A Global Tide and Storm Surge Model with a Parallel Unstructured-Grid Shallow Water Solver

Abstract:

Accurate hydrodynamic forecasting of tides and storm surges on a regional scale near coasts is receiving increased attention in the 21st century. For this purpose, regional models are typically used where open-boundary conditions are prescribed. However, if a tide or storm-surge model covers a much larger area than the local coastal region, some additional physical processes become relevant. This is most clear on a global scale, where there are no open boundary conditions to force the model anymore; instead the tides are forced by the gravitational potential within the model.

We present a new global 2D hydrodynamic model to simulate tide and storm surges. The main goal of this model is to zoom in from global to regional scale and to study the impact of various assumptions in regional models. To enable this, the model uses an unstructured grid shallow-water solver capable of representing coastal areas in more detail than the open oceans. This is of importance as much of the tidal energy is dissipated on the shelf, even on a global scale.

The solver used here is D-Flow FM, capable of computing 2D and 3D shallow-water flow on unstructured meshes and developed for applications in e.g. regional seas, coastal areas and estuaries. Here we apply D-Flow FM on a global scale for the first time.

The global tide model was calibrated using the OpenDA toolbox for automated parameter estimation and FES2012 tidal solution. Depth, bed friction and linear friction parameterizing dissipation by internal tides were optimized to improve overall tidal characteristics. Calibration was performed using a coarser version of the global model to reduce computational time, which is then input for a limited number of calibration runs with the original model.

For the purpose of developing an operational version of the model for storm-surge forecasting, we investigate the performance of the parallel version of D-Flow FM for this global model on supercomputers. Here we show some initial results.