We have created a modeling community testbed to help with the development of improved predictive capabilities for coastal waves, storm surge and inundation associated with tropical and extra-tropical storms. The testbed enables the sharing of model grids, forcing data, observational data, model results, and model evaluation tools. Outcomes include quantification of model skill, computational efficiency, and sensitivity to implementation issues such as the physics included in the model and the grid resolution.

This presentation focuses on the tropical component of the testbed and compares a variety of models as well as grids in simulating wave and surge environments for two well documented historical hurricanes, Hurricanes Rita (2005) and Ike (2008). The testbed domain includes the Gulf of Mexico with higher resolution provided along the Louisiana – Texas shelf to model the regional responses to hurricanes. Models included in the testbed are ADCIRC, FVCOM, SELFE, SLOSH, SWAN, SWAVE, WWM and WWIII. The Interactive Model Evaluation and Diagnostics System (IMEDS) toolkit is being used to implement a common set of skill assessment metrics across the testbed. Testbed activities include both the evaluation of model skill vs observed data and the inter-comparison of model results. As important, extensive work has been done to better understand the necessary spatial resolution both offshore and onshore, needed to reliably simulate waves and surge fields. In addition to model validation, computer resource performance and scalability of the inter-model and inter-grid comparisons are examined.

The ultimate goal is to provide guidance on model selection and implementation to end users that have operational modeling needs in this field. The testbed was initiated and has been substantially supported by the Southeastern Universities Research Association (SURA) with funding provided by the NOAA IOOS program.