Flank-collapse on Ta'u Tsland, Samoan Archipelago: Timing and hazard implications



Publication Details

Williams, S. P., Davies, T. R., Barrows, T. T., Jackson, M. G., Hart, S. R. & Cole, J. W. (2014). Flank-collapse on Ta'u Tsland, Samoan Archipelago: Timing and hazard implications. In K. Sassa, P. Canuti & Y. Yin (Eds.), Landslide Science for a Safer Geoenvironment. Volume 3: Targeted Landslides (pp. 583-588). Cham, Switzerland: Springer.

Additional Publication Information

ISBN: 9783319049960


A discrepancy between the cartographic depiction of Ta'u Island, Samoan archipelago, in 1849 and its present geomorphology, leads to the impression that a massive collapse involving an estimated 30 km3 occurred on the island's southern flank less than 170 years ago. It is likely that this flank-collapse, whenever it occurred, generated a tsunami with regional impacts. Here we apply exposure dating to the remnant landslide scarp using the cosmogenic nuclide 36Cl, to show that the flank-collapse occurred 22.4 1.8 ka during the last glacial maximum (LGM). The collapse may have been triggered due to volcanic-related processes, but it is also possible that climatic-eustatic sea-level during the LGM may have played a role in influencing failure of the flank. We confirm that the initial cartographic depiction of Ta'u in 1849 was incorrect, and that this prehistoric landslide-tsunami was not a societal hazard at the time of its occurrence. This is because the Samoan and surrounding Island Nations were only inhabited about 3 ka or so. Nevertheless, we suggest that geomorphic features similar to the Ta'u flank-collapse on analogous islands and seamounts in the Pacific likely represent signatures of landslide-tsunamis in the past. We conclude that there is a need to identify and date other such features in the Pacific, in order to further improve our spatial and geochronological understanding of these events. There is also a need to identify flank features that have not yet failed, and assess the likelymechanisms that could potentially trigger failure. By doing this, we can start assessing with more confidence the hazard potential of similar flank-collapses in future-a risk that is presently under-represented.

Please refer to publisher version or contact your library.



Link to publisher version (DOI)