Improving the P-Surge Ensemble System by Incorporating the Relationship between Tropical Cyclone Intensity and Radius of Maximum Wind Errors

The radius of maximum wind (RMW) is an important structural parameter for tropical cyclones (TCs) that plays a critical role in assessing storm surge. Explicitly forecasting the RMW has been shown to improve probabilistic storm surge forecasts; however, the error statistics used to define the Probabilistic Storm Surge (P-Surge) model ensemble of RMW forecasts do not consider the relationship between TC intensity and RMW. This study explores the relationship between TC intensity and RMW forecast errors for Atlantic TCs. A significant correlation is identified between TC intensity errors and RMW errors with intensity errors accounting for roughly 23% of the variance in RMW forecast errors at 72 h. Consideration of the RMW–intensity relationship is found to benefit the P-Surge ensemble. Utilizing a bivariate method based on binned forecast errors shows that an ensemble of TC structure should have reduced intensity uncertainty for weak and large systems but increased intensity uncertainty for intense and small systems. The opposite is true for the RMW where more uncertainty in the RMW distribution is needed for weak and large systems but less uncertainty is re-quired for intense and small TCs. The bivariate method is applied to the P-Surge ensemble and is shown to improve the P-Surge track ensemble by better accounting for forecast uncertainty in both the RMW and TC intensity and by more ac-curately matching the expected wind probability distribution.