Megatsunami Deposits on the Island of Hawaii: Implications for the Origin of Similar Deposits in Hawaii and Confirmation of the `Giant Wave Hypothesis'

Deposits from giant waves in the Hawaiian Islands have been linked to giant submarine landslides (GSLs) offshore, for which there is abundant geophysical evidence. The setting of the deposits is ambiguous, however, so variations in sea level, local tectonics, stream deposition, or anthropogenic modification, have each been posed as alternatives to deposition from a landslide-generated tsunami. During a reconnaissance in May 2002, we found sedimentary sequences of marine and non-marine origin on the northwest flank of Kohala Volcano, Island of Hawaii. Our preliminary fieldwork at these locations, previous isotopic age dating of Kohala rock sequences, and the known local tectonism of the island, together provide compelling evidence that the marine sequences were deposited by at least one, if not two, megatsunamis during the past 300,000 years. That inference has been largely supported by sedimentary facies analysis based on petrography and paleontology. Based upon the present maximum known exposure elevations and the known subsidence history of Kohala, we show that two distinct calcareous basalt boulder conglomerates, of probable ages of 110 +/- 10 ka and 255 +/- 10 ka, were deposited at paleo-elevations of greater than 330 m and 750 m above MSL. These ages correspond to the eustatic sea level rises at stages 5e and 7, but the large elevations discount sea level high stands as the cause of deposition. The run-ups from tsunamis resulting from GSLs offshore are modeled in hundreds of meters, in broad agreement with the elevations of the calcareous basalt conglomerates we have found. Preliminary interpretations suggest that the conglomerates are of similar age to the known GSLs offshore: Alika phase 2 for the younger deposit (ca. 130 ka BP) and Alika phase 1 or South Kona for the older deposit (ca. 240 ka BP). Some of the Kohala marine conglomerates are very fossiliferous and some are barren; they unconformably overlie paleosols. For age control, we presently rely upon a U-series age of an in situ coral fragment determined by alpha spectrometry (Stearns, 1973) and on stratigraphic relations with mugearite lavas that were previously dated by K-Ar techniques. Our results have important implications for hydrodynamic models of megatsunamis. Using a rough 1/3 rule, the inferred wave heights of greater than 100 m (stage 5e) and 250 m (stage 7) are 2 to 4 times in excess of recent modeling for this area. Our results also imply that similar deposits previously described on the islands of Lanai, Molokai, Maui and Oahu may indeed originate from megatsunamis. We strongly suspect that when more precise ages for these deposits are forthcoming, they will support the hypothesis of a changing climate trigger within the transgressions of the last two interglacial periods, which also suggests why these deposits are sometimes confused with sea level high stands.