Providing quantification measures for geohazardous events is of utmost importance not only for saving lives, but also for sustainable urban or industrial development in potentially vulnerable environments and many other fields in which geohazards play a role. This work implements an uncertainty quantification (UQ) workflow within the ExaHyPE 2 engine for the shallow-water equations (SWE). The SWE are a set of strictly hyperbolic partial differential equations, which are most commonly used in tsunami modeling. Therefore, an SWE application has been developed within ExaHyPE 2, and been verified for correctness. UQ has then been provided in the form of an UM-Bridge model server, which allows for running an SWE simulation with a displaced origin of the earthquake. The validity of this model server has also been verified. Additionally, an approach to guided adaptive mesh refinement (guided AMR) using the adjoint SWE developed by Davis and LeVeque has been implemented. The F-Wave solver mainly used in the SWE application is found to approximate the solution of the SWE reasonably well. But tsunami events failed to be solved correctly for reasons shown to be unrelated to the SWE application. The guided AMR has also been found to work to a certain degree, but here as well, the error is shown to not lie with the SWE application.     «

Providing quantification measures for geohazardous events is of utmost importance not only for saving lives, but also for sustainable urban or industrial development in potentially vulnerable environments and many other fields in which geohazards play a role. This work implements an uncertainty quantification (UQ) workflow within the ExaHyPE 2 engine for the shallow-water equations (SWE). The SWE are a set of strictly hyperbolic partial differential equations, which are most commonly used in tsu...     »