Characterization and statistical modelling of tropical cyclone wind inflow angles for joint wind speed and direction hazard assessment

The accurate characterization and modelling of tropical cyclone (TC) wind inflow angles is crucial for various lines of scientific research and engineering applications. While numerous studies have delved into TC wind fields, much less attention has been paid to the distributions of wind directions within TCs. Moreover, a comprehensive comparison and quantitative examination of existing inflow angle models in the literature, assessing their performance and efficacy, is notably lacking. In this study, 483 snapshots from the H*Wind database were employed to explore the symmetric and asymmetric characteristics of TC inflow angle distributions. The analysis revealed that, among common TC-related variables, relative angular momentum exhibits the strongest (negative) correlation with inflow angles. Subsequently, a new parametric inflow angle model was developed, demonstrating superior goodness-of-fit in terms of the bias, root-mean-square-error (RMSE), and linear correlation when compared to four existing models. The proposed model was further validated using the National Data Buoy Center (NDBC) data for several historical hurricanes. Finally, the developed model was applied to assess the joint TC wind and direction hazards for three coastal sites in the United States. The findings and the model developed herein possess broad-ranging applications for wind-resistant design and risk assessment of engineering structures within TC-prone regions.