Evaluation of the WRF‐ARW Model Physical Parameterization Schemes for Heavy Flood Events Over Bushehr‐Province of Iran

Abstract

This study evaluates the performance of different WRF model configurations in simulating all 19 heavy precipitation events over Bushehr Province, Iran, over two decades (2000–2019). For the first time, the sensitivity of combinations of three physical parameterization options—seven cumulus (Cu) schemes, three planetary boundary layer (PBL) schemes, and three microphysics (MP) schemes—has been analyzed both individually and seasonally over a long‐term period. A total of 209 simulations, each with a 24‐hr lead time, were conducted using 11 configuration sets. The Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) in the Multi‐Criteria Decision Making (MCDM) method with eight performance indicators was applied. Results revealed that Cu schemes have the largest influence on precipitation simulations compared with MP and PBL schemes. Among them, Kain‐Fritsch, Grell‐Freitas, and Grell‐Devenyi schemes produced the most accurate precipitation, respectively, while KIAPS and BMJ performed the worst, highlighting the dominant role of the Cu scheme in simulating extreme precipitation events in the region. Ensemble‐mean analysis of atmospheric variables, including CAPE, vertical velocity, specific humidity, boundary layer height, and temperature, further indicated that seasonal and multi‐scale forcing strongly affect convective development and model performance. These findings, emphasizing the seasonal characteristics of extreme precipitation events, provide new insights into the physical processes underlying flood dynamics in Bushehr Province and underscore the importance of selecting appropriate parameterizations for reliable regional flood prediction. We expect that this study will serve as a conceptual guideline for selecting suitable configurations for realistic operational precipitation forecasts in Bushehr Province.