The TRR 181 seminar is held by Prof. Dr. Jörn Behrens (Universität Hamburg) on December 2, 11am.
Model-Adaptivity - how to combine increasing model complexity in an adaptive way
In oceanic wave propagation (in particular tsunami simulation) a common model approximation uses shallow water wave theory for pretty accurate results. However, in near shore regimes or in situations of drastically changing bathymetry more sophisticated modeling approaches involving non-hydrostatic effects are more appropriate. While the shallow water approach allows for very efficient large scale simulations, non-hydrostatic equations are computationally much more involved and often not feasible in situations, where fast simulations are required (such as early warning, or probabilistic assessments with many ensemble members).
In this presentation I will develop a strategy to seamlessly compbine hydrostatic (shallow water) and non-nydrostatic (Boussinesq-type) modeling approaches. These two different models will then be used in an adaptive way, switching from (efficient) hydrostatic to (computationally more expensive) non_hydrostatic computation, wherever necessary. An important question then arises, on how to detect those domain areas necessitating the model refinement. A number of indicators will be tested and some preliminary conclusions be drawn.