GEOLOGIC EVIDENCE FOR SUBDUCTION EARTHQUAKES

Some of Earth’s largest earthquakes occur at subduction zones. At the margin of the Pacific Ocean, hundreds of millions of people and trillions of dollars of economic infrastructure are at risk from these earthquakes. The frequency, magnitudes, and potential damaging effects of subduction earthquakes can be inferred using four types of geologic evidence. The first type of evidence is land-level change, which is documented in tidal marshes. Peaty marsh soils are sharply overlain by tidal muds in areas that experience more than about 0.5 m of coseismic subsidence. The amount of subsidence can be estimated from assemblages of foraminifera, diatoms, and pollen within the sediments. In areas of coseismic uplift, tidal muds are abruptly overlain by marsh or forest soils. As in the zone of coseismic subsidence, fossil assemblages can be used to estimate the amount of uplift. The second type of evidence is tsunami deposits. Subduction earthquakes trigger teletsunamis that spread sheets of coarse sediment in tidal marshes, low coastal lakes, and other low-lying coastal environments. In the case of subduction earthquakes, coarse tsunami sediments directly overlie coseismically subsided or elevated soils. The third type of evidence is liquefaction features. Dykes and mounds (“blows” or "volcanoes") of gravel, sand, and silt are produced by all large earthquakes and provide direct evidence of strong ground motion. However, attribution of these features to subduction earthquakes can be indirect or unproven. The fourth type of evidence is landslides. Ground motion during subduction earthquakes may trigger landslides over large areas. However, as in the case of liquefaction features, attribution of landslides to subduction earthquakes is problematic. Seismic shaking may trigger slope failures at the edge of the continental shelf. These failures, in turn, may produce large turbidity currents, which leave deposits in basins at the base of the continental slope. The strongest arguments for past subduction earthquakes are based on a temporal and spatial association of two or more of the four types of evidence mentioned above. In addition, the evidence should come from a large area, spanning sites located over distances of at least 100 km along the length of the subduction zone.