An Improved Physics-Based Hurricane Track Model over the North Atlantic Basin with Its Application for Wind-Hazard Assessment

A stochastic tropical cyclone (TC) track model is an essential tool to robustly assess the TC hazards and associated risks for various applications. The present study developed a physics-driven track model for the North Atlantic (NA) basin, which captures explicitly the effects of a comprehensive set of environmental variables. The model has three fully coupled modules. First, the random seeding technique was utilized for TC genesis modeling. Second, a new regionalized beta drift model was developed to be compatible with a physics-based TC trajectory model, and for the first time it was calibrated for the NA basin and serves as a refinement over the existing constant or latitude-weighted beta drift (BD) model. Third, an improved physics-based FAST intensity model for landfalling TCs was derived, based on the parametrization of the time-dependent topographic effect, resulting in a 53% improvement in predicting the time of decaying to half of the initial TC intensity over the previous model. An extensive examination of the performance of the proposed model was carried out using the historical best-track data from the Hurricane Database (HURDAT). The new model shows excellent temporal agreement with historical tracks and better spatial consistency when compared to the existing state-of-the-art physics-based models. The model validation was extended to the comparison of the statistics of key TC variables for kilometer posts (KPs) along the coastline, confirming the overall good performance of the proposed model; the additional sensitivity analysis also reveals that the performance is affected by the small sample size of historical tracks and the consideration of the different radii for extracting track data. The validation using historical tracks and comparison with the ASCE 7-22 design code values show that the results using our model are comparable and have similar trends, while the performance varies between kilometer posts (KPs). Finally, on the basis of this model, the 50- and 500-year return period TC wind-hazard curves were evaluated and discussed for the considered region.