Performance of Physics Based and Dynamics Based Cyclone Tracking Algorithm Over the North Indian Ocean

ABSTRACT Ocean‐atmospheric conditions associated with the tropical cyclones are widely studied using reanalysis datasets. However, the tracking of tropical cyclones using the reanalysis in the different ocean basins is still debatable, and basin‐specific track accuracy remains underexplored. In this study, we have demonstrated that the tropical cyclone detection and their tracking using both physics based and dynamics based algorithms display significant variations in the north Indian Ocean from the diagnostics of ERA5 reanalysis for the period 1990–2022. Comparison of observed TC with reanalysis reveals that, among all the algorithms tested (UZ, CNRM, OWZ, TRACK) the physics based algorithm (i.e., CNRM and UZ) outperforms the dynamics based algorithm (i.e., OWZ), and offers a realistic representation of TC tracks with approximately 80% probability of detection. Additionally, the track error is less than 50 km for approximately 65% of TC points, highlighting the physics based tracker's ability to reproduce observed trajectories with high spatial accuracy. Also, physics based algorithm better represents spatial distribution and interannual variability of TC frequency, and TC duration compared to dynamics based algorithm. Given the fact that TC metrics are highly sensitive to the choice of tracker, this study identifies physics based (CNRM) algorithm as the most effective tracking tool for the north Indian Ocean, thereby providing the first detailed, track‐oriented performance for this basin. This study advocates from reanalysis and climate models that physics based algorithms is the reliable method for TC detection and tracking in North Indian Ocean with implications for future projections.