PERSPECTIVES AND CHALLENGES FOR ACOUSTIC DETECTION AND NUMERICAL MODELS OF LANDSLIDE INDUCED LANDSLIDES IN THE GULF OF MEXICO
The studies of the Papua New Guinea tsunami (2200 casualties) of July 1998 have been groundbreaking in demonstrating how a relatively small deepwater mass failure can cause devastating local tsunamis without any warning. A development of the cost efficient and reliable landslide monitoring system in the GoM high risk areas and the establishment of a tsunami warning network associated with the GoM landslides are of great necessity. I will discuss the challenges associated to setting a monitoring system combined to numerical procedures for accurately detecting the initiation and the development of marine landslides, and the consequent potential underwater landslide.
The region in the Gulf of Mexico where marine landslide are most frequent is the Mississippi Canyon. Detection of underwater soil movement can done with remote technique, such as underwater acoustics, or onsite technologies, such as high precision GPS. I will discuss the potential of data collection for performing three-dimensional computations to relate the morphology and slope of the scarp. Fast parallel 3D software is required for modeling the dynamics of the landslide geometry and the dynamics of the water free surface (Grilli et al, 2002). We will show some examples of tsunami propagation software run with the Boundary Element Method, and test different sources.