||The devastating consequences of recent tsunami events in Indonesia (2004) and Japan (2011) have prompted a scientific response in assessing tsunami hazard even in regions where an apparent low risk and/or lack of complete historical tsunami record exists. Although a great uncertainty exists regarding the recurrence rate of large-scale tsunami events in the Gulf of Mexico (GoM) due to sparsity of data, geological and historical evidences indicate that the most likely tsunami hazard could come from a submarine landslide triggered by a moderate earthquake.
In this work, tsunami generation from submarine landslides is modeled using a simplified 3D hydrostatic/non-hydrostatic Navier-Stokes model and the Volume of Fluid method to track fluid interfaces. Once the tsunami wave is developed, wave information is passed to a 2D hydrostatic/non-hydrostatic, depth-averaged model for propagation and inundation of coastal cities.
The models are applied to historical GoM landslide sources as well as failures determined by a probabilistic approach. For the probabilistic assessment, Monte Carlo Simulation is used to determine landslide parameters based on distributions and correlations obtained from observed data. Along fixed transects over the continental slope of the GoM, these parameter values are used together with seismic data to perform slope stability analyses in order to determine the most likely extreme tsunamigenic landslide events. Hazards are identified for specific coastal locations, including maximum runup height, inundation depth/extent, and strong currents affecting the maritime industry, in accordance with the guidelines of the National Tsunami Hazard Mitigation Program.