large earthquakes in the Cascadia subduction zone. Because no large earthquakes—or locally generated tsunamis—have occurred there within the short historical record, there has heretofore been essentially no planning by public agencies for such a contingency. Because the effects of a large tsunami could indeed be devastating in many low-lying, highly populated areas throughout this region, the issue has now become one of great societal as well as scientific interest. Perhaps the most revealing lines of evidence have come, and will continue to come, from geological field studies of the direct effects of recent abrupt changes in elevation and the resulting tsunamis in the very young sediments of coastal areas as well as changes in the coastal morphology.
Closely related to the tsunami hazard is seiching, or the oscillation of closed and partially open bodies of water caused by long-period surface waves, which are often produced by the same large earthquakes that generate significant tsunamis. For example, the 1964 Alaskan earthquake caused waves as high as 2 m in bays and channels along the Gulf of Mexico, with sizes and shapes suitable for resonating with the wavelength of the arriving long-period seismic energy. In addition to a tsunami that might be caused by a large subduction-zone earthquake off the Pacific Northwest coast, damaging seiches might be generated by the same seismic event on water bodies such as Lake Washington in Seattle, in the Straits of Juan de Fuca and Georgia, on individual arms of Puget Sound, and in large bays such as Grays Harbor and San Francisco Bay.
Although the basic physics of tsunami generation, wave propagation, modification, and run-up on the shallowing shore are generally understood, the oceanic waves and their effects have had little systemic and comprehensive examination. A tsunami research planning group commissioned by the National Science Foundation in 1985 recommended, with highest priority, that a major effort be made to gather meaningful field data related to the generation and propagation of tsunamis. Specifically, the group suggested that arrays of seafloor pressure sensors be placed in two areas where there is a high likelihood of large earthquakes that could result in tsunamis—the Shumagin and Yakataga seismic gaps in Alaska. Their purpose would be to capture the onset of a tsunami and provide critical data on the time histories of shallow-water surface elevation and relative bottom displacement. As part of the same experiment, one or more deep-water arrays of bottom-pressure sensors should be established at greater distances, perhaps north of the Hawaiian Islands and off the California shore, for measuring water-surface elevation, wave directions, and seismic spectra characteristics. Such data are essential for evaluating theoretical models of tsunami generation, propagation, and run-up.
Tsunamis are generated by long-period seismic processes, and it is therefore important to determine the long seismic period nature—those in excess of 100—of earthquakes in areas that might generate tsunamis. Whereas abundant long-period records of such earthquakes are available from instruments located at great distances from the source, almost no close-in, long-period records are currently available because no appropriate instrumentation is in place. The research planning group recommended that