Enhanced Global Tropical Cyclone Identification in ERA5 through Bayesian Inference and Dynamic Tracking (BIDTrack) Algorithm

Abstract In this study, the Bayesian Inference and Dynamic Programming Tracking (BIDTrack) algorithm is developed for enhanced global tropical cyclone (TC) tracking in reanalysis datasets, particularly the fifth major global reanalysis produced by the European Centre for Medium-Range Weather Forecasts (ECMWF) (ERA5). BIDTrack addresses challenges like trajectory discontinuities and parameter sensitivity in traditional methods by combining Bayesian inference with dynamic programming. The algorithm is optimized through a Bayesian interval optimization (BIO) process, which refines the parameters to retain cyclone candidates that are statistically significant and physically meaningful. Results indicate a strong spatial correlation between BIDTrack-derived trajectories and International Best Track Archive for Climate Stewardship (IBTrACS) observations, especially in cyclone-prone regions like the North Atlantic and western Pacific. BIDTrack captures both major hurricanes and weak storms, providing a reliable tool for cyclone path reconstruction and climate impact assessments. This research demonstrates BIDTrack’s potential in improving TC tracking and enhancing the understanding of cyclone dynamics in ERA5. Significance Statement Tropical cyclones, such as hurricanes, are powerful storms that pose significant risks to coastal communities. Tracking their paths accurately is crucial for understanding their behavior and mitigating their impacts. In this study, an emerging method, Bayesian Inference and Dynamic Programming Tracking (BIDTrack), is introduced by combining Bayesian inference with dynamic programming to enhance cyclone tracking in the fifth major global reanalysis produced by the European Centre for Medium-Range Weather Forecasts (ECMWF) (ERA5). BIDTrack generates cyclone paths with probability estimates, providing a more precise assessment of whether a given track point corresponds to the actual cyclone. This algorithm is effective in tracking both strong hurricanes and weaker storms, making it a valuable tool for researchers and decision-makers interested in cyclone behavior and climate impacts.