Co-Occurrence of Extreme Precipitation, Sea Level and Wind: Statistical Perspective

Cost-effective strategies to mitigate the increased risks of coastal hazards from climate change require reliable joint likelihood estimates of high precipitation and sea levels across various wind velocities. This paper describes a statistical approach for computing the extremal dependence between such variables using trivariate copulas, employing the method to a long stretch of daily maxima (from 1973 to 2023) at Bridgeport, Connecticut (a location on the Long Island Sound). Our results indicate that wind conditions substantially affect the extremal dependence between precipitation and sea level, with higher westward wind velocities resulting in stronger dependence. Also, while a single-parameter Archimedean family of copulas (example, the Gumbel copula) is sufficient for westward wind velocities, other wind regimes require the use of the more flexible t-copula. Overall, our findings offer valuable insights into linking wind-modulated tail dependence between precipitation and sea levels to compound flood risk modeling, enabling more accurate estimation of joint exceedance probabilities under storm conditions, and inform the development of more resilient coastal flood risk mitigation strategies.