Dynamically adaptive data-driven simulation of extreme hydrological flows

Hydrological hazards such as storm surges, tsunamis, and rainfall-induced flooding are physically complex events that are costly in loss of human life and economic productivity. Many such disasters could be mitigated through improved emergency evacuation in real-time and through the development of resilient infrastructure based on knowledge of how systems respond to extreme events. Datadriven computational modeling is a critical technology underpinning these efforts. This investigation focuses on the novel combination of methodologies in forward simulation and data assimilation. The forward geophysical model utilizes adaptive mesh refinement (AMR), a process by which a computational mesh can adapt in time and space based on the current state of a simulation. The forward solution is combined with ensemble based data assimilation methods, whereby observations from an event are assimilated into the forward simulation to improve the veracity of the solution, or used to invert for uncertain physical parameters. The novelty in our approach is the tight two-way coupling of AMR and ensemble filtering techniques. The data assimilation system is implemented on various test cases that delve into the aspects of ensemble based assimilation filters. Additionally, data assimilation on tsunami models is analyzed and a methodology to map the uncertainties in the seabed deformation due to the associated earthquake to the water surface elevation forecast has been presented. Further, using other simulated environments such as the Chile tsunami event of February 2010, a systematic way to calibrate the assimilation system is presented. Finally, the technology is tested by assimilating actual gauge data from the Tohoku tsunami event. These advances offer the promise of significantly transforming data-driven, real-time modeling of hydrological hazards, with potentially broader applications in other science domains.Engineering Mechanic