Lohithakshan Parambath, Lisha (2014-04). Identification of Submarine Landslide for Tsunami Hazard Assessment in the Gulf of Mexico Using a Probabilistic Approach. Master's Thesis.
Thesis
The eastern coast of USA, including the Gulf of Mexico (GOM), is more prone to tsunamis caused by submarine landslides than earthquakes. The Tsunami Hazard Assessment research program lead by ten Brink, 2009, reported the presence of ancient submarine landslides deposit in the GOM dating back to the post glacial period which indicates that there is a likelihood for tsunami events in the future. In fact, the GOM has some of the largest submarine landslides when compared to landslides off the coast of Oregon, central California and New Jersey. Moreover, the high population density and the ongoing industrial development in the GOM, makes it necessary to assess the hazard and develop mitigation plans that involve the development of inundation map, education, early warning and evacuation plans. Specifically in the GOM, assessing the tsunami hazard is to develop tsunami inundation map to identify potential submarine landslide sources, either by using a probabilistic approach or a deterministic approach that uses worst case landslide-tsunami scenarios. A probabilistic approach in the GOM is more suited due to the lack of earlier records of tsunami caused by submarine landslides. Thus the probabilistic model can mimic or create tsunami scenarios based on distribution of physical and geometrical variables involve in the landslide-tsunami mechanisms. Monte Carlo Simulation (MCS) is the tool used to generate random variables under certain distribution, and the MCS Model for the GOM generates a large number of submarine landslides with randomized parameters (like location, runout length, depth, headscarp height, width, slope etc.) capable of producing tsunamis. Parameter results are validated to verify if their distribution follow the same distribution from observed landslide events.
