The Sensitivity of SLOSH Storm Surge Prediction to Wind Forecast Model Selection

Operational forecasting of storm surge combines the uncertainties and approximations of the predicted wind and pressure fields with those of the coastal surge model. While many studies have addressed the performance of different storm surge models forced by high-quality, data-assimilated, hindcast wind / pressure fields, much less has been done to isolate the impact of the wind / pressure input on storm surge model performance, particular in forecast scenarios. To address this, we evaluate storm surge predictions from SLOSH, the official operational storm surge forecast model of the National Weather Service, using three storm forcings. Specifically, we compare storm surge predictions that are forced by data-assimilated, hindcast wind / pressure fields and two parametric models that are appropriate for forecasting. One of the parametric wind models is included in SLOSH while the other is a generalized version of the Holland parametric model that utilizes all of the information available in the National Hurricane Center's official forecasts, including the distances to all isotachs specified in the four quadrants of the storm. Predicted winds and water levels are compared to observed data for several storms and regions to identify trends in the performance of wind model and storm surge model prediction.