Performance of GPM-IMERG precipitation products under diverse topographical features and multiple-intensity rainfall in an arid region

Abstract. The influence of topographical features and rainfall intensity on the accuracy of precipitation values estimated by earth observing satellites has attracted attention in the past decade. Assessment of rainfall products delivered by the Integrated Multi-satellitE Retrievals of Global precipitation measurement (IMERG) against ground observations has risen as an important endeavour since the accuracy of these products remain unreliable. This study comprehensively evaluated the three GPM IMERG products (near and post-real-time), over the period March 2014 to June 2018. The evaluation approaches were carried out for different seasons, rainfall intensities, topographical features, and hydrological regions over an extremely arid and semiarid country of Saudi Arabia. In general, the results confirmed that the performance of the final-run product surpassed the near-real-time products in terms of consistency and estimated errors. The evaluation results showed that for seasonal-based evaluation, the precipitation products exhibited better performance in spring and summer, while having relatively lower accuracy and higher biases in fall and winter. In addition, the results showed that the IMERG products had high performance in capturing the various rainfall intensities, with light rain having the highest accuracy. This is particularly important for arid regions as most of the rainfall is of the low-intensity class. Overall, the higher the rainfall intensity, the higher the detection errors in the IMERG products. Moreover, the hydrological evaluation results showed that the hydrological regions with low density of rain gauge stations hinders the proper evaluation of satellite products and tends to underestimate the performance of the products. Furthermore, the accuracy of the precipitation products was affected by topography to different extents. IMERG precipitation products exhibited high detection accuracy over moderate elevation areas (inland regions); whereas it had poor performance over flat plains (coastal regions) and high altitudes (foothills and mountainous regions). The outcomes of this evaluation could help developers in improving the GPM IMERG calibration to achieve better detection accuracy over arid and semiarid regions. More importantly, these results are of interest for local authorities to help manage development activities and to plan precautionary measures for extreme rainfall events.