Assessment of Satellite-based Precipitation Products Performance over the Hyper-arid Climate of Kuwait

Abstract

AbstractPrecipitation is a complex natural parameter that is essential for water and environmental systems. Due to its variability on the spatial and temporal scales, satellite-based precipitation products (SPPs) have arisen interest in hydrology and meteorology applications. This study measures the performance of six high resolutions SPPs (GPM IMERG products (IMERG-E, IMERG-L, and IMERG-F), TMPA products (3B42 V7, 3B42RT V7), and PERSIANN product) in producing the observed precipitation over a hyper-arid climate, water scarce region for the period 2013-2018. It also evaluates their performance dependency on the aggregation time-step and topographic elevations. According to a number of continuous and categorical evaluation metrics: (a) SPPs overestimate the observed daily annual and seasonal precipitation, particularly with near real-time products, (b) all SPPs estimates depict correlation ranging from 0.68 to 0.84 with the annual and seasonal precipitation and weak correlations in dry season, and (c) their ability to detect rain/no-rain events is measured by Peirce Skill Score (PSS), ranging from 0.73 to 0.92 across annual and seasonal scales, whereas 3B42RT V7 reproduces lower PSSs. Furthermore, the study finds that aggregation to a monthly time-step improves only SPPs correlations. The performance of near real-time products shows significant dependency on elevations, especially with 3B42RT V7 that shows low skills at coastlands. The TMPA products ability to detect rain/no-rain events dramatically drops from highlands to coastlands, with low skills to generate observed no/tiny and light precipitation classes. The study addresses an adequate ability of IMERG-F and PERSIANN to be utilized in water and environmental studies over hyper-arid climate regions, with highlighting for the superiority of IMERG-F.