Pure appl. geophys. 157 (2000) 1227–1256 0033–4553/00/081227–30 $ 1.50+0.20/0
Tsunami Generated by the Late Bronze Age Eruption of Thera (Santorini), Greece FLOYD W. MCCOY1 and GRANT HEIKEN2
Abstract—Tsunami were generated during the Late Bronze Age (LBA) eruption of the island of Thera, in the southern Aegean Sea, by both caldera collapse, and by the entry of pyroclastic surges/flows and lahars/debris flows into the sea. Tsunami generated by caldera collapse propagated to the west producing deep-sea sedimentary deposits in the eastern Mediterranean Sea known as homogenites; open-ocean wave heights of about 1.9– 17 m are estimated. Tsunami generated by the entry of pyroclastic flows/surges and lahars/debris flows into the sea propagated in all directions around the island; wave heights along coastal areas were about 7 – 12 m as estimated from newly identified tsunami deposits on eastern Thera as well as from pumice deposits found at archaeological sites on northern and eastern Crete. Key words: Tsunami, volcanic eruptions, tephra, pyroclastic surges, pyroclastic flows, debris flows, lahar, calderas, tsunami deposits, pumice, ash, Late Bronze Age.
1. Introduction An explosive eruption on the island of Thera, also known as Santorini, in the Cycladic Islands, about 3600 years B.P. during the Late Bronze Age (LBA) had profound effects in the Aegean and eastern Mediterranean region. Local effects included the devastation of the island and a civilization that inhabited that island leading to the demise of cultural centers throughout the region, such as the Minoan culture on Crete (thus the eruption is often referred to as the ‘‘Minoan’’ eruption) and the Cycladic culture in the Cycladic Islands. Regional effects included widespread dispersal of ash, possible climate change associated with the injection of ash into high altitudes, seismic activity, dispersal of pumice rafts by oceanic surface currents, and generation of tsunami. Evidence comes from both the geological and archaeological record of the circum-Mediterranean area. Deposits of tephra are widespread throughout the Aegean and eastern Mediterranean region from the Nile 1 University of Hawaii, Windward Community College, Kaneohe, HI 96744. E-mail:
[email protected] 2 Los Alamos National Laboratory, Earth and Environmental Sciences Division, Los Alamos, NM 87545.
1228
Floyd W. McCoy and Grant Heiken
Pure appl. geophys.,
Delta to the Black Sea (various papers in HARDY, 1990; STANLEY and SHENG, 1986; GUICHARD et al., 1993) and provide dramatic documentation of the explosivity of this Plinian-type eruption. Atmospheric temperature decrease of about 0.5°C in the Northern Hemisphere is estimated from sulphur aerosols injected into the stratosphere (SIGURDSSON et al., 1990). Seismic activity possibly associated with the eruption caused widespread damage to LBA Minoan buildings (DOUMAS, 1983; various papers in Hardy, 1990). Pumice accumulations in coastal deposits and archaeological sites throughout the eastern Mediterranean and Aegean littoral suggest extensive rafting of pumice following the LBA eruption (FRANCAVIGLIA, 1990; and others). That tsunami were generated has been documented by the discovery of a distinctive sedimentary deposit, homogenites, interbedded in deep-sea sediments of the Mediterranean Sea (CITA et al., 1984; and others), by chaotic accumulations of tephra and other sediments at archaeological sites in the Aegean region (ASTROM, 1978; NEEV et al., 1987; FRANCAVIGLIA, 1990; and others), as well as by newly identified deposits within the LBA eruption sequence on Thera (MCCOY, 1997; MCCOY and HEIKEN, 1997). Geological and geophysical studies on Thera have lead to an understanding of the eruptive mechanics, vent placement and topographic changes associated with the eruption, and inferences for the generation of tsunami (BOND and SPARKS, 1976; HEIKEN and MCCOY, 1984; DRUITT et al., 1989; SPARKS and WILSON, 1990; and others). Tsunami generated during the eruption has been assumed by archaeologists as one of the primary destructive effects of the eruption, although identifiable damage to man-made structures by tsunami remains elusive (MARINATOS, 1939; PLATON, 1971; DOUMAS, 1983; BARBER, 1987; various papers in HARDY, 1990; and others). This paper reviews evidence for tsunami from the archaeological and geological record, and presents preliminary data on volcaniclastic deposits found on Thera that are possibly the result of tsunami inundation. Possible tsunamigenic phases of the LBA eruption are discussed with inferences for wave propagation into the Aegean and eastern Mediterranean Seas; estimates for tsunami wave heights are made from sedimentary deposits considered to have originated from tsunami passage at sea and impact on land.
2. Background 2.1 Geologic and Tectonic Setting of Thera Thera is one of several volcanic centers active during the Quaternary along the Hellenic island arc of the southern Aegean Sea (Fig. 1). Five islands surround (Thera, Therasia, Aspronisi) or occur within (Kameni Islands) two water-filled calderas to form the archipelago of Santorini (Fig. 2). Thera, Therasia, and
Vol. 157, 2000
Tsunami Generated by the Late Bronze Age Eruption of Thera
1229
Aspronisi are remnants of a larger island destroyed during the LBA (Minoan) eruption. The Kameni Islands have been constructed by post-Bronze Age volcanism, the latest eruption in 1950 on Nea Kameni; these islands today have manifestations of volcanic activity with hot springs, fumeroles, and low-level seismic activity (FYTIKAS et al., 1990). Volcanic activity during the past 700,000 years constructed pyroclastic cones and lava shields of dominantly dacitic composition, to form the Thera volcanic field (PICHLER and KUSSMAUL, 1980; SEWARD et al., 1980; HEIKEN and MCCOY, 1984;
Figure 1 The southern Aegean and eastern Mediterranean Seas showing the locations (circles) of volcanoes along the Aegean island arc; the circled location northeast of Santorini is the underwater volcano at Colombo Bank. Islands mentioned in the text are identified. Large open arrow indicates the direction of maximum tectonic closure between the African and Aegean/Europe plates. Lines and patterns, from north to south, indicate: dashed lines (Aegean Sea) = 150 (northern line) and 100 km (southern line) contour on the subsurface depth of the subducted crustal slab beneath the Aegean arc; dotted lines south and west of Crete =sea floor troughs (Hellenic ‘‘trench’’ system); fine stipple pattern through the central Mediterranean Sea = Mediterranean Ridge, an accretionary wedge of sediment produced by the closure between Africa and Europe; thick dark line with sawtooth pattern = approximate position on the sea floor of the downgoing slab. Tectonic indicators are from LE PICHON and ANGELIER (1979), RYAN et al. (1982), KASTENS et al. (1992), and MEIJER and WORTEL (1997).
1230
Floyd W. McCoy and Grant Heiken
Pure appl. geophys.,
Figure 2 Generalized geologic map of Santorini. ‘‘Basement’’ refers to Mesozoic metamorphic rocks; the Akrotiri volcanic complex is a sequence of pillow lavas and associated sediments of the Pliocene age. Megalo Vouno volcanic complex, the Micros Profitis Elias, Therasia and Skaros volcanoes are eruption centers active over the past 900,000 years. Data from PICHLER and KUSSMAUL (1980), HEIKEN and MCCOY (1984), DRUITT et al. (1989), and unpublished field data. The LBA eruption is often referred to as the ‘‘Minoan’’ eruption, after the Minoan culture that existed on Crete at the time of the eruption.
DRUITT et al., 1989; BARTON and HUIJSMANS, 1986; FRIEDRICH, 1994; MCCOY and HEIKEN, 2000). The latest pyroclastic eruption in this sequence was the LBA (Minoan) eruption that excavated a large caldera north of the present-day Kameni
Vol. 157, 2000
Tsunami Generated by the Late Bronze Age Eruption of Thera
1231
Islands and deposited a thick tephra layer on the islands surrounding the caldera (HEIKEN and MCCOY, 1984) (Fig. 2). These volcanic deposits are built upon Mesozoic metamorphic rocks and Pliocene submarine pillow lavas and associated sediments (Fig. 2). The Mesozoic terrain is part of the Hellenide-Anatolide orogenic belt deformed during the early Cenozoic closure of the African and European plates (MCKENZIE, 1972; LE PICHON and ANGELIER, 1979; FYTIKAS et al., 1984). An extensional crustal regime during the Pliocene, in response to this closure, formed the Aegean basin with the islands of the Aegean basin being the subaerial portions of a series of PaleogeneMiocene nappes (MAKRIS, 1978; BONNEAU, 1984; HALL et al., 1984; ROBERTSON and DIXON, 1984; ARMIJO et al., 1992). Contemporary plate-closure rates are about 10– 40 mm/yr., with intra-arc extension rates of about 3–8 cm/yr. (DEWEY et al., 1973; LIVERMORE and SMITH, 1985; ORAL et al., 1995; KAHLE and MULLER, 1996). A northwesterly-dipping Benioff zone occurs beneath the arc (Fig. 1) (AGARWAL et al., 1976; BATH, 1983; SPAKMAN, 1986). Seismic activity in response to this tectonic activity, as well as to volcanism, is centered in the southern Aegean Sea in the vicinity of Thera, often with destructive tsunami (BATH, 1983; TAYMAZ et al., 1990).
2.2 Generation of Tsunami by Volcanism and Slumping Tsunami created during historic volcanic eruptions have been documented, particularly those generated during the 1883 eruption of Krakatau. Accounts from this eruption come from both survivors and from later studies (see SIMKIN and FISKE, 1983, for an excellent compilation; see also SIGURDSSON et al., 1991; CAREY et al., 1996, 2000). Tsunami were produced by collapse of the caldera (WHARTON, 1888), entry of pyroclastic surges and flows into the ocean (SELF and RAMPINO, 1981; LATTER, 1981; SIGURDSSON et al., 1991; CAREY et al., 1996, 2000), slumping of parts of the underwater volcanic edifice (SELF and RAMPINO, 1981), in addition to the explosive interaction of magma and seawater (not considered as significant as other effects in generating tsunami by SELF and RAMPINO, 1981). Wave heights exceeded 40 m; runup distances reached 11 miles inland over low slopes (on the order of 1 – 5°) of coastal benches (SIMKIN and FISKE, 1983). Tsunami produced during the 1815 eruption of Tambora in Indonesia were generated by pyroclastic flows entering the ocean (SELF and RAMPINO, 1981); coastal inundation reached up to 4 m elevations (VAN PADANG, 1971; WALKER, 1973). The 3500 yBP eruption of Aniakchak volcano in Alaska produced tsunami with runups as high as 15 m over low slopes (coastal peat bog), apparently also by entry of pyroclastic flows into the ocean (WAYTHOMAS and NEAL, 1997). Tsunami generated by sea-floor displacement due to gravity-driven slumps and slides are well documented. The collapse of volcanic edifices to produce tsunami are the focus of numerous papers in this volume and elsewhere (MCGUIRE et al., 1996).
1232
Floyd W. McCoy and Grant Heiken
Pure appl. geophys.,
Massive underwater slides generating tsunami are documented, the best studied perhaps being the Storegga slides off Norway about 7000 yBP (BUGGE et al., 1987; JANSEN et al., 1987) leaving deposits in Scotland (DAWSON et al., 1988, 1993) and Norway (BONDEVIK et al., 1997). Runup heights from this tsunami are estimated at up to an elevation of 11 m (BONDEVIK et al., 1997). An underwater slide in the Aleutian trench south of Alaska in 1946 produced devastating tsunami with runups reaching 35 m in Alaska and 17 m in Hawaii, and with estimated maximum wave-heights in Hawaii up to 15 m (MACDONALD et al., 1947). Tsunamis generated by debris avalanches accompanying or following volcanic activity have produced waves of the order of 9 – 20 m in height (reviewed in MCGUIRE, 1996). In the Aegean, there are accounts describing historic volcanic eruptions and consequent tsunami. In 1650, as one example, Jesuit priests on Thera described with some awe a submarine eruption at Colombo Bank northeast of Thera (Fig. 1) ‘‘the ocean burned . . . , flames leaping from the sea . . . fiery clouds lowered upon us . . . cast up into the heavens enormous rocks, which fell to earth again some two leagues away . . . in a field we saw a boulder spewed from the bowels of the earth, of a size that 40 men could not move it . . . the sea covered with pumice . . . sulfurous steam’’ (GALANOPOULOS and BACON, 1969). Tsunamis associated with this eruption were apparently generated by caldera collapse; recent bathymetric surveys of Colombo Bank indicate collapse was on the order of 400 m (PERISSORATIS, 1985). On Thera, waves deposited pebbles and dead fish as much as two miles inland in valleys (GALANOPOULOS and BACON, 1969); tsunamis were reported to be 16 m high on the island of Ios immediately north of Thera (LUCE, 1969); at Patmos, northeast of Thera ‘‘the sea rose 50 m and 30 m on the west and east coasts, respectively’’ (GALANOPOULOS and BACON, 1969); on the northern coast of Crete, ships were torn from their moorings and sank in the harbor at Iraklion (the town was under siege by Turkish forces at the time and their ships were destroyed by the tsunami, thus ending the siege) (GALANOPOULOS and BACON, 1969; GALANOPOULOS, 1960; DOUMAS, 1983).
2.3 Obser6ations of Tsunami Wa6e Heights Accounts describing tsunami often use two different observations: (1) wave height, as would be observed at the coastline with the tsunami approaching, and (2) elevation or distance onshore of runup by the wave, usually easy to distinguish by limits of damage or of displaced debris inland. Wave height is a difficult measurement, potentially dangerous to the observer, and is reliably measured by high-water levels or damage levels marked on surviving structures after inundation—such measurements are rare; good data on tsunami wave heights are correspondingly rare. Runup is not a reliable indicator of wave height, rather is a function of numerous variables: slope and topography offshore and onshore; soil/sediment and rock types offshore and onshore; roughness factors determined offshore by reefs,
Vol. 157, 2000
Tsunami Generated by the Late Bronze Age Eruption of Thera
1233
bars, banks, etc., and determined onshore by rivers, hills, vegetation, buildings, etc.; in addition to wave characteristics such as wave type, height, angle of approach, onshore velocity, etc. (LOOMIS, 1976; COX, 1979). Experimental studies have found amplification factors of runup over wave height as high as 5.7 (SPIELVOGEL, 1973).
3. The LBA Eruption of Thera 3.1 Archaeological Setting In the archaeological chronology of the southern Aegean derived from potterysherd types, the eruption occurred, and defines, the boundary between the Late Minoan IA and Late Minoan IB of the Late Bronze Age (LBA) (see various papers in HARDY, 1990). Dates center around two intervals, 1630 BC and 1500 BC (see MANNING, 1995, for discussion of the controversy concerning these dates). Archaeological evidence provides a rich account of the effects of the eruption on civilizations in the region, particularly the Cycladic culture on Santorini and the Minoan culture on Crete. On Santorini, an archaeological excavation at Akrotiri (Fig. 3) uncovered what appears to have been a prosperous small city (DOUMAS, 1983) buried within tephra of the LBA eruption. In the aftermath of the eruption, both the Cycladic and Minoan cultures were replaced about two centuries after the volcanic disaster by the Mycenaean culture from mainland Greece. Whether this cultural change was due entirely to the eruption or to other cultural factors, such as war or economics, continues to be argued by archaeologists and historians—that widespread devastation did occur concurrently with the eruption from earthquakes, fire, and possibly tsunami, however, is clear in the archaeological record (see, for example, MARINATOS, 1939; GALANOPOULOS and BACON, 1969; LUCE, 1969; HOOD, 1971; PLATON, 1971; BARBER, 1987; DOUMAS, 1983). Certainly the consequences of an eruption of such magnitude near the center of two thalassocratic LBA societies would have had a major impact on their future. One need only compare the effects of equivalently large Plinian eruptions on post-Bronze Age civilizations to see the potentially huge effect of the Thera eruption and its regional effects on these LBA societies (see, e.g., MCCOY and HEIKEN, 2000).
3.2 Explosi6ity, Magnitude and Tephra Volume The explosivity index of the LBA eruption of Thera is estimated at 6.9 on the Volcanic Explosivity Index (VEI; NEWHALL and SELF, 1982) and as such represents one of the largest eruptions in the past few millennia (only seven eruptions have had higher VEI values in the past four millennia; SIMKIN and SIEBERT, 1994; DECKER, 1990; VOGEL et al., 1990).
1234
Floyd W. McCoy and Grant Heiken
Pure appl. geophys.,
Volcanic plume heights are estimated to have been up to 36 km (SIGURDSSON et al., 1990). Injection of ash to these heights, in conjunction with estimates of a high sulfur content within the plume, suggests that global climatic cooling for a few years was a likely consequence of the eruption (SIGURDSSON et al., 1990). The duration of the major part of the eruption ranged from six hours to four days with no extended interruptions in this activity, erupting at an intensity of 1.4–4.2× 108 kg/sec.; accumulation rates of tephra were as high as 3 cm/min. (PYLE, 1990; SPARKS and WILSON, 1990; SIGURDSSON et al., 1990). Volume of caldera collapse is calculated at 18 – 39 km3 (HEIKEN and MCCOY, 1984; PYLE, 1990); modern depths in the caldera are 390 m (PERISSORATIS, 1985) indicating at least this amount of collapse during the eruption, but certainly much more, perhaps as much as 690 m depending upon reconstructions of the pre-eruption topography of the island (PICHLER and FRIEDRICH, 1980; HEIKEN and MCCOY, 1984; FRIEDRICH et al., 1988; DRUITT and FRANCAVIGLIA, 1992).
Figure 3 Map of the southern portion of Thera showing the Akrotiri region and the archaeological excavation.
Vol. 157, 2000
Tsunami Generated by the Late Bronze Age Eruption of Thera
1235
Figure 4 Stratigraphic section of the deposits from the LBA eruption (adapted from MCCOY and HEIKEN, 2000; nomenclature follows RECK, 1936; HEIKEN and MCCOY, 1984; and DRUITT et al., 1989). Arrows in the right column indicate suggested tsunamigenic phases of the eruption. At lower right in the lithologic section, a man-made wall is depicted to indicate phases of the eruption that preserved (BO1 and lower BO2) then destroyed (succeeding phases) archaeological structures.
3.3 Eruption Sequence and Tsunamigenic Phases The LBA eruption occurred in four major phases, preceded by a minor precursor phase; eruption activity during three of these likely produced tsunami (Fig. 4). The discussion below is summarized from HEIKEN and MCCOY (1984), unless otherwise noted. A precursor eruptive phase of magmatic and phreatomagmatic activity some weeks or months prior to the major phases of the eruption deposited a thin tephra layer extending to 8 cm over southern Thera (Fig. 5a), and may have provided advance warning of the impending disaster to local inhabitants (HEIKEN and MCCOY, 1990). Bedding structures in the precursor deposit are suggestive of air-fall deposition with local dispersal by atmospheric winds (HEIKEN and MCCOY, 1990). Although a significant lithic component in the ash might indicate smaller collapse near the vent, there is no evidence of larger collapse structures during this precursor phase. This evidence in combination with the lack of major destruction to manmade structures (DOUMAS, 1983) and the inferred lower level of eruptive activity, suggest no tsunami generation during this initial eruption phase. Intense magmatic activity of the first major phase of the eruption (BO1/Minoan A) deposited up to 7 m of pumice and ash, with a minor lithic component, to the
1236
Floyd W. McCoy and Grant Heiken
Pure appl. geophys.,
Figure 5 Depictions of vent positions (suture line) and tephra deposits (fine stippled pattern) from the precursor and major phases of the LBA eruption of Thera; also shown is the modern coastline (solid line) and the approximate LBA pre-eruption coastline (dotted line) (modified from HEIKEN et al., 1990). Figure 5a. Precursor eruption phase (BO0).
southeast and east (Fig. 5b). Dispersal was by atmospheric and stratospheric winds; dispersal patterns and isopachs imply a vent north of the modern Kameni Islands, perhaps an extension and enlargement of the vent of the precursor phase. Archaeological evidence indicates burial of man-made buildings with limited damage mainly from collapsed roofs. Tephra characteristics indicate fragmentation and discharge of magma by magmatic gases in a setting where there was no water interaction at the vent. Such depositional mechanisms, in combination with the inferred vent position on land and a lack of evidence for major collapse of the volcanic edifice, are not those mechanisms and settings noted for tsunami generation during historic volcanic eruptions. Accordingly it is inferred that tsunami were not generated during this first major phase of the LBA eruption. Pumice deposited on the sea around Thera during this first phase of the eruption may have formed enormous rafts of floating pumice that drifted on surface currents throughout the Aegean and eastern Mediterranean Seas. For
Vol. 157, 2000
Tsunami Generated by the Late Bronze Age Eruption of Thera
1237
Figure 5b First major eruption phase (BO1/Minoan A).
example, pumice rafts following the 1883 eruption of Krakatau, a less explosive eruption than LBA Thera (VEI= 6.0; SIMKIN and SIEBERT, 1994) were as much as 15 ft thick and floated across the Indian Ocean transporting skeletons and trees (SIMKIN and FISKE, 1983). With the second phase (B02/Minoan B), however, the eruption character changed to phreatomagmatic activity with numerous thin (B 1 m) pyroclastic surges and flows leaving a deposit of up to 12 m; thickest accumulations are on southern Thera (Fig. 5c). Changes in bedform characteristics indicate an increasing velocity of surges and flows as second phase activity continued, which is reflected in the complete destruction of any man-made structures not buried within first-phase or early second-phase deposits (MCCOY and HEIKEN, 2000). Flow directions and the pattern of thickness variations suggest a vent sited south of the first-phase vent in a setting where water had unlimited access into the vent. By comparison with tsunamigenic phases of historic eruptions, such as Krakatau in 1883, it is entirely likely that tsunamis were generated during the second phase of the LBA eruption, particularly during later portions of this phase, wherever surges and flows entered the sea (assuming surge/flow bulk densities
1238
Floyd W. McCoy and Grant Heiken
Pure appl. geophys.,
Figure 5c Second major eruption phase (BO2/Minoan B); also shown are pyroclastic surge and flow directions (from MCCOY and HEIKEN, 2000), with inferred sites of tsunami generation and directions of wave propagation.
greater than sea water to diminish buoyancy and allow flow into the water rather than on the sea surface). Thicker accumulations along southern Thera indicate pyroclastic activity focused in this direction and thus tsunami perhaps directed more towards the south and southeast, although tsunami would have been generated anywhere these surges and flows entered the ocean along the outer perimeter of the island. Tsunami generated within the center of the island in the southern embayment would have had limited entry into the Aegean Sea through the narrow channel in the pre-collapse landscape (Fig. 5c). Pyroclastic flow activity continued into the third phase of the eruption (BO3/Minoan C) with the initiation of caldera collapse (Fig. 5d). The deposit from this phase of the eruption is a massive unit (up to 55 m thick on central-southeastern Thera) within which individual pyroclastic-flow units are not distinguishable. A high proportion of lithic fragments, some many meters in size, and the composition of these fragments indicate major collapse of the north-central portion of the LBA
Vol. 157, 2000
Tsunami Generated by the Late Bronze Age Eruption of Thera
1239
Figure 5d Third major eruption phase (BO3/Minoan C); also depicted is the area of initial caldera collapse (modified from MCCOY and HEIKEN, 2000, and HEIKEN et al., 1990), with inferred sites of tsunami generation and directions of wave propagation due to entry of pyroclastic flows into the sea.
island to form the caldera. Thickness variations and ballistic trajectories of lithic blocks indicate a vent situated approximately in the same position as that for the second-phase of the eruption. Late in this phase, lahars and debris flows were produced. As during second-phase activity, tsunamis may have been generated wherever pyroclastic flows entered the sea around the perimeter of the island(s). Tsunamis could have been larger than those produced earlier in the eruption because the characteristics of third phase deposits on land suggest a few massive slurry-like mixtures that were thick enough to flow over caldera walls close to 400 m high (HEIKEN and MCCOY, 1984; SPARKS and WILSON, 1990)—the entry of such a pyroclastic slurry into the ocean could have produced huge tsunami propagated in all directions. In addition, large tsunamis might have been produced by sea water filling the collapsing caldera, particularly if the two peripheral fault blocks to the caldera (Fig. 2) had partially collapsed in this phase of the eruption, thus widening the ocean channels to the north and east-southeast and focusing tsunami in these directions.
1240
Floyd W. McCoy and Grant Heiken
Pure appl. geophys.,
Figure 5e Fourth eruption phase (BO4/Minoan D); also depicted are areas of final caldera collapse, pyroclastic flow directions (modified from Heiken et al., 1990, and MCCOY and HEIKEN, 2000), and inferred sites of tsunami generation. The position of the tsunami deposit at Pori is identified.
The final phase of the LBA eruption (BO4/Minoan D) was marked by varied activity: lithic-rich base surge deposits, lahars, debris flows, and some co-ignimbrite ash-fall deposits (Figs. 5e and 5f). That tsunami was generated by this activity, towards the east by both a pyroclastic flow and a lahar/debris flow entering the ocean, is documented by tsunami deposits within the fourth phase deposit (described below; (MCCOY, 1997, 1997b). This phase also marked the completion of caldera collapse (BOND and SPARKS, 1976; HEIKEN and MCCOY, 1984; MCCOY and HEIKEN, 2000). Vertical collapse of at least 400 m in the caldera (perhaps up to 700 m assuming about 300 m elevation to the LBA island that was collapsing into the caldera), and collapse of the two fault blocks peripheral to the caldera by 300 m, also produced tsunami (vertical displacements relative to modern sea level using bathymetric data from PERISSORATIS, 1985). This is documented by sedimen-
Vol. 157, 2000
Tsunami Generated by the Late Bronze Age Eruption of Thera
1241
Figure 5f Fourth eruption phase (BO4/Minoan D); also depicted are flow directions of lahars, debris flows and lithic-rich base surges (modified from HEIKEN et al., 1990, and MCCOY and HEIKEN, 2000), inferred sites of tsunami generation, and directions of wave propagation.
tary deposits within the deep-sea stratigraphic succession of the eastern Mediterranean Sea that have been identified as tsunami deposits produced by waves propagating to the west and west-southwest (KASTENS and CITA, 1981; CITA et al., 1984; HIEKE, 1984).
4. Sedimentary Deposits Produced by Tsunami Generated during the LBA Eruption of Thera 4.1 Homogenites: Deep-sea Tsunami Deposits The formation of a distinctive deep-sea sedimentary deposit found in cores from the eastern Mediterranean Sea, homogenites, has been attributed to tsunami
1242
Floyd W. McCoy and Grant Heiken
Pure appl. geophys.,
(KASTENS and CITA, 1981; CITA et al., 1984; HIEKE, 1984). A single homogenite layer can be up to 9 m thick. Homogenites are homogeneous in color (thus the name; usually a light to medium brown-grey), soupy, contain a microfossil assemblage of mixed ages, have a thin basal zone of silt or sand, show normal size grading through much of the layer, and have no sedimentary structures except for burrows in the upper few centimeters (Fig. 6). Homogenites are distinctive with their homogeneous appearance, thickness, and color that contrasts with the pelagic stratigraphic section of the eastern Mediterranean of turbidites (white to brown), sapropels (dark grey to black), volcanic ash layers (light tan to grey), burrowed pelagic muds (light green to dark brown) and oozes (white). Homogenites are found only in cores taken of sediments from floors of enclosed basins on the sea floor. Homogenites appear in seismic reflection profiles as an acoustic transparent layer constrained to the basin floor (Fig. 7). They occur at the stratigraphic level of the ash layer from the LBA eruption; the LBA ash layer is not present in cores containing a homogenite. Cores from surrounding slopes of the basin show an unconformity/paraconformity at the stratigraphic interval of the LBA ash layer.
Figure 6 Typical sedimentological characteristics of homogenites from two cores taken southwest of Crete on the Mediterranean Ridge; thickness of both homogenite layers is close to 5 m (from CITA et al., 1984).
Vol. 157, 2000
Tsunami Generated by the Late Bronze Age Eruption of Thera
1243
Figure 7 Near-bottom 4 kHz seismic-reflection profile showing restriction of the homogenite layer (acoustically transparent interval) within and on a basin floor (from KASTENS and CITA, 1981).
These factors suggest slumping of sediments onto basin floors from the surrounding walls at the time of the LBA eruption, triggered by the passage of tsunami generated during the LBA Thera eruption (KASTENS and CITA, 1981) (Fig. 8). Tsunami would have had wavelengths adequate to provide near-bottom current velocities sufficient to erode and slump deep-sea sediments. In the intervening 3600 years since the eruption, compaction of the turbid bottom water ponded within the basin has formed the homogenites. The occurrence of homogenites only in ocean-floor basins southwest of Crete, and the alignment of these basins with pathlines calculated for tsunami propagation to the west from Thera (Fig. 9), infer wave generation by collapse of the caldera and the fault block west of the caldera (third and fourth phases of the eruption; Figs. 5e,f). Collapse of the latter would have widened and deepened the channel into the Aegean Sea, focusing wave propagation to the west.
4.2 Tsunami Deposits on Thera A 3.5-m thick volcaniclastic deposit intercalated between third and fourth phase deposits of the LBA eruption (see Fig. 4), near the village of Pori on the east coast of Thera (Fig. 5), has been interpreted as tephra reworked by tsunami generated during the eruption (MCCOY, 1997; MCCOY and HEIKEN, 1997). The deposit is a mixture of pumice, ash and lithic (lava) fragments, characteristic components of all phases of the eruption, in two distinct layers (Fig. 10). The lower layer has a
1244
Floyd W. McCoy and Grant Heiken
Pure appl. geophys.,
minimum thickness of 1.5 m and extends inland 36 m to an elevation of 6 m on a 2–16° slope. It has a lenticular shape open towards the ocean but which thins inland; internal bedding structures include planar and cross-bedded layers, some with normal size-grading, clastic tails in the lee of larger clastic particles, and layers of lithic-rich gravels, often in sharp contact with erosional features. The basal contact of the lower layer is sharp with erosional scour into the underlying tephra. The upper layer appears to be about 2 m thick; the upper contact is gradational and indistinct. It is a chaotic mixture of particle sizes and composition with no internal bedding structures, but with a higher content of coarser lithic fragments than found in the lower layer. The contact between the lower and upper layer is distinct with some features indicative of erosion. Detailed sedimentological work on the deposit continues and will be reported elsewhere. A tsunami origin for this deposit rather than one related to eruptive mechanisms is suggested by: (1) the lens-shaped geometry of the deposit open towards the ocean which is a unique geometry within the volcanic succession, (2) sedimentary structures in the lower layer of the deposit that are indicative of a water-laid deposit under oscillatory flow, in contrast to structures in surrounding pyroclastic and volcaniclasitc deposits, and (3) the chaotic upper layer of the deposit which appears to be reworked rather than the product of deposition from pyroclastic activity, lahars or debris flows. A fluvial or debris flow origin for the deposit is discounted
Figure 8 Diagrammatic representation of sediment disturbance and slumping into isolated basins on the Mediterranean Ridge to produce homogenite layers with passage of tsunami generated during the LBA eruption of Thera (from CITA et al., 1984).
Vol. 157, 2000
Tsunami Generated by the Late Bronze Age Eruption of Thera
1245
Figure 9 Pathlines of inferred tsunami propagation. Solid arrows trace tsunami possibly generated by the collapse of the caldera and a large fault block west of the caldera (see Figs. 6e and 6f). Shaded areas in the eastern Mediterranean Sea identify basins on the deep-sea floor where cores have sampled homogenites (11 shaded areas identify 35 different basins). Dashed arrows indicate pathlines for tsunami possibly generated by the entry of pyroclastic flows and lahars/debris flows into the sea (from KASTENS and CITA, 1981).
by these observations: (1) cross-bedding structures indicate oscillatory flow rather than unidirectional flow; (2) abundant pumice is indicative of limited abrasion during flow, inferring short duration and distance of flow; and (3) the deposit does not appear constrained within a channel, rather appears to be planar in bed geometry, indicative of sheet-wash type motion. The stratigraphic position of the tsunami deposit, in lower contact with pyroclastic flow deposits of the third (BO3/Minoan C) phase of the eruption, and mixed along the upper contact with debris-flow deposits of the fourth (BO4/Minoan D) eruption phase, infer tsunami origin related to the entry of pyroclastic flows and debris flows into the sea. Initial tsunami flooding is marked by the erosional basal contact. Internal packets of cross-bedded and planar beds within the lower portion of the tsunami deposit might indicate multiple inundation episodes related to repeated entry of pyroclastic flows into the sea, or seiche during maximum inundation. The upper portion of this deposit may represent tsunami generated by
1246
Floyd W. McCoy and Grant Heiken
Pure appl. geophys.,
the entry of a debris flow into the sea (lack of internal sedimentary structures; chaotic assemblage of pumice and lithic fragments; higher proportion of angular, pebble and cobble-sized lithic fragments; somewhat gradational upper contact into a debris flow). If the latter, this is one of few indications of tsunami generated by the entry of debris flows into the sea during a volcanic eruption.
4.3 Tsunami Deposits Elsewhere Tsunami deposits related to the LBA eruption have been inferred from numerous coastal exposures and archaeological sites where rounded pumice has been found, often mixed with seashells and sometimes in a cultural destruction layer dated to the LBA (Fig. 11): Anaphi island east of Thera (MARINOS and MELIDONIS, 1971), Crete (at Amnissos, MARINATOS, 1939; PICHLER and SCHIERING, 1977; FRANCAVIGLIA and DI SABATINO, 1990; at Pitsidia, VALLIANOU, 1996), Cyprus (ASTROM, 1978; MESZAROS, 1978), and Israel (near Tel Aviv, PFANNENSTIEL, 1960; other coastal sites, NEEV et al., 1987). However, the use of pumice in coastal exposures as indicators of tsunami deposition must be done with caution. The disparity between travel times for tsunami generated by the eruption and for pumice rafts drifting on surface currents
Figure 10 Sketch (from a photograph) of the tephra stratigraphy and tsunami layer near Pori, Thera. Stratigraphic identification and approximate scale on the cliff face are shown on the right. Note the lack of internal stratification in tephra layer BO3 and the series of deposits from lahars and debris flows in BO4, in contrast to the lenticular shape and internal stratification of the tsunami deposit. The two layers composing the tsunami deposit are shown: a lower layer characterized by numerous planar beds, cross-beds, and gravel-rich beds, most with erosional contacts including the basal contact; an upper layer characterized by the lack of internal stratification in a chaotic, poorly-sorted volcaniclastic mixture.
Vol. 157, 2000
Tsunami Generated by the Late Bronze Age Eruption of Thera
1247
Figure 11 Regional ash distribution of tephra from the LBA eruption (shaded area). Symbols identify sites and sampling locations where ash (deep-sea cores, lake cores, exposures on land, archaeological sites) and pumice (exposures on land and at archaeological sites) have been found. Numbers next to the two triangles in Anatolia represent layer thicknesses in cm for ash deposits found in lake sediments; curved lines over the eastern Mediterranean Sea are isopachs of ash thickness, in cm, of the tephra layer found in the deep-sea stratigraphic sequence sampled by coring. Data are from: NINKOVITCH and HEEZEN (1965), NEEV et al. (1987), CADOGAN (1972), RAPP et al. (1973), VITALIANO and VITALIANO (1974), ZEIST et al. (1975), FORNASERI et al. (1975), ASTROM (1978), WATKINS et al. (1978), HOOD (1978), DOUMAS and PAPAZOGLOU (1980), MCCOY (1980, 1981), SULLIVAN (1988, 1990), STANLEY and SHENG, 1986), KELLER (1980), BETANCOURT et al. (1990), MARKETOU (1990), SOLES and DAVARAS (1990), WARREN and PUCHELT (1990), FRANCAVIGLIA (1990), and GUICHARD et al. (1993).
(5–20 cm/sec., O8 ZSO¨Y et al., 1989) constrains distances from Thera where tsunami would have arrived coincident with rafted pumice. As one example, open-ocean velocities for tsunami of the order of 1000 km/hr would have waves arriving in the Levant (approximately 1000 km, straight line distance from Thera) in about 60–100 mm (Fig. 12; YOKOYAMA, 1978). Pumice rafted by surface currents of 5–20 cm/sec for 1500 km (typical surface currents in the eastern Mediterranean Sea; approximate float path following eastern Mediterranean gyres; O8 ZSO¨Y et al., 1989) would
1248
Floyd W. McCoy and Grant Heiken
Pure appl. geophys.,
arrive along the Levant coastline in about 85–350 days, long after the eruption and eruption-generated tsunami (assuming a four-day duration of the LBA eruption; SPARKS and WILSON, 1990; SIGURDSSON et al., 1990). Accordingly, pumice found along the Levant littoral could not have been deposited by tsunami generated by the LBA eruption. In a general case (applying assumptions noted above on surface current patterns and velocity, drift paths and rates for pumice rafts, 4-day duration of the eruption, tsunami generated during late-phase eruptive activity), rafted pumice would be at distances of only about 70 – 100 km or less from Thera (north-central coast of Crete, Anaphi, Melos, Amorgos) after four days of drifting and in position for tsunami arrival and deposition on land.
5. Wa6e Characteristics Inferred for Tsunami Generated by the LBA Eruption of Thera Wave heights near the eruption site on Santorini (no water depth specified or distance offshore) have been estimated at 50–86 m (YOKOYAMA, 1978). The lower wave-height estimate was based upon two pumice deposits about 1 m thick found by MARINOS and MELIDONIS (1971) on the island of Anaphi due east of Thera (Fig. 1) at an elevation of 40 – 50 m. It is not clear if the Anaphi deposit represents
Figure 12 Tsunami time-distance relationships and pathlines from Thera; contours are in minutes (from YOKOYAMA, 1978).
Vol. 157, 2000
Tsunami Generated by the Late Bronze Age Eruption of Thera
1249
a pumice-fall accumulation or a deposit left from runup by tsunami. The upper wave-height estimate was derived by YOKOYAMA (1978) from the occurrence of pumice 5 m above sea level in a coastal terrace near Tel Aviv (presumed related to the Thera eruption by PFANNENSTIEL, 1960; see section above), assuming LBA sea-level about 2 m below present levels and a wave height of 7 m offshore of Tel Aviv, then calculating an original wave height of 86 m near Santorini. By averaging the 50 – 86 m wave-height estimate, then ‘‘weighting twice’’ the Anaphi estimate because the former ‘‘might be more reliable,’’ YOKOYAMA (1988) refined the wave-height calculation to 63 m near Santorini. Tsunami directed towards Anaphi and the Levant, including south towards Crete, likely were generated by the entry of pyroclastic flows and lahars/debris flows in the sea (as noted previously). Tsunami deposits found near Pori on Thera suggest these estimates by YOKOYAMA (1978) may be too high. This deposit has an upper contact 7 – 8 m above modern sea level, providing a minimum indication of tsunami wave heights. Adding to this an LBA (3600 yBP) sea level perhaps as much as 2 meters lower than modern levels in the Mediterranean and southern Aegean region (VAN ANDEL and SHACKLETON, 1982; FLEMMING and WEBB, 1986; GALILI et al., 1988; STANLEY and WARNE, 1993), and evidence of perhaps up to 2 m of subsidence of Thera since the LBA (GALANOPOULOS and BACON, 1969), then tsunami wave heights at coastal impact might have been as much as 11–12 m. Given uncertainties between runup and wave height, wave heights somewhere between 7 m and 12 m might be suggested. Such an estimate is close to those by PICHLER and SCHIERING (1977) of 8–10 m wave heights along the northern and eastern coasts of Crete at various archaeological sites. KASTENS and CITA (1981) used YOKOYAMA’s (1978) minimum estimate of a 50 m wave near Thera, and assumed this wave amplitude at a 200 m water depth off Santorini, to calculate open-water wave heights of between 1.9 and 17 m, with wave lengths \ 100 km, in the eastern Mediterranean Sea over the abyssal homogenite sites. Sea-floor current velocities of up to 49 cm/sec would result from such a wave, and would exceed critical erosional velocities needed to erode and cause slumping of fine-grained pelagic sediments off basin walls onto basin floors to form homogenite deposits (KASTENS and CITA, 1981). Nearshore wave heights as openocean tsunami amplitudes of this magnitude entering shallow coastal waters might be considerably higher than those suggested from the tsunami deposit at Pori, Thera, a reflection of generation of the homogenite-forming tsunami by caldera collapse rather than by pyroclastic and lahars/debris flows entering the sea.
6. Summary Tsunamis generated by the LBA eruption of Santorini were generated by (1) caldera collapse and (2) the entry of pyroclastic flows and lahars/debris flows into
1250
Floyd W. McCoy and Grant Heiken
Pure appl. geophys.,
the sea. Volcanogenic tsunamis produced by caldera collapse and the entry of pyroclastic flows into the sea are well documented from historic eruptions; tsunami generation by the entry of lahars and debris flows into the sea is poorly documented. Evidence for tsunami comes from deep-sea deposits left by open-ocean passage of the wave, as well as from deposits on land where inundation has left pumice accumulations, has produced destruction layers at archaeological sites, or has left distinctive volcanoclastic deposits within the tephra stratigraphy of the eruption on eastern Thera. Pumice accumulations in coastal exposures or archaeological sites at distances greater than about 100 km likely were not, however, emplaced by tsunami generated during the LBA Thera eruption considering drift rates via surface currents in the Aegean and eastern Mediterranean Seas and travel times for tsunami. Wave propagation following caldera collapse (third and fourth phases of eruption activity) would have been to the west, considering somewhat asymmetric collapse of the volcanic edifice and the occurrence of tsunami deposits (homogenites) on the eastern Mediterranean sea-floor; open-ocean wave amplitudes are calculated to have been 1.9 – 17 m. Wave propagation following ocean entry of pyroclastic flows and lahars/debris flows (second through fourth phases of eruption activity) would have been in all directions around Thera, and certainly towards the east as indicated by tsunami deposits on eastern Thera; wave amplitudes are estimated at about 7 – 12 m along coastal Thera and Crete. The regional effects of tsunami must have been significant, considering the cataclysmic event which occurred within the core of two thalassocratic cultures, Minoan and Cycladic, and which destroyed the center of one, the Cycladic culture on Thera.
Acknowledgements Our thanks to numerous colleagues for discussions, assistance and comments: C. Doumas and the archaeologists at the Akrotiri archaeological excavation, Meg Watters, Eleni Argy Murray, and other field helpers and students (especially Tom Lowther, Mary Lewis, and Ian Robins). Reviews by B. Keating, W. Dudley, W. Friedrich and anonymous reviewers improved the manuscript considerably. We thank those who made life in the field a pleasant experience (N. Yannaros, A. Kambouris, F. Colombo, K. Damigos, M. Karnavoulis and others). Field work was permitted by the Institute of Geology and Mineral Exploration (IGME), and supported by the National Geographic Society, Earthwatch/Center for Field Studies and the Associated Scientists at Woods Hole. Production of graphics was supported in part by the Colgate Fund of the Associated Scientists at Woods Hole. Portions of this study were done while F. McCoy was Senior Research Associate at the Wiener Laboratory of the American School of Classical Studies in Athens. This is a contribution from the School of Oceanography and Earth Sciences and
Vol. 157, 2000
Tsunami Generated by the Late Bronze Age Eruption of Thera
1251
Technology, University of Hawaii (no. 4955) and the Associated Scientists at Woods Hole (Mass.).
REFERENCES AGARWAL, N. K., JACOBY, W. R., and BERCKHEMER, H. (1976), Teleseismic P-wa6e Tra6el-time Residuals and Deep Structure of the Aegean Region, Tectonophysics 31, 33 – 57. ARMIJO, R., LYONCAEN, H., and PAPANASTASSIOU, D. (1992), East-west Extension and Holocene Normal-fault Scarps in the Hellenic Arc, Geology 20, 491 – 494. ASTROM, P., Traces of the eruption of Thera on Cyprus? In Thera and the Aegean World II (Doumas, C., ed.) 1 (London, Thera and the Aegean World 1978) pp. 31 – 34. BARBER, R. L. N., The Cyclades in the Bronze Age (London, Gerald Duckworth & Co., Ltd. 1987) 283 pp. BARTON, M., and HUIJSMANS, J. P. P. (1986), Post-caldera Dacites from the Santorini Volcanic Complex, Aegean Sea, Greece: An Example of the Eruption of La6as of Near-constant Composition o6er a 2,200 -year Period, Contr. Min. Pet. 94, 472 – 495. BATH, M. (1983), Earthquake Frequency and Energy in Greece, Tectonophysics 95, 233 – 252. BETANCOURT, P., GOLDBERG, P., HOPE SIMPSON, R., and VITALIANO, C. J., Exca6ations at Pseira: The e6idence for the Theran eruption. In Thera and the Aegean World III, 3 Chronology (Hardy, D. A., ed.) (London, The Thera Foundation 1990) pp. 96 – 99. BOND, A., and SPARKS, R. S. J. (1976), The Minoan Eruption of Santorini, Greece, J. Geolog. Soc. London 132, 1–16. BONDEVIK, S., SVENDSEN, J. I., and MANGERUD, J. (1997), Tsunami Sedimentary Facies Deposited by the Storegga Tsunami in Shallow Marine Basins and Coastal Lakes, Western Norway, Sedimentology 44, 1115–1131. BONNEAU, M., Correlation of the Hellenide nappes in the southeast Aegean and their tectonic reconstruction. In The Geological E6olution of the Eastern Mediterranean (Dixon, J. E., and Robertson, A. H. F., eds.) Special Publication of the Geological Society of London, no. 17 (Oxford, Blackwell Scientific 1984) pp. 517–527. BUGGE, T., BEFRING, S., BELDERSON, R. H., EIDVIN, T., JANSEN, E., KENYON, N. H., HOLTEDAHL, H., and SEJRUP, H. P. (1987), A Giant Three-stage Submarine Slide off Norway, Geo-Marine Lett. 7, 191–198. CADOGAN, G. (1972), Volcanic Glass Shards in Late Minoan I Crete, Antiquity 46, 310 – 313. CAREY, S., SIGURDSSON, H., and MANDEVILLE, C. (1996), Pyroclastic Deposits from Flows and Surges which Tra6elled o6er the Sea during the 1883 Eruption of Krakatau Volcano, Bull. Volcanol. 57, 493–511. CAREY, S., SIGURDSSON, H., and MANDEVILLE, C. (2000), Pyroclastic deposits from flows and surges that tra6elled o6er the sea from the 1883 Krakatau eruption. In Volcanic Hazards and Disasters in Human Antiquity (McCoy, F. W., and Heiken, G., eds.) Geological Society of America Special Paper No. 345. CITA, M.B., CAMERLENGHI, A., KASTENS, K. A., and MCCOY, F. W. (1984), New Findings of Bronze Age Homogenites in the Ionian Sea: Geodynamic Implications for the Mediterranean, Marine Geology 55, 47–62. COX, D. C. (1979), Local Tsunamis in Hawaii — Implications for Hazard Zoning, Hawaii Institute of Geophysics Report HIG-79-5, University of Hawaii, Honolulu, 46 pp. DAWSON, A. G., LONG, D., and SMITH, D. E. (1988), The Storegga Slides: E6idence from Eastern Scotland for a Possible Tsunami, Marine Geology 82, 271 – 276. DAWSON, A. G., LONG, D., SMITH, D. E., SHI, S., and FOSTER, I. D. L., Tsunamis in the Norwegian Sea and North Sea caused by the Storegga submarine landslide. In Tsunamis in the World (Tinti, S., ed.) (Netherlands, Kluwer Academic Publishers 1993) pp. 31 – 42.
1252
Floyd W. McCoy and Grant Heiken
Pure appl. geophys.,
DECKER, R. W., How often does a Minoan eruption occur? In Thera and the Aegean World III (Hardy, D. A., ed.) 2 (Earth Sciences: London, The Thera Foundation 1990) pp. 444 – 452. DEWEY, J. F., PITMAN, W. C., RYAN, W. B. F., and BONNIN, J. (1973), Plate Tectonics and the E6olution of the Alpine System, Geolog. Soc. of Am. Bull. 84, 3137 – 3180. DOUMAS, C. G., Thera: Pompeii of the Ancient Aegean (London, Thames and Hudson 1983) 168 pp. DOUMAS, C. G., and PAPAZOGLOU, L. (1980), Santorini Tephra from Rhodes, Nature 287, 322 – 324. DRUITT, T. H., and FRANCAVIGLIA, V. (1992), Caldera Formation on Santorini and the Physiogeography of the Islands in the Late Bronze Age, Bull. Volcanol. 54, 484 – 493. DRUITT, T. H., MELLORS, R. A., PYLE, D. M., and SPARKS, R. S. J. (1989), Explosi6e Volcanism on Santorini, Greece, Geological Magazine 126, 95 – 126. FLEMMING, N. C., and WEBB, C.O. (1986), Tectonic and eustatic coastal changes during the last 10,000 years deri6ed from archaeological data. In Dating Mediterranean Shorelines (Ozer, A. and Vita-Finzi, C., eds.) Zeitschrift fu¨r Geomorphologie, Supplement 62, 1 – 29. FORNASERI, M., MALPIERI, L., and TOLOMEO, L. (1975), Pro6enance of Pumices in the North Coast of Cyprus, Archaeometry 17, 112–116. FRANCAVIGLIA, V., Sea-borne pumice deposits of archaeological interest on Aegean and eastern Mediterranean beaches. In Thera and the Aegean World III, 3 Chronology, London (Hardy, D. A., ed.) (The Thera Foundation 1990) pp. 127–134. FRANCAVIGLIA, V., and DI SABATINO, B. (1990), Statistical study on Santorini pumice-falls. In Thera and the Aegean World III, 3 Chronology, London (Hardy, D. A., ed.) (The Thera Foundation 1990) pp. 29–52. FRIEDRICH, W. L., Feuer im Meer6ulkanismus und die Naturgeschichte der Insel Santorin (Heidelberg, Spektrum Akademischer Verlag 1994) 256 pp. FRIEDRICH, W. L., ERIKSEN, U., TAUBER, H., HEINEMEIER, J., RUD, N., THOMSEN, M. S., and BUCHARDT, B. (1988), Existence of a Water-filled Caldera prior to the Minoan Eruption of Santorini, Greece, Naturwissenschaften 75, 567–569. FYTIKAS, M., KOLIOS, N., and VOUGIOUKALAKIS, G., Post-Minoan 6olcanic acti6ity of the Santorini 6olcano. Volcanic hazard and risk, forecasting possibilities. In Thera and the Aegean World II, 2 Earth Sciences, London (Hardy, D. A., ed.) (The Thera Foundation 1990) pp. 183 – 198. FYTIKAS, M., INNOCENTI, F., MANETTI, P., MAZZUOLI, R., PECCERILO, A., and VILLARI, L., Tertiary to Quaternary e6olution of the 6olcanism in the Aegean region. In The Geological E6olution of the Eastern Mediterranean (Dixon, J. E. and Robertson, A. H. F., eds.) Special Publication of the Geological Society of London, no. 17 (Oxford, Blackwell Scientific 1984) pp. 687 – 699. GALANOPOULOS, A. G. (1960), Tsunamis Obser6ed on the Coasts of Greece from Antiquity to Present Time, Annali di Geofisica 13, 369–386. GALANOPOULOS, A. G., and BACON, E., Atlantis —the Truth behind the Legend (Indianapolis, BobbsMerrill Co. 1969) 216 pp. GALILI, E., WEINSTEIN-EVRON, M., and RONEN, A. (1988), Holocene Sea-le6el Changes Based on Submerged Archaeological Sites off the Northern Carmel Coast of Israel, Quaternary Research 29, 36–42. GUICHARD, F., CAREY, S., ARTHUR, M. A., SIGURDSSON, H., and ARNOLD, M. (1993), Tephra from the Minoan Eruption of Santorini in Sediments of the Black Sea, Nature 363, 610 – 612. HALL, R., AUDLEY-CHARLES, M. G., and CARTER, D. J., The significance of Crete for the e6olution of the eastern Mediterranean. In The Geological E6olution of the Eastern Mediterranean (Dixon, J. E., and Robertson, A. H. F., eds.) Special Publication of the Geological Society of London, no. 17 (Oxford, Blackwell Scientific 1984) pp. 499 – 516. HARDY, D. A., editor, Thera and the Aegean World III, Archaeology 1, Volume, Earth Sciences 2 and Chronology 3 (London, The Thera Foundation 1990) 511 pp., 487 pp. and 242 pp. HEIKEN, G. H., and MCCOY, F. W. (1984), Caldera De6elopment during the Minoan Eruption, Thera, Cyclades, Greece, J. Geophys. Res. 89, 8441 – 8462. HEIKEN, G. H., and MCCOY, F. W. (1990), Precursory acti6ity to the Minoan eruption, Thera, Greece. In Thera and the Aegean World III, 2 Earth Sciences (Hardy, D. A., ed.) (London, The Thera Foundation 1990) pp. 79–88.
Vol. 157, 2000
Tsunami Generated by the Late Bronze Age Eruption of Thera
1253
HEIKEN, G. H., MCCOY, F. W., and SHERIDAN, M. (1990), Palaeotopographic and palaeogeologic reconstruction of Minoan Thera. In Thera and the Aegean World III, 2 Earth Sciences (Hardy, D. A., ed.) (London, The Thera Foundation 1990) pp. 370 – 378. HIEKE, W. (1984), A Thick Holocene Homogenite from the Ionian Abyssal Plain (Eastern Mediterranean), Marine Geology 55, 63–78. HOOD, S., The Minoans (London, Thames and Hudson 1971), 453 pp. HOOD, S., Traces of the eruption outside Thera. In Thera and the Aegean World II, 1 (Doumas, C., ed.) (London, Thera and the Aegean World 1978) pp. 681 – 690. JANSEN, E., BEFRING, S., BUGGE, T., EIDVIN, T., HOLTDAHL, H., and SEJRUP, H. P. (1987), Large Submarine Slides on the Norwegian Continental Margin: Sediments, Transport and Timing, Marine Geology 78, 77–107. KAHLE, H.-G., and MULLER, M. V. (1996), Trajectories of Crustal Deformation of Western Greece from GPS Obser6ations 1989 – 1994, Geophys. Res. Lett. 23, 677 – 680. KASTENS, K. A., and CITA, M. B. (1981), Tsunami-induced Sediment Transport in the Abyssal Mediterranean Sea, Geolog. Soc. Am. Bull. 92, 845 – 857. KASTENS, K. A., BREEN, N. A., and CITA, M. (1992), Progressi6e Deformation of an E6aporite-bearing Accretionary Complex: SeaMARC I, SeaBeam and Piston-core Obser6ations from the Mediterranean Ridge, Marine Geophys. Res. 14, 249– 298. KELLER, J., Prehistoric pumice tephra on Aegean islands. In Thera and the Aegean World II, 2 (Doumas, C., ed.) (London, Thera and the Aegean World 1980) pp. 49 – 56. LATTER, J. H. (1981), Tsunamis of Volcanic Origin: Summary of Causes, with Particular Reference to Krakatau, 1883, Bull. Volcanol. 44, 467 – 490. LE PICHON, X., and ANGELIER, J. (1979), The Hellenic Arc and Trench System: A Key to the Neotectonic E6olution of the Eastern Mediterranean Area, Tectonophysics 60, 1 – 42. LIVERMORE, R. A., and SMITH, A. G. (1985), Some boundary conditions for the e6olution of the Mediterranean region. In Geological E6olution of the Mediterranean Basin (Stanley, D. J., and Wezel, F.-C., eds.) (New York, Springer-Verlag 1985) pp. 83 – 98. LOOMIS, H. G., Tsunami Wa6e Runup in Heights in Hawaii, Hawaii Institute of Geophysics Report HIG-76-5, University of Hawaii, Honolulu, 13 pp. LUCE, J. V., The End of Atlantis (London, Thames and Hudson 1969) 187 pp. MAKRIS, J. (1978), The Crust and Mantle of the Aegean Region from Deep Seismic Soundings, Tectonophysics 46, 269–284. MACDONALD, G. A., SHEPARD, F. P., and COX, D. C. (1947), The Tsunami of April 1, 1946, in the Hawaiian Islands, Pacific Science 1, 21 – 37. MANNING, S. W., The Absolute Chronology of the Aegean Early Bronze Age: Archaeology, Radiocarbon, and History (Sheffield, U.K., Sheffield Academic 1995) 345 pp. MARINATOS, S. (1939), The Volcanic Destruction of Minoan Crete, Antiquity 13, 339 – 425. MARINOS, G., and MELIDONIS, N. (1971), On the Strength of Seaquakes (Tsunami ) during the Prehistoric Eruption of Santorini, Acta of the First International Scientific Congress on the Volcano Thera, Athens, The Archaeological Society, 277 – 282. MARKETOU, T., Santorini tephra from Rhodes and Kos: Some chronological remarks based on the stratigraphy. In Thera and the Aegean World III, 3 Chronology (Hardy, D. A., ed.) (London, The Thera Foundation 1990) pp. 100–113. MCCOY, F. W. (1980), The upper Thera (Minoan) ash in deep-sea sediments: distribution and comparison with other ash layers. In Thera and the Aegean World II, 2 (Doumas, C., ed.) (London, Thera and the Aegean World 1980) pp. 57–78. MCCOY, F. W., Areal distribution, redeposition and mixing of tephra within deep-sea sediments of the eastern Mediterranean Sea. In Tephra Studies (Self, S., and Sparks, R. S. J., eds.) (Hingham, Mass., D. Reidel 1981) pp. 245–254. MCCOY, F. W. (1997), Tsunami Deposits on Santorini (Thera) Island, Greece, from the Minoan Late Bronze Age Eruption of Thera Volcano, abst., Geological Society of America abstracts with Programs, 29, no. 5, 28. MCCOY, F. W. (1997b), Tsunami Deposits on Santorini (Thera) Island, Greece, from the Late Bronze Age Eruption of Thera Volcano, abstr., Geol. Soc. Am. Abstr. with Programs 29 no. 5, 28.
1254
Floyd W. McCoy and Grant Heiken
Pure appl. geophys.,
MCCOY, F. W., and HEIKEN, G. (1997), Tsunami Generated by the Late Bronze Age Volcanic Eruption of Santorini, Greece, abstr., EOS Trans., Am. Geophys. Union 78, 452. MCCOY, F. W., and HEIKEN, G. (2000), The Late-Bronze Age explosi6e eruption of Thera (Santorini ), Greece: Regional and local effects. In Volcanic Hazards and Disasters in Human Antiquity (McCoy, F. W., and Heiken, G., eds.) Geological Society of America Special Paper No. 345. MCGUIRE, W. J. (1996), Volcano instability: a re6iew of contemporary themes. In Volcano Instability on the Earth and Other Planets (McGuire, W. J., Jones, A. P., and Neuberg, J., eds.) Geolog. Soc. Special Publication 110, 1–23. MCGUIRE, W. J., JONES, A. P., and NEUBERG, J. (1996) eds., Volcano Instability on the Earth and Other Planets, Geolog. Soc. Special Publication 110, 1 – 23. MCKENZIE, D. (1972), Acti6e Tectonics of the Eastern Mediterranean Region, Geophys. J. Roy. Astron. Soc. 30, 109–185. MEIJER, P. TH., and WORTEL, M. J. R. (1997), Present-day Dynamics of the Aegean Region: A Model Analysis of the Horizontal Pattern of Stress and Deformation, Tectonics 16, 879 – 895. MESZAROS, S., Some words on the Minoan tsunami of Santorini. In Thera and the Aegean World II, 2 (Doumas, C., ed.) (London, Thera and the Aegean World 1978) pp. 257 – 262. NEEV, D., BAKLER, N., and EMERY, K. O., Mediterranean Coasts of Israel and Sinai (New York, Taylor and Francis 1987) 130 pp. NEWHALL, C. G., and SELF, S. (1982), The Volcanic Explositi6ity Index (VEI): An Estimate of Explosi6e Magnitude for Historical Volcanism, J. Geophys. Res. 87, 1231 – 1238. NINKOVITCH, D., and HEEZEN, B. C. (1965), Santorini tephra, Proceedings of the 17th Symposium of the Colston Research Society, London, Butterworths Scientific Publications, 413 – 453. ORAL, M. B., REILINGER, R. E., TOKSO¨Z, M. N., KING, R. W., BARKA, A. A., KINIK, I., and LENK, O. (1995), Global Positioning System Offers E6idence of Plate Motions in Eastern Mediterranean, EOS, Trans. Am. Geophys. Union 76, 9–11. O8 ZSO¨Y, E., HECHT, A., and U8 NLU¨ATA, U. (1989), Circulation and Hydrography of the Le6antine Basin. Results of POEM Co-ordinated Experiments 1985 – 1986, Progress in Oceanography 22, 125 – 170. PERISSORATIS, C. (1985), Surficial Sediment Map of the Aegean Sea Floor, Institute of Geology and Mineral Exploration (1 : 200,000), Athens, Greece. PFANNENSTIEL, M. (1960), Erla¨uterungen zu den bathymetrischen Karten des o¨stlichen Mittelmeeres, Bull. Instit. Oceanographique 1192, Monaco. PICHLER, H., and FRIEDRICH, W. L., Mechanism of the Minoan eruption of Santorini. In Thera and the Aegean World II, 2 (Doumas, C., ed.) (London, Thera and the Aegean World 1980) pp. 15 – 30. PICHLER, H., and KUSSMAUL, S., Comments on the geological map of Santorini Islands. In Thera and the Aegean World II, 2 (Doumas, C., ed.) (London, Thera and the Aegean World 1980) pp. 413 – 427. PICHLER, H., and SCHIERING, H. (1977), The Thera Eruption and Late Minoan-IB Destructions on Crete, Nature 267, 819–822. PLATON, N., Zakros. The Disco6ery of a Lost Palace of Ancient Greece (New York, Charles Scribners and Sons 1971) 345 pp. PYLE, D. M. (1990), New estimates for the 6olume of the Minoan eruption. In Thera and the Aegean World III, 2 Earth Sciences (Hardy, D. A., ed.) (London, The Thera Foundation 1990) pp. 113 – 121. RAPP, G., COOKE, S. R. B., and HENDRICKSON, E. (1973), Pumice from Thera (Santorini ) Identified from a Greek Mainland Archaeological Exca6ation, Science 179, 471 – 473. RECK, H., Santorini: Der Werdegang eines Insel6ulkans und sein Ausbruch 1925 – 1928 (Berlin, D. Reimer 1936). ROBERTSON, A. H. F., and DIXON, J. E., Introduction: aspects of the geological e6olution of the eastern Mediterranean. In The Geological E6olution of the Eastern Mediterranean (Dixon, I. E., and Robertson, A. H. F., eds.) Special Publication of the Geolog. Soc. London 17 (Oxford, Blackwell Scientific 1984) pp. 1–74. RYAN, W. B. F., KASTENS, K. A., and CITA, M. B. (1982), Geological E6idence Concerning Compressional Tectonics in the Eastern Mediterranean, Tectonophysics 86, 213 – 242. SELF, S., and RAMPINO, M. R. (1981), The 1883 Eruption of Krakatau, Nature 294, 699 – 704. SEWARD, D., WAGNER, G. A., and PICHLER, H., Fission track ages from Santorini 6olcanics (Greece). In Thera and the Aegean World II, 2 (Doumas, C., ed.) (London, Thera and the Aegean World 1980) pp. 101–108.
Vol. 157, 2000
Tsunami Generated by the Late Bronze Age Eruption of Thera
1255
SIGURDSSON, H., CAREY, S., and DEVINE, J. D., Assessment of mass, dynamics and en6ironmental effects of the Minoan eruption of Santorini 6olcano. In Thera and the Aegean World III, 2 Earth Sciences (Hardy, D. A., ed.) (London, The Thera Foundation 1990) pp. 100 – 112. SIGURDSSON, H., CAREY, S., and MANDEVILLE, C. (1991), Krakatau, Nat. Geog. Res. Expl. 7, 310 – 327. SIMKIN, T., and FISKE, R. S., Krakatau 1883: The Volcanic Eruption and its Effects (Washington, D.C., Smithsonian Institution Press 1983) 464 pp. SIMKIN, T., and SIEBERT, L. (1994), Volcanoes of the World (Tucson, Geoscience Press 1994) 349 pp. SOLES, J., and DAVARAS, C., The ran ash in Minoan Crete: new exca6ations on Mochlos. In Thera and the Aegean World III, 3 Chronology (Hardy, D. A., ed.) (London, The Thera Foundation 1990) pp. 89–95. SPAKMAN, W. (1986), Subduction beneath Eurasia in Connection with the Mesozoic Tethys, Geologie Mijnbouw 65, 145–153. SPARKS, R. S. J., and WILSON, C. J. N. (1990), The Minoan deposits: a re6iew of their characteristics and interpretation. In Thera and the Aegean World III, 2 Earth Sciences (Hardy, D. A., ed.) (London, The Thera Foundation 1990) pp. 89–99. SPIELVOGEL, L. Q. (1973), Runup of Single Wa6es on a Sloping Beach, Hawaii Institute of Geophysics Report HIG-73-21, University of Hawaii, Honolulu, 21 pp. STANLEY, D. J., and SHENG, H. (1986), Volcanic Shards from Santorini (Minoan Ash) in the Nile Delta, Egypt, Nature 320, 733–735. STANLEY, D. J., and WARNE, A. G. (1993), Sea Le6el and Initiation of Predynastic Culture in the Nile Delta, Nature 363, 435–438. SULLIVAN, D. G. (1988), The Disco6ery of Santorini Minoan Tephra in Western Turkey, Nature 333, 552–554. SULLIVAN, D. G., Minoan tephra in lake sediments in western Turkey: dating the eruption and assessing the atmospheric dispersal of the ash. In Thera and the Aegean World III, 3 Chronology (Hardy, D. A., ed.) (London, The Thera Foundation 1990) pp. 114 – 119. TAYMAZ, T., JACKSON, J., and WESTWAY, R. (1990), Earthquake Mechanisms in the Hellenic Trench near Crete, Geophys. J. Internat. 102, 695 – 731. VALLIANOU, D. (1996), New e6idence of earthquake destructions in Late Minoan Crete. In Archaeoseismology (Stiros, S., and Jones, R. E., eds.) Athens, Institute of Geology and Mineral Exploration, and The British School at Athens, Fitch Laboratory Occasional Paper 7, 153 – 168. VAN ANDEL, TJ. H., and SHACKLETON, J. C. (1982), Late Palaeolithic and Mesolithic Coastlines of Greece and the Aegean, J. Field Archaeology 9, 445 – 454. VAN PADANG, M. (1971), Two Catastrophic Eruptions in Indonesia, Comparable with the Plinian Outburst of the Volcano of Thera (Santorini ) in Minoan Time, Acta of the 1st Intl. Sci. Cong. on the Volcano of Thera; Arch. Services of Greece, General Direction of Antiquitiest and Research, Athens, 51 – 63. VITALIANO, C. J., and VITALIANO, D. B. (1974), Volcanic Tephra on Crete, Am. J. Archaeology 78, 19–24. VOGEL, J. S., CORNELL, W., NELSON, D. E., and SOUTHON, J. R. (1990), Vesu6ius/A6ellino, One Possible Source of Se6enteenth Century BC Climatic Disturbances, Nature 344, 534 – 537. WALKER, G. P. L. (1973), Explosi6e Volcanic Eruptions — A New Classification Scheme, Geologische Rundschau 62, 431–446. WARREN, P. M., and PUCHELT, H. (1990), Stratified pumice from Bronze Age Knossos. In Thera and the Aegean World III, 3 Chronology (Hardy, D. A., ed.) (London, The Thera Foundation 1990) pp. 71–81. WATKINS, N. D., SPARKS, R. S. J., SIGURDSSON, H., HUANG, T. C., FEDERMAN, A., CAREY, S., and NINKOVITCH, D. (1978), Volume and Extent of the Minoan Tephra from Santorini Volcano: New E6idence from Deep-sea Sediment Cores, Nature 271, 122 – 126. WAYTHOMAS, C. F., and NEAL, C. A. (1997), Tsunami Generation during the 3500 yr BP Caldera-forming Eruption of Aniakchak Volcano, abstr. EQS Trans., Am. Geophys. Union 78, F816. WHARTON, W. J. L. (1888), On the seismic sea wa6es caused by the eruption of Krakatoa August 26th and 27th, 1883. In The Eruption of Krakatoa, and Subsequent Phenomena (Symons, G. J., ed.) Report of the Krakatoa Committee of the Royal Society, London, the Royal Society, 494 pp.
1256
Floyd W. McCoy and Grant Heiken
Pure appl. geophys.,
YOKOYAMA, I. (1978), The tsunami caused by the prehistoric eruption of Thera. In Thera and the Aegean World II, 2 (Doumas, C., ed.) (London, Thera and the Aegean World 1978), 277 – 283. YOKOYAMA, I. (1988), The tsunami caused by the prehistoric eruption of Thera. In (Mourtzos, N. and Marinos, P., eds.) International Symposium on Engineering Geology and Protection of Ancient Works, Monuments and Historic Sites, Athens, Field Guide Tour T3, The Aegean World-Greek Islands. ZEIST, W. VON, WOLDRING, H., and STAPERT, D. (1975), Late Quaternary Vegetation and Climate of Southwestern Turkey, Palaeohistoria 17, 53 – 143. (Received June 4, 1998, revised March 5, 1999, accepted May 18, 1999)