Rapid tsunami propagation and inundation models from time-dependent earthquake source inversions from land- and ocean-based geophysical sensors

Tuesday, 16 December 2014
Diego Melgar1, Yehuda Bock2 and Jessie K Saunders1, (1)Scripps Institution of Oceanography, La Jolla, CA, United States, (2)UCSD/IGPP 0225, La Jolla, CA, United States
Significant improvements in geophysical sensor deployments and rapid response have made it possible to compute a suite of earthquake source models in the first seconds to minutes after rupture initiation. With an example application for the 2011 Mw9.0 Tohoku-oki event we present results from kinematic earthquake source inversion of regional geophysical data including strong motion, GPS, buoy, and ocean bottom pressure data. We will show that the time dependent earthquake source models thus derived provide suitable initial conditions for tsunami propagation and inundation modeling. In addition to the time-varying deformation of the sea-floor, the algorithm we present also accounts for the tsunamigenic effects of the horizontal advection of steep bathymetric features. With a detailed comparison to post-event tsunami survey measurements we will show that kinematic earthquake source models from strictly land-based data are a significant improvement over static inversions and provide reasonable estimates of expected tsunami intensity in the near source region. We will then show that adding ocean-based observations dramatically improves the inundation forecast. We will conclude that with current land-based deployments of geophysical sensors it should be feasible to produce rapid estimates of tsunami intensity which are reliable enough and useful for guiding early warning. We also conclude that adding just a few ocean-based sensors into the estimation process can drastically improve the reliability of such forecasts and warnings.