Author(s): Dao My Ha; Pavel Tkalich; Chan Soon
Linked Author(s):
Keywords: Tsunami forecast; Data-driven; Proper Orthogonal Decomposition
Abstract: In this paper, we present a data-driven tsunami forecast method which is based on Proper Orthogonal Decomposition (POD). The method allows for quick and accurate estimations of tsunamis based on the real-time deep ocean buoy sea surface observations such as DART by NOAA. Using a wave propagation model, tsunami propagation scenarios need to be run one time to fill a POD-based computational matrix, which plays role of a high fidelity data base. In forecasting mode POD is capable of estimation of initial tsunami parameters as well as maximum wave height and travel time at every grid point of high resolution grid over the whole computational domain, including nearshore areas. The forecasting method requires minimum computation time and is constrained only by physical assumptions of the utilized tsunami propagation model.
Year: 2010