Tropical Cyclone Rainfall Asymmetries Inferred from GPM-IMERG: A Focus on Lesser Antilles

Abstract

This study utilizes GPM-IMERG satellite rainfall estimates to assess the asymmetric rainfall patterns in 27 tropical cyclones (TCs) across the Lesser Antilles region from 2000 to 2020. The aim is to evaluate whether there is a persistent relationship between precipitation and wind characteristics, which could support improved TC-related flood risk assessment for these islands. With a focus on hurricane and tropical storm categories, the 30-minute precipitation variability was assessed within a radius of 500 km from the TC’s eye during its path in the study area. In addition, TC’s forward speed and wind characteristics, like  TC’s category and the extent of 34-knot winds (R34), are included. The analysis reveals temporal trends, indicating increased TC rainfall events in the study area during the second decade. Correlations show positive relationships between rainfall total (RT), rainfall area (RA), and rainfall intensity at the 90^th percentile (RI_0.9), with RT and RI_0.9 showing the strongest link in the majority of the observations. Contrary to conventional assumptions, this research challenges the idea that highest category TCs in the wind intensity always produce higher rainfall, as we see that higher-category hurricanes such as H4 (209-251km/hr) and H5 (>=252km/hr) were often associated with lower rainfall values in RT and RI_0.9 compared to tropical storms (63 - 118 km/hr). Tropical storms, like higher-category hurricanes, can display large rainfall areas. In addition,  quadrant analysis of rainfall zones around the TC eye highlights that the NE and SE quadrants in TC have significantly more rainfall impact. However, it also reveals the danger posed by weaker quadrants in wind characteristics such as SW and NW, as they can exhibit high rainfall values in RA and RT. The study indicates complex, non-linear relationships between TC’s wind and precipitation characteristics in the Lesser Antilles region. Incorporating the rainfall variability observed in TC dynamics into early warning systems and risk assessment is essential for a more effective emergency response and mitigation planning.

General methodology

The satellite rainfall estimates were obtained within a defined buffer of a diameter of 500km around the TC eye while following the TC trajectory. The buffer was further dissected into quadrant spatial zones  (NE, SE, SW and NW) to provide a detailed perspective on rainfall distribution in different parts of the TC impact area. For each eye position, rainfall characteristics (RT, RA and  RI_0.9) were computed for the whole buffer and later individual quadrant partitions. The computed rainfall characteristics were then investigated for potential correlation relationships with the TC wind intensity. In addition, quadrant rainfall patterns were analyzed for persistence throughout the TC duration.