Storm surges and tsunamis both produce short-lived, severe inundations of coasts by marine waters, witness the Indian Ocean tsunami of 26 December 2004. Storm surges and tsunamis leave lithic or bioclastic sediment (or both) deposits above the highest astronomical tide (Plates 5.1-5.3) (Chappell et al. 1983; Atwater 1987; Bryant et al. 1992; Dawson 1994). Smaller tsunamis cause run-up and backwash, and leave traces in coastal regions (see Dawson 1994). For instance, tsunamis generated by the Lisbon earthquake of 1755 left traces in estuarine deposits in the Scilly Isles (Foster et al. 1991), and those generated by submarine slides off Iceland left vestiges in the Scottish and Norwegian landscapes (Dawson et al. 1988; Dawson et al. 1993). Along coasts where both hazards occur, such as Western Australia, tsunami sedimentation is distinguishable where deposition occurred well beyond the reach of storm waves and surge, or where the sediment contains clasts too large for carriage by storm waves and surge (Nott and Bryant 2003). Nonetheless, tsunami sediments in coastal environments are extremely difficult to detect (Scheffers and Kelletat 2003). Wave-induced sediments located some distance inland or accumulations of very coarse material can be indicative of tsunamis. Tsunamigenic fine sediments are far more difficult to study and interpret, particularly as wind or ordinary storm waves might have deposited them (see Dawson and Shi 2000). A promising approach to recognizing the vestiges of old tsunamis is to consider the tsunamigenic origin of unusual deposits or geomorphic features in coastal areas. Figure 5.3 shows sites and regions with trustworthy evidence of tsunami deposits or tsunamigenic geomorphic features. Around the shores of the Atlantic, sedimentary evidence of tsunamis comes from the Caribbean, Scotland, western Norway, and the southern Portuguese coast. In the Mediterranean Sea, tsunami deposits come from southern Italy, the Aegean Sea, and Cyprus. In the Indian Ocean, sedimentary evidence is restricted to northwestern Australia. The Pacific Ocean, with many active plate boundaries, has a high frequency of tsunamis, with the best evidence of their passage found in Indonesia, New Guinea, northern and southeastern Australia, New Zealand, Tuamotu (in the South Seas), the Hawaiian Islands, the northwest coast of North America, Kamchatka, and Japan and the Kuriles Islands.
Although researchers have made progress in identifying deposits attributable to tsunamis, they have fared less well in seeking geomorphic imprints of tsunamis around the world's coastlines. Field evidence of geomorphic alterations resulting from tsunamis is confined to four regions and, up to 2003, described in about 15 articles (see Scheffers and Kelletat 2003). This could be because tsunamis have had little effect upon Pleistocene and Holocene coastal development. Ted Bryant (2001) commented that the effects of tsunamis on coastal processes are normally limited, but occasionally they may play a major role. For instance, very high magnitude Holocene tsunamis may have dominated coastal development in southeastern Australia, where they were primarily responsible for the formation of barrier islands, cliffs, canyons, and sculptured bedrock forms (Bryant et al. 1996). Extreme flow velocities are required to carve such bedrock forms as flutes and vortexes, a fact pointing to tsunamis rather than large storm waves as the agents involved. The largest old tsunami waves in Australia swept sediment across the continental shelf, obtaining flow depths of 15-20 m at the coastline with velocities in excess of 10 m/s. At Jervis Bay, New South Wales, waves attained elevations exceeding 80 m and may have has flow depths above 10 m (Bryant and Nott 2001).
Plate 5.2 Boulder ridge (up to 50 t) accumulated 100 m distant from the shoreline at 7 m above sea-level on the southern coast of Anguilla. Photograph: © Anja Scheffers.
Plate 5.3 Tsunami ridge at Eleutherea, Bahamas, consisting of sand and boulders at 15 m above sea-level, deposited by a tsunami approaching from the open Atlantic ocean approximately 3,000 BP; extreme hurricane waves expose the bimodal material. Photograph: © Anja Scheffers.
The Western Australian coast lies not far from the convergent plate margin near Indonesia. For this reason, it has been Australia's most tsunami-prone region, with two 4-6-m-high tsunami run-up inundations occurring over the past 30 years. In addition, Western Australia also regularly experiences very intense tropical cyclones (willy-willies) that can produce large marine floods. Plainly, to study the sedimentological and geomorphic effects of tsunamis, it must be possible to distinguish them from the effects of storm surges in prehistoric records.
Jonathan Nott and Bryant (2003) surveyed more than 2,500 km of the Western Australian coast looking for evidence of prehistoric ephemeral marine inundations (storm surges or tsunamis). They found wave-transported shell, coral, sand, and boulder deposits atop 30-m-high headlands, elevated sand deposits containing large boulders up to 10 m above sea-level, shell and coral deposits several kilometres inland, and fields of large imbricated boulders across shore platforms. Two facts pointed to a tsunami origin: (1) the elevations of the deposits; and (2) the size of individual clasts. Radiocarbon dating of the deposits allowed the construction of a tsunami chronology. Several tsunamis have occurred over the past millennium. Prehistoric wave events (probably tsunamis) were considerably larger in this region than those that have occurred over the past 115 years since European settlement. Tsunamis with run-up heights of between 10 and 30 m or more seem to have occurred with recurrence intervals of about once every four to five centuries. At some locations, such as Cape Leveque, two major tsunamis have occurred over the past millennium.
The source of the tsunamis remains debatable. Earthquakes emanating from the Timor Trench and the Krakatau volcanic eruption of 1883 caused recent tsunamis in Western Australia. Similar mechanisms of considerably higher magnitude may have generated larger prehistoric tsunamis along the Western Australian coast. Submarine landslides are another possibility. However, undersea landslides usually cause localized tsunamis, and the long stretch of coast bearing evidence of old tsunamis, coupled with the shallowness of the continental shelf between Cape Leveque and North West Cape, where the 500-m isobath lies over 400 km offshore, hint that submarine landslides are an unlikely source. Another possibility is bolide impacts. In a 1,000-year period, any coastal site with 180° exposure and reach of 6,000 km has a 1:12 chance of experiencing a 2-m amplitude tsunami (with a run-up height of about 10 m), and 1:35 chance of experiencing a 5-m or greater tsunami (with a run-up height of 25 m or more) from asteroid impact in the ocean (Ward and Asphaug 2000). This means that Perth, in southwestern Western Australia, has a 9.95 per cent chance and 3.41 per chance of experiencing 2-m and 5-m tsunami amplitudes, respectively, from an oceanic asteroid impact (Nott and Bryant 2003). Thus, asteroid impact in the Indian Ocean is a possible source for the tsunamis that seem to have struck the Western Australian coast.
Several deposits on islands of the Hawaiian group suggest transport by giant waves, possibly generated by giant landslides on the submarine flanks of the Hawaiian Ridge, which attain lengths of 200 km (Moore et al. 1989, 1994). Such waves apparently laid down gravel deposits on Lanai and Molokai some 100,000 years ago (Moore and Moore 1984; Moore et al. 1994). On the island of Lanai, a tsunami during the last interglacial might be responsible for depositing gravel at heights of more than 300 m. The origin of these coastal and high-elevation marine gravels on the Hawaiian islands of Lanai and Molokai is controversial, because the vertical tectonics of these islands is poorly constrained (Felton et al. 2000). Massive tsunamis from offshore giant landslides might have produced them, but highstands of sea levels are another possibility. However, at Kohala on the island of Hawaii, continuous subsidence is well established. Lithofacies analysis and dating of a fossiliferous marine conglomerate 1.5-61 m above present sea level support a tsunami origin and indicate a run-up of more than 400 m that would reach more than 6 km inland (McMurtry et al. 2004). The conglomerate is 110,000 years old (± 10,000 years), which makes a tsunami caused by the approximately 120,000-year-old giant Alika 2 landslide from nearby Mauna Loa volcano a possible cause.
Giant landslides on the Hawaiian Islands would produce tsunami trains that would move out radially across the Pacific Ocean, eventually reaching continents around the Pacific Rim. The eastern seaboard of Australia seems to bear evidence of their passage (Bryant et al. 1992). A catastrophic tsunami almost totally demolished sand barriers along the coast of southern New South Wales, and vestiges of catastrophic wave erosion on coastal abrasion ramps are evident at least 15 m above present sea-level. The barriers, which date from the last interglacial, appear to have been destroyed about 105,000 years ago, probably by the tsunamis generated near Hawaii.
On the evening of 26 January 1700, a 'tremendous upheaval offshore from Oregon and Washington caused 600-1000 km of the coast to drop up to l-2-m below sea level'
(Goldfinger et al. 2003, 556). An enormous earthquake had occurred along the Cascadia subduction zone, where the Juan de Fuca plate slips below the North American plate. It generated a tsunami that was locally 10-12-m high. It spread across the Pacific, and historical records note its arrival along the Japanese coast. No Europeans were in that part of North America to witness the event, but Native Americans recorded it in oral histories. Historical and palaeoseismic data from the Cascadia coast suggest that the earthquake was a magnitude 9 subduction earthquake (Goldfinger et al. 2003).
In the late 1980s, research in westernmost Washington State revealed intertidal mud buried extensive, well-vegetated lowlands (represented by peaty layers in estuarine sediments) at least six times in the last 7,000 years. In three cases, sheets of sand also buried the lowlands. Tsunamis created by rapid tectonic subsidence (in the range 0.5-2 m) along the outer coast of Washington State may have caused these burials (Atwater 1987; Atwater et al. 1991). The subsidence was associated with large earthquakes (magnitude 8 or 9) emanating from the Cascadia subduction zone. Later research based on sediment cores collected along the continental margins of western North America found evidence for a 10,000-year earthquake record from two major fault systems (Goldfinger et al. 2003). Thirteen earthquakes seem to have ruptured the entire margin from Vancouver Island to at least the California border since the eruption of the Mazama ash 7,700 years ago. The 13 events above this ash layer have an average repeat time of 600 years, the youngest event around 300 years ago coinciding with the coastal record. Other earthquakes occurred after the base of the Holocene (at least 9,800 years ago), bringing the total to 18, each of which triggered turbid flows in several Washington channels.
Was this article helpful?