A SYNTHESIS OF CHARACTERISTICS OF SUBMARINE LANDSLIDES GENERATED BY THE 1964 GREAT ALASKA EARTHQUAKE IN SEVEN FJORDS

Submarine landslide-generated tsunamis were the single largest cause of fatalities in the M_w9.2 1964 Great Alaska earthquake. In the last decade, we studied submarine slope failures in seven fjords: Resurrection Bay, Port Valdez, Passage Canal, Dangerous Passage, Aialik Bay, Port Bainbridge, and Harris Bay. The seven fjords lie 20 to 30 km above the Alaska-Aleutian megathrust, providing an ideal landslide trigger mechanism. To characterize the landslides, we use multibeam bathymetry data, pre- and post-event bathymetry differencing, sparker and chirp seismic data, wave runup directions and heights, S-wave velocity profiles, onland sedimentary record of the tsunamis, coseismic observations, sediment cores, and tsunami models. All slides originated at the margins of the fjords, mostly in unconsolidated sediment of fjord-head deltas. The slides transported material up to ~15 km, resulting in slide deposits up to 20 m thick and a megaturbidite deposit up to 15 m thick. These slides resurfaced the entire fjord bottom and the resultant flow of sediment and water brought numerous deep dwelling fish to the surface, killed by the sudden pressure changes. Typical fjord sedimentation resulted in conditions primed for slope failures. Fjord-head deltas deposited unconsolidated sediment at the upper margins of the fjords, which composed the majority of sediment that failed. However, we find the highest tsunami runups are correlated with blocky landslides that required unique depositional conditions. The Little Ice Age glacial expansion led to significant sedimentation at the margins of the fjords. Near Shoup Bay in Port Valdez, in Passage Canal, and probably in Dangerous Passage, ice overrode till and sediment deposited in front of the glaciers. After ice retreat, the consolidated sediment was perched at the margins of the fjords. In the 1964 earthquake, this sediment failed and was transported to the fjord bottoms as large blocks, but did not disaggregate to the extent of sediment deposited at more distal fjord-head deltas, generating tsunamis with high runup. Lastly, we find a significant correlation between glaciated area of the catchment and submarine slide volume. This indicates glacially generated sediment is not stored in valleys and could potentially predict of slide volume for areas that have not yet failed.