Preservation potential of tsunami deposits on arid siliciclastic coasts

Numerous post-tsunami surveys have been conducted in the last two decades, especially since the 2004 Indian Ocean tsunami. These studies have documented a variety of characteristic sedimentary and erosional features that can be ascribed to known events. Nevertheless, the question arises whether these structures are just ephemeral or have a potential to be preserved in the geological record. This review describes the changes that have affected muddy to sandy siliciclastic tsunami deposits in Peru. Each of these was surveyed in the first months after the tsunami: Chimbote (1996), Camand (2001) and Pisco-Paracas (2007). Here, we describe the changes we observed during re-surveys in 2007 and 2008. It has long been recognized that onshore tsunami deposits may suffer from surficial processes, tectonic movements and anthropogenic alteration. Earthquake-induced uplift or subsidence may subject a tsunami deposit to erosion or burial, respectively. Quick burial in rapidly subsiding coastal areas may enhance preservation. Deposits of the last or most landward-reaching wave may be preferentially preserved if they escape erosion by subsequent tsunami waves; however, inland areas are also vulnerable to subaerial reworking, including by wind and by humans. The Peruvian examples reviewed here show that the preservation of arid-coast tsunami deposits depends on interactions that are more complex that hitherto perceived. These involve sediment type, grain size, depositional setting, co-seismic movement, bioturbation, winds, and anthropogenic modification. In one example, all traces of the tsunami have been removed or reworked by flash floods and ocean waves. In another example, clasts on a coastal plain from tsunami-backwash began to be rounded and abraded by eolian sands immediately after the event. Eolian processes also smoothed and filled tsunami scours. By contrast, muddy tsunami deposits in certain areas escaped erosion by wind, probably because of their greater cohesion. In still another example, 0.5 m of co-seismic uplift was not enough to prevent ocean waves from removing a tsunami sand sheet that had mantled a coastal marsh. The buried record of tsunami deposits on modem coasts may therefore not fully represent the vulnerability of these regions to tsunamigenic hazards.