Evaluation of Three Genesis Potential Indices for Tropical Cyclogenesis in the Arabian Sea: Two Case Studies Using WRF and ERA5

Abstract

Tropical cyclones (TCs) generated over the Arabian Sea can cause significant damage to infrastructure, human lives, landfall, and property near inshore and maritime trade route areas. A key to successful prediction of TCs is a skillful prediction of potential cyclogenesis locations. This study focuses on evaluating three genesis potential indices (GPIs) derived from a global reanalysis (ERA5) and dynamically downscaling using a regional model (WRF) for two TC cases: Gonu in 2007 and Kyarr in 2019, selected by analyzing the accumulated cyclone energy trend from the International Best Track Archive for Climate Stewardship (IBTrACS) dataset over the period of 1981–2019. The two TCs belong to category 4 and above on the Saffir–Simpson scale. To test the sensitivity of downscaling to cumulus parameterizations, two WRF experiments were conducted using the Kain–Fritsch and New Tiedke cumulus schemes, respectively. The calculated genesis locations with help of the three GPIs were compared with IBTrACS. The results show that 1) all indices have reasonable skills in reproducing genesis locations, although their performances differ somewhat; 2) the dynamic downscaling with two WRF experiments added value to the study by comparing two numerical schemes for estimating genesis locations; and 3) WRF with the New Tiedke and Kain–Fritsch schemes showed good skill in reproducing the spatial distribution of the most relevant dynamical parameters. The pattern correlations are well correlated with environmental parameters of untransformed GPI and higher correlations with binary logarithmic transformed GPI. The applicability to other cyclones is also tested (e.g., TC Nilofar in 2014) with encouraging results. This study demonstrates the usefulness of GPIs for forecasting TC genesis in the region.