Minutes after the 1964 magnitude-9.2 Great Alaska Earthquake began shaking, a series of tsunami waves swept through the village of Chenega in Prince William Sound, destroying all but two of the buildings and killing 23 of the 75 inhabitants. Fifty years later, scientists from the U.S. Geological Survey revealed the likely cause of the tsunami, a large set of underwater landslides. The scientists used detailed seafloor images not only clear up a decades-old mystery, but also underscore the tsunami hazard that submarine landslides can pose in fjords around the world where communities and ports are commonly located.
USGS geologists mapping Alaska's south coast shortly after the 1964 Great Alaska Earthquake had speculated that a submarine landslide might have triggered the Chenega tsunami, just as known landslides triggered tsunamis that devastated the Alaskan towns of Valdez, Seward and Whittier. But a bathymetric survey at the time, which only imaged relatively shallow seafloor, down to 180 meters (330 feet) deep, did not reveal evidence of a landslide in nearby Dangerous Passage or the other waterways around Chenega. Alternate explanations involving seafloor movement during the earthquake did not fit the timing and severity of the Chenega tsunami as described by eyewitnesses, explained Daniel Brothers, USGS geophysicist and lead author of the study.
The tsunami origin remained uncertain until a team led by Brothers mapped a large submarine landslide complex in Dangerous Passage, mostly in water deeper than what was studied in 1964.
What makes this slide unusual is that much of the material that slid was at a water depth of 250 to 350 meters (820 to 1150 feet), said Peter Haeussler, USGS geologist and a coauthor of the report. The deeper initiation depth made it particularly good at generating a tsunami.
The scientists used multibeam sonar technology to collect high-resolution bathymetric (seafloor depth) data, and a single-channel seismic-reflection system to collect sub-bottom profiles (cross-sectional views of sedimentary layers and other features beneath the seafloor).
The researchers calculated the time it would take for a tsunami triggered by a large landslide in the mapped areas to reach the village of Chenega and found a good fit with eyewitness reports: a tsunami wave triggered in the areas where they found landslide evidence would take three to four minutes to reach the village, consistent with the arrival time of the most destructive waves.
It is exciting to see the technology evolve so we can now get high-resolution images of the seafloor that we could not back then and to pinpoint the most likely source for the waves. After 50 years, this new work confirms our original inference that it was probably landslide-generated waves that devastated Chenega so many years ago, but we had no adequate submarine data to define either the size or location of the landslide sources, said USGS geologist emeritus George Plafker who, with colleague Larry Mayo, was one of the first responders and wrote some of the early geological field reports on surface effects of the Chenega waves in 1965.
The modern bathymetric data revealed three sedimentary basins at progressively deeper levels toward the open waters of Prince William Sound. These basins were originally carved by descending glaciers when sea level was lower. As sea level rose and the glaciers retreated after the last ice age, the basins filled with sediment eroded and washed off the land. The basins are separated by sills, likely terminal glacial moraines, and bounded by steep, rugged slopes. Details of the underwater terrain include scarps, blocks, hummocky surfaces and other features typical of landslides. Similarly, the sub-bottom seismic profiles show that the basins are filled with fine-grained sediment, with layers disrupted in ways that indicate multiple landslides.
The new report by USGS scientists and their colleagues from Boise State University and the Alaska Department of Fish and Game, A submarine landslide source for the devastating 1964 Chenega tsunami, southern Alaska, was published in the journal, Earth and Planetary Science Letters.