Coupling a Parametric Wave Model into Hydrodynamic Ocean Circulation Forecasts to Improve Efficiency of Nested Estuarine Storm Surge Predictions

Abstract

Efficiency in storm surge modeling is crucial for forecasting coastal hazards in real-time. While computation cost is trivial to organizations with ample resources, the robustness of forecasts generated by most parties are restricted by wall-clock time. The Parametric Wave Model (PARAM) was developed by Boyd and Weaver (2021) as an alternative to computationally expensive wind-wave models when modeling estuarine environments. For this study, PARAM has been tightly coupled with the ADCIRC hydrodynamic model to become ADCparam, then integrated into the Multistage ensemble forecast system, a one-way nesting framework for modeling waves and circulation in coastal estuaries developed by Taeb and Weaver (2019). ADCparam was applied only to the embedded estuarine fine mesh, while the rest of the Multistage system was left unaltered. ADCparam has greatly reduced computation time for the embedded fine mesh sub-model in relation to the third-generation wave model originally used. While the PARM wave solution shows dissimilarities with the SWAN solution, significant wave height and wave period results are consistently similar enough to further pursue the parametric wave ensemble method. ADCparam models demonstrated run times up to 51% faster than ADCIRC coupled with SWAN, an established iterative wave model that comes standard with the ADCIRC package and Multistage repos. ADCparam wall time is comparable to ADCIRC running alone, making the cost of including the wave solver negligible. Computational efficiency is greatly increased, and solution integrity is generally maintained. ADCparam and its application to Multistage are still in the stages of refinement and validation, but have proven to be an efficient, viable path for implementing waves in any estuarine ocean circulation model.