Abstract by Adam Robertson
Combining Topography Files for Tsunami Wave Back-propagation Estimation
In the tsunami wave backpropagation problem, the accuracy of topography files modeling the ocean floor affects the reliability of the prediction. As part of our model of a tsunami (1852) in Indonesia, we combined data from multiple topography sources to produce the best possible approximation. We have developed a method to (i) slice accurate land data from relevant regions of a higher resolution file, (ii) interpolate over the corresponding regions of a low resolution file to approximate the sea floor, and (iii) replace missing values in the high resolution file with the approximations from the interpolated data. Because some data sources provide non-uniform samples of the topography, we implement methods of interpolating over non-uniform samples. We discuss several iterations of this process and provide estimates of the success of our current model.