiordian-Callorian magnelic stretigraphy. 2w OxfordianKimmeridgian houndsry, 196. 203 Oxygen isotopc zones, 140. 161

P-wave propagation, 23 Pacfic anomalies, 168 Pakistatl, 146 Paleocene-Eocene boundary, 136 Pabbclimatology. 160 Palrugme age, 159 magnetic stratigraphy. 132 Paleogeography, 160 Pakointcnsity dctcrminations, 270 high. 241 Inws, 273 magnet,= strattgraph). 2,273 Paleomagneta pole, 16 Paleopreclpltauon. 265 Palcogcnc magnctlc strstlgraphy, 132 Pnramagnct-tlc~ n n s 26 . Paramapetism. 26 Parameter estlrnatlon, 64 Pass through cryogemc magnetometen. 87-88,234 P a s v r mam n unconfrlmtt~ra.161 u Paterson Volcanlca 220 Pattern bt. 91 Pedogcmc and dcrr~talsusceptlb~htv,265 Pelagnc deep-sea sed~ments,101 h e s t o n e s , 39 Per~od~crtv of -100 kv. 105 Pcrmo-Carhunnfrroua Reverse Supcrchron (PCRSI. ~. 215 Perno-Triassic boundary, 214-215 Physical disturbance of s e d h c n t ~243 , Pictou scdirnmts, 219

Piston core collections. 99 PlatIc>m carbonates. 101 Plienshachian-early Bajocian inlewal, 201 Plio-Pleistocene boundary, 102 climatic cycles. 154 glaciation. 145 lakc and loess sequencer. 145 land mammal sequences. 144 magnetic stratigraphies, 101 polarity chrons, Y5 Polarity, 9 hias, 215, 229, Z32 chron norncnclaturc. 75 transition rrcrrrds and VGP ~ a t h s20 . zone, 76 Potwar basin, 167 Pre-Callovian Jurassic magnetic straligraphy, 200 Prcccssional hxcing, 125 variance. 263 Precision parameter, 59 Wcfcrrcd longitudinal VGP paths, 21 PTingle Falls excursion. 245 Prohahilily density P, 58 Progressive deforestation, 2h7 Pyrrhotite. 34

Quadrupole, 24 Qual~tycrlterra 93 Quiet zonc. 168

Rates of sediment accumulation. 88 Red clay facies. 43 Rclativc gcomagnctic palcointcnsitics, 272 Relaxation lime. 34 51 Remagnetizatim circles, 56 Rcmanent cuccivity. 28 magnetizalion. 27 Rebetltive waveform. 2% Reproduc~blhtrof magnetozones. 89 Rr%rrlutlonof ilstrnchronoloplcill tlmescale 112 Resultant direction (R). 59 RCunion event. 96

International Geophysics Series EDITED BY

RENATA DMOWSKA Division ofApplied Scieprce Harvard University Cambridge, Mussuchusrm

JAMES R. HOLTON 1)cpartnlent of Atmosphrrrc Scrences Unzvercrly of Warhrrrgton Seattle, Washington

Volume I

BETOGI;TEXBER(~. Phpics 1)f the Earth's lnleriar. 1959*

Volume2

~ J S E ~W. H

Volume 7.

S. K. K u ~ c n R s(rd.) Cantinental Drifl. 196?*

Volr,lrcmeJ

C. E. JUNFE. Air Chemistry and Kadioactivlty. 1963'

Volume 4

KUHF.HI. G. F I ~ A G AS"I C 1 0 0 s ~A. BUS~SOZR. A n introduction to Atniosphcric Phpicr. 10hIx

Vo/,(urnr6

L. D u r o u ~.AND R. DEF.&Y. 'Ibcrmodyn~miclnf Cloud-. IOhl*

Volume 7

H. U.Rol

Vr,lr
RlcrrnD A. Cnalc. l b c Clppcr Afmosphvie: Mctn~n,lopyilnd Physics. I965*

Volume 9

W r ~ ~ L. r sWrno. Structure of the Stratosphere and Mcsoiphcrc. 196h'

Vo'oDsnr I0

MICHFI.FC:AYIIII. The G ~ i l ~Field i t ~ O[ ths Earth from Classical nnd Modern Methods. 1967"

V

S. M a r s u s ~ nawn ~ WAN r.Ac6. H. C ~ ~ ~ 8 c (eds.) . l . r Physics DI Geomagnetic Phcnorncna. (In two v<,lumra.) 1967"

I

Volume I 2

CHA~IBERLAIN. Physics of fhc A111ora and Airglow. 1961'

I..

Physic+ or the Marine Atmosphere. 1965'

K. YA KONDRATYEV. Kadiation in

-our ui Prillf

tk

Atmuspherc. 1969-

Internattonal G e o p h v s ~ aSenes Volmrre I3

E. P I I ~ I E N *st>C. W. NE\VTOC. A t ~ ~ l o ~ p hCirculation ~ric Sysrcms: 'l'hcir Struclure and Physical Intrrprctntion. l')h9*

Vollr,rrc -76

W. C ~ A Y B E R W N IU N DONALD M. HIIN~~N. n l e o r y o f Planetary Atmospheres: An Introduction t o l h r i r Physics and Chemistry, Srcund Edilion. 1987

JIISFPH

Voltrm~I4 I Irxx?-R~sllniTrlm u OWENIC,G,mmm.rl Introductio~~ to Iml~xphericPhysim. 1969%

knlrirnc 77 1 A Jncoer I hc E ~ r l h Core, '~ Sccond Ed~ltrln 1987

Voiulnr 15 C . S. RAM~AGC. MOIISOO~ Meteorology. IU71'

\/oItfime 38 J. R. Ark!. Principles of Occiln Phykiur. 1987

Voleme 16 i x u + , s R.

Volume 3Y MAallN A IJ\t\or The Llghtnmg D~\chdcge 1987

Voltrrne I 8

T T O l m N . An

Introduction t o Dynamic Meteorolog?. 197ZC

4 N m C.ON%I\Y B. ZCOVY. Middlc AtmoVolrcnw 40 Uavw G . A s n R r w r Jnblrs K. HULLON. sphcrc Dynamics. 1987

bL I, R u i ~ u r oClimate . and Life. 1974*

V,,lurnr, 1 0 MELVLN E. S I ~ K NOcean . Circulatim~Physics. 1975

I

UavrnH. M I LI rn. Walrr at ihe Surface of 1heEaFtli:An Introduction t o tcosysrcm Hydrmlyoamic~.1977

Vol,iu,tlc 22 lustvH W. C H A M I I F R L Theory ~ ~ ~ - .of Plallctary Armosphcrc%:An Tnlroduclion to Thrir Physics and C1~cmirtry.1978'

VoJr#mr24 ARxrrr S D ~ N N\Ccathcr I ~ Mnd~firalir~n i-n Cloud Szeding I Y 8 U I.'oirrrni. 25

R o o ~ ti ~ rFLEAOLIA\,) JIIOF~A BI P h ~ ~ ~ Sccond ~ c s . Fdltli,n 1980

FINGER

i\n llltroductlon 10 A t m o ~ p h ~ r I ~

L'ullmie 77 D4vm H. M I I . I . ~Pnergy R. at thc Surface of rhc hsrth: A n Introduction Lo lhc Energrticr <,ICcosyslemr. 1981

Vr>l!~nze 29 M I BLDYKO The Earth 9 Chmntc Pilrt dnd Fulurr 1982' 1 1

.

7

ADRT..\N E. GILL.Atmasphcrc-Ocean Dynamics. 1982

3 P ~ o l o~ N Z A N Ul)vft,rmdl~r,n\ ~ r m l Elastxc Earth 1Y8?*

Vul~rnie32 RowaLnI. MF~n1l.l. A N O MrnrdnL\V. MCEII~CNNY. Ihc E a ~ t h i sMagnctic Field. Its Hislor).. Origin. 2nd Planetat!' PcrspccIi~,~. 1983 Volirm? .i.iJnHN S. Lrwa .A\" RONALDti. l'alN\. and Evolutia~l.1983

Plnnzb and Their Atmospheres: Origin

li,lirn>e 74 ROLFM E L S ~ The N ~rirnuncntal K Crust "% tilophys~calApproach 1986 Volunlc .B

M.U.

Sar;llllv,

D.B o u ~ r V, . S, N.*z&a~Nno, .A\n

Grarimdr). 1986

W a n ~ r c r Chemistry . of the Natural Alm~rsphrre.1988 Volornr 41 PETER

Volrrmv J2 S. PRI,A R Y * Inlroducticx> to Micromctcorolrrgy. 1988

V,,l,iun,e LO J . A. JALOKS. Thc Earth's Core. 1975:"

1

345

KH. G. TAIIPHITIIN~\.. Lunar

V~?lt'nrr44 WII.LLIM R. COTTON N I , R I C H ~ R DA~THES. A. Storm and O o u d Dyna~~>ics. 1989 Volrnre 47.

WJI.T.IAM MLSKE.ticoph>~icalDxla A ~ ~ a l y sDiscrete i~: I~YCTSC. Thcury. RevLIIed Edition. 198484

l/nlzl,,zu W

S . GEoacE PHlt.avllb~.El Niiio, La NiAa. and the Southern Oscillation. 1YYU

I.'rrliane 47

ROBERT A

I

J 4 ~ K 5

t

R K I ~ h~ iN d Mechanics of the Almcnpherz 1991 HOTTON 2\11 lntroductlnn 11, Dvnamc Meteorologv, Third E d ~ u o n

1 9'12

V o l r ~ a r49 ~ A l . E x s o ~ nA , K A U ~ MGeophysical AV. Ficld T h r t ~ qand Method. Part A: tiravitati<,nill. Electric, and .Mngnrtic Fields. 1992 Part B: Ebclr~rmagncticFicids 1. 1094 Part C1: Flrclroma~neticFiclds 11. 1994 S. BUICIILR, GUNDCJN 51. ORI,UUS. ROBERIJ. CH\RLX)N.A N D V. W n ~ r c Global . Bioguochrmicvl Cycles. 1992

Voioll4me 50

Si\\rtlb.l.

Volume 51

O R ~ AENV , ~ N ASN D ' ~ F \ G - FWOSC. O N G F w l t Mechanics and Transport Properties of Roclts. 1992

GoKI~ON

~ N . I ~ l t r . + \ ~ Remote ~ ~ l e t bcnslng 1092 Vol~rmr52 R O s m F H I ~ F ~Atmosphcnc

RIP 53 K D H F KA.~ HOUZE.IR. Cloud Dynumic~.1993 Vqliwie .54 P E I ~V.R HOODS. .4~ra8oI-<-li~ud-Climate i ~ l t ~ r a c l iIcJY3 ~n~. Voltzmw 55 S. J. G I B D ~ X ~A N CnZ A. KIJKO.i l l , Introductirin tr, Minirig Scismolos?. 1Y91 Voliinre F6 D E N N ~I..S H A K I M A NGlobal N. Physical Clirnatolngy. 1494

346

I I

Voltcme 5'9

I n t e r n a t o ~ Geophysics l Scrlcs

D a ~ t t t5. . WIIKS.Statistical Methods in thr Atrnoxpheric Sciences. 1995

Vnlumr 60 Fnt~rRlKhLoLKm Calculating the Wcvther 1995

Volirme 61 MURKY L. SALBS.Funda~llentalsof Atmospheric Physics. 1YYb Volutnr 62 JAMES P. MCC.,\LPIN Pdeoseismolo~g.1996.

I

I

l'olvmtc 63 Ro~a1.nT. MERRILL, MICIIACL W. ~ I c E L ~ ~ T~N SS Y PIIIUIT D~ L. MCFADDEY. The Mayetic Field of the Earth: Paleomagnetism, the Core, and the Deep Mantle. 1996.

ISBN 0 - 1 2 - 5 2 7 4 7 0 - X