Validation of the COAMPS-TC Ensemble Prediction System and a Summary of Recent Upgrades

The COAMPS-TC Ensemble Prediction System (EPS) is a regional dynamical model ensemble designed to provide probabilistic forecast guidance for tropical cyclone (TC) track, intensity, and structure. An 11-member version of the COAMPS-TC EPS has been run operationally since 2020 by the Fleet Numerical Meteorology and Oceanography Center (FNMOC). NRL-Monterey has also run an experimental real-time 21-member version of the COAMPS-TC EPS in recent years, accumulating a unique dataset consisting of thousands of regional dynamical model ensemble TC forecasts.

Here we utilize the aforementioned large sample of 21-member NRL-Monterey COAMPS-TC EPS forecasts to comprehensively assess the performance of the system. We compare the accuracy and bias of the COAMPS-TC EPS ensemble mean forecast to deterministic COAMPS-TC and other operational deterministic regional dynamical TC models. We also use probabilistic validation metrics to assess the quality of COAMPS-TC EPS predictions, in particular those for intensity and intensity change. As part of the aforementioned validation work, we examine the relative performance of 11-member and 21-member ensembles. Finally, we describe the latest updates to the COAMPS-TC EPS that were transitioned to FNMOC in 2024, including a greatly expanded suite of graphical forecast products.