Physical Modeling of 3D Tsunami Evolution Using a Landslide Tsunami Generator
PI: Fritz, Hermann, Georgia Tech
Co-PI: Puzrin, Alexander, Institute for Geotechnical Engineering
Co-PI: Germanovich, Leonid, Georgia Tech

NEES Consortium, Inc.
NEES at Oregon State
Tsunami Research

Tsunamis are gigantic ocean waves, normally associated with submarine earthquakes. Ten devastating tsunamis claiming more than 4000 lives have been recorded only in the last decade of the twentieth century (Gonzalez, 1999), mostly in the Pacific Ocean. In the latest event on July 17, 1998, a 15 m high tsunami wave hit the coast of Papua New Guinea and took more than 2200 lives (Synolakis et al, 2002).

Tsunamis can be generated directly by seismic impact (Kanamori and Kikuchi, 1993). However, in some seismic events, termed "tsunami earthquakes," tsunamis appear to be much larger than expected from their seismic waves (Kanamori, 1972). Examples of tsunami earthquakes include the 1896 earthquake in Sanriku (Japan), the 1946 Aleutian earthquake, the 1963 and 1975 events in Kuriles, and the 1960 earthquake in Peru (Okal, 1993). For some of these earthquakes, it has been proposed (Kanamori, 1985) that the caused tsunamis were triggered by the massive failure of the sea floor in the form of giant submarine landslides. This hypothesis has been supported by direct observations of such tsunami generating or tsunamigenic landslides as those triggered by earthquakes off the Grand Banks of Newfoundland (1929) (Heezen and Ewing, 1952), Alaska (1958) (Miller, 1960; Fritz et al., 2001), Alaska (1964) (Coulter and Migliaccio, 1966; Plafker et al., 1969), and Nothern California (1980) (Field et al, 1982). According to the National Geophysical Data Center and World Data Center for Solid Earth Geophysics as well as Intergovernmental Oceanographic Commission (Brayant, 2000), in the Pacific Ocean alone, 65 tsunami events attributed to submarine landslide caused a total number of 14,661 deaths.

Our ultimate long-term goal is to develop a fundamental understanding of the mechanism of tsunamigenic landslides and subsequent tsunami generation, propagation, and run-up. This would allow for improved assessment and possible mitigation of the landslide and tsunami hazard. Unfortunately, the field data from real world observations are limited to very few cases, while an important part of this data related to the tsunami generation stage is widely missing. The goal of the proposed research is to compensate for this lack of data by the physical modeling of 3-dimensional tsunami evolution using a novel landslide tsunami generator, which will complement the existing NEES tsunami facilities at OSU, Oregon. The design parameters for the proposed landslide tsunami generator will be determined using governing model similitude and dimensional analysis of the landslide and tsunami characteristics. The size of the NEES tsunami wave basin will provide us with the unique opportunity to study scale effects of tsunami generation and propagation based on a non-dimensional scale series. This apparatus will enable individual control of the dynamic landslide parameters such as landslide location, geometry and acceleration.

For design of the experimental program some of the geometries and velocities of tsunamigenic landslides will be determined from the case histories. However, in order to cover the whole range of possible parameter combinations we propose to use the novel analytical and numerical models of the shear-band propagation mechanisms of tsunamigenic landslides. In the proposed experiments we are going to measure the characteristics of the subaqueous landslide motion and the near-field tsunami generation, propagation and run-up in three dimensions. In addition the resulting landslide deposits will be mapped and their thickness recorded. The measured landslide and tsunami characteristics will be compared to the existing real world observations providing a validation for physical, analytical and numerical models used in this research. This will justify the use of these models in the development of fundamental understanding of the landslide and tsunami mechanisms. This improved understanding will allow us to revisit the conventional assumptions used in the numerical modeling of landslide generated tsunamis and to develop recommendations for the future experimental and theoretical research.