Implementation and calibration of a nested WW3-SWAN model for simulating waves under varying and extreme weather conditions: Case study of the Algerian coast

Abstract

This study presents an implementation and calibration of high-resolution coupled WaveWatch III and SWAN wave models for the southern coast of the Mediterranean Sea (Algerian coast). Both coupled models were calibrated to ensure more accurate estimates of the extreme wave events, which are the primary concern of all offshore and coastal activities. WaveWatch III model was considered for the coarse grid to provide the boundary conditions for the nested SWAN model, which is more recommended for nearshore wave modeling. The source term packages (ST4 and ST6) proposed in WW3 were evaluated and calibrated by varying the tunable coefficients. Seven offshore buoys were used during the calibration of WW3, and the validation was performed against four Algerian buoys. The calibrated WW3 model was nested within four SWAN model grids (G1 in the Center, G2 and G3 in the East and G4 in the West) along the Algerian coasts. The implemented nested SWAN model for four nearshore areas was calibrated by evaluating different wind growth and whitecapping source term packages and turning the dissipation coefficient (Cds). The SWAN and WW3 models were calibrated, taking into account their performance in estimating severe wave conditions above the 95th percentile and their performance in estimating varying sea state conditions. For the coarse grid, the results indicate that WW3-ST4 physics with beta max equal to 2.5 is more suitable to generate boundary conditions for SWAN model in case of storm waves. For the nested SWAN model physics, the results indicate different performances across each grid: The Komen and Janssen combinations for wind growth and whitecapping source terms, with Cds = 1.5 is recommended at G1 and G3, while Cds = 6.5 with the Komen and Janssen source terms showed a better performance for G2 and the Komen and Komen combinations with Cds = 0.36E-5 is more appropriate for the G4. The results reflect the high sensitivity of the nested WW3-SWAN model parametrization to the coastal study location.