Synonyms

Seismic sea wave; Tidal wave (erroneous)

Definition

Tsunami: Japanese “harbor wave.”

Tsunami: long-length waves in a body of water (usually large lake or sea), generated by the substantial displacement of water.

Introduction

Human populations have long shown a preference for settling in coastal zones. The coastline is a meeting point between different resource zones and a doorway to efficient and economical water-based transport for trade. However, this rich and attractive location can also be vulnerable to forces of nature with gradual or rapid disastrous consequences such as sea-level change, hurricanes, and tsunamis. Tsunamis occur in a relatively short period of time. The warnings, if recognized, are rarely more than a half day in advance and more likely a matter of minutes to hours. The size of a tsunami varies greatly, ranging from waves that only slightly alter a typical day’s tide gauge record to mega-tsunamis with the potential of wave heights in the tens of meters and inundation values over 1 km. In recent years, multiple mega-tsunamis (e.g., December 26, 2004, and July 17, 2006, in the Indian Ocean and March 11, 2011, in Tohoku, Japan; see recent discussion in Goff et al., 2012) have heightened the attention and awareness of this phenomenon and directed greater interest into the impact tsunamis have had on coastlines in the past. Archaeological coastal sites have provided some insight into understanding the impact of tsunamis over time, and geoarchaeological studies of coastal sites are at the forefront of this field.

Tsunami dynamics

Tsunamis are propagated by a sudden displacement of water that results in a wave characterized by a substantially long wavelength and greater potential for inshore run-up relative to meteorologically driven waves. The causes of tsunamis include earthquakes (subaerial and terrestrial), landslides, subaerial slumping, volcanic eruptions, meteorites, glacial felling, and explosions. While, at sea, a tsunami’s wave height might be considered negligible, its wavelength can be as long as hundreds of kilometers. In contrast, large storm waves have a wavelength of 30–40 m. As the tsunami wave approaches a coastline, the shallower water depth slows the overall speed of the wave and compresses its full volume, resulting in increasing wave heights. This volume of water can then flow far inland until the wave’s energy dissipates. Damages to coastal settlements resulting from tsunami impact range from insignificant to completely devastating, depending on the specific dynamics of the wave, the preparedness of the human population (particularly whether evacuation has occurred), the extent of coastal development, and the specific character of the coastline. The most vulnerable sites are those located along the coastlines, especially within harbors or estuaries where the tsunami flow is known to travel even further inland, become magnified due to reflective waves, and sometimes gain wave height.

Historical evidence of tsunami events

Written descriptions of events are important resources for reconstructing tsunami events of the past. Sources range from religious tracts, governmental records, requests for assistance, insurance maps, and historical accounts. These descriptions have been compiled into more comprehensive catalogues, traditionally presented in tandem with earthquake records, but sometimes presented independently (e.g., Lockridge and Dunbar, 1996). In areas more extensively studied, separate catalogues exist for discrete regions (e.g., Australia, Italy, Cyprus, Greece, the Eastern Mediterranean, Turkey, the Pacific Ocean, Central America, the Middle East; Soloviev et al., 1986; Molina, 1997; Fernandez et al., 2000; Riad et al., 2003; Tinti et al., 2004; Dominey-Howes, 2007; Fokaefs and Papadopoulos, 2007; Papadopoulos et al., 2007; Ambraseys and Synolakis, 2010; Altinok et al., 2011; Papadopoulos et al., 2014). These catalogues present a summary of information and known sources that provide information about past tsunamis. Some catalogues contain the source information translated into a single language, together with the original source reference, but unfortunately, the original source is sometimes difficult to access, and only translations are available. This, and other variables, can lead to complications when assessing the reliability of the information.

Common methods for determining the reliability of a written text include determining whether (1) the source is first, second, or third hand; (2) it is contemporaneous with the event or a later recollection; (3) a number of sources corroborate the same event; and (4) related accounts of simultaneous trigger events exist (e.g., earthquakes or volcanic eruptions; see Salamon et al., 2007) and (5) more recently the publication or report of field evidence. Reliability is a problem, as written records sometimes provide erroneous information for a multitude of reasons, including issues with varying calendar systems, tendencies toward exaggeration, poor information transfer, bad translation, the merging of near-contemporaneous events, or the separation of single events into multiple occurrences (Karcz, 2004). Despite these issues, the historical record remains a critical source for reconstructing historical tsunami events.

Instrumentally recorded events are available for the twentieth century to the present. These records come from tidal and wave records, aerial photographs, and more recently satellite imagery, deep-sea wave recorders, and tsunami and earthquake warning systems. Over time, more instruments are being added around the world, and in the future there should be fewer events that occur without being recorded and measured instrumentally.

Paleotsunami research

Reconstructing the occurrence of tsunamis in the past offers a means to estimate the risk for future tsunamis. Tsunamis that date prior to the historical record, or for which there are no written observations, are called “paleotsunamis” (International Tsunami Information Center, 2011). The majority of field research directed at identifying paleotsunamites (sedimentological markers of paleotsunami events) has focused on evidence for marine incursion into terrestrial areas (see discussion in Scheffers and Kelletat, 2003; Dawson and Stewart, 2007). These deposits are especially reliable, as they demonstrate specific movement of marine water above and beyond the usual storm wave zone of influence, an important criterion for properly labeling the event a tsunami rather than storm related (Goff et al., 2004; Morton et al., 2007; Ramírez-Herrera et al., 2012). Stratigraphic sequences in the terrestrial realm, such as those identified during coring or trenching campaigns, might reveal signs of previous tsunami events. Among the proxies that have been used for recognizing these changes include (1) the presence of marine-specific fauna in a setting that is otherwise terrestrial, brackish, or freshwater, (2) changes in sediment grain size, (3) marine-associated organisms within the sands or sediments, (4) erosional bottom contacts, and (5) unusual inclusions of foreign origin (Goff et al., 2012). While, in some cases, affected sites are abandoned after a tsunami strike, some sites are also regularly rebuilt and renewed following the destructive events, and this makes recognizing tsunami deposits or damage within an archaeological site complicated. Therefore, nearby sediment traps in less-inhabited areas (lake bottoms, lagoons, and artificial reservoirs) sometimes preserve more evidence of the event, which can be extrapolated to destruction sequences within archaeological sites based on age.

Evidence of tsunamis at archaeological sites

Coastal archaeological sites are particularly vulnerable to tsunami-related damages. Written records of a tsunami event might be produced when there is an occupied site that is impacted by the event. If a coastline lacks any settlements (or settlements of consequence to those recording the history), it is less likely that the event will be documented in the written record. Simultaneously, the preservation of tsunami deposits is problematic, particularly at occupied sites where rebuilding and post-tsunami cleanup further alters or erases the deposits that may be ephemeral to start.

Evidence for paleotsunami events using field techniques with geoarchaeological components has been published at a few archaeological sites worldwide including Egypt (Guidoboni et al., 1994; Stanley and Bernasconi, 2006; Hamouda, 2009), Lebanon (Morhange et al., 2006), Israel (Reinhardt et al., 2006; Goodman-Tchernov et al., 2009), Crete (Scheffers and Scheffers, 2007; Bruins et al., 2008), the UK (Dawson et al., 1988, 1990), Greece (Vött et al., 2010), and New Zealand (Goff and McFadgen, 2003; Goff et al., 2010). In some cases, studies have been conducted by reviewing and reinterpreting the causes of destruction or abandonment phases published in earlier excavation reports, such as the case of Atlit Yam (Pareschi et al., 2007; as refuted by Galili et al., 2008) and Knossos, Crete (Antonopoulos, 1992; as refuted by Minoura et al., 2000), and New Zealand prehistoric Maori sites (the reassessment by Goff and McFadgen, 2003, of the conclusions of Leach and Leach, 1979). In these cases, a reconsideration of the original findings – and in the Maori case including consideration of nearby physical findings in the non-anthropogenically altered environment – led the researchers to assign a tsunami-related cause to the damage seen at the sites. Tectonic activity and/or tsunami inundation was suggested as the major factor behind the abandonment and movement of North American west coast sites in British Columbia, Canada (Hutchinson and McMillan, 1997), as well as coastal sites in Washington State and Oregon (Woodward et al., 1990; Cole et al., 1996).

A multitude of indicators have been used to suggest tsunami-related damage at archaeological sites. With regard to physical evidence, the most important aspect appears to be the context of the deposit rather than any single feature of the deposits themselves. It is important to note that tsunami deposits can reflect only the materials and sediments supplied at that particular site, so many features may be site specific. For example, if the inhabitants of a site were regularly discarding marine shell into a freshwater lake, then marine shell in the freshwater lake itself cannot be indicative of a past tsunami. However, a sudden change in quantity, species, or individual character of those shells (breakage/wear) would have significance as would the sudden appearance of a shell horizon at a site where such activities were absent. Among the features that have been used to represent tsunamis at archaeological sites (see Goff et al., 2012, for a full summary) are (1) tilted, damaged, or altered marine installations such as harbor features or coastal constructions, (2) unusual changes in grain size distribution within sediments, (3) hiatuses or eroded horizons, (4) burials that appear disturbed or do not fit recognized cultural norms, (5) marine micro- and/or macro-paleontological deposits (e.g., pollen, thecamoebians, foraminifera, ostracods, mollusks, gastropods) distributed out of context or with unusual taphonomic features, (6) abandonment sequences contemporaneous with such features, and (7) unbroken or freshly broken and deposited ceramics or other cultural material offshore without explanatory context (shipwreck, construction fill, garbage pits). All investigations benefit from a wide geographical study of stratigraphic sequences to be able to demonstrate the presence of these unusual horizons, particularly given the common changes in their appearance depending on the dynamics of the particular geography and the variations in local source materials. In Caesarea, Israel, for example, over three tsunamis were identified in offshore cores and in excavations, and each one varied somewhat with regard to tsunami-related indicators. Many of these variations could be attributed to differences in the landscape at the time of the event, especially for events that occurred prior to heavy human occupation and therefore lacked many indicators of anthropogenic origin (ceramics, damaged harbor features, etc.; Goodman et al., 2009).

Summary

The study of tsunamis at archaeological sites is a young field. In time, as the study of field deposits from modern tsunami events advances and knowledge of their expression and dynamics grows, we can expect more descriptions of tsunami-derived deposits at archaeological sites as well. The possibility of tsunami events should always be considered when studying archaeological sites located on coastlines, particularly those in seismically active zones.

Cross-references