Validation of Satellite Precipitation Estimates over the Western Cape Region, South Africa

Abstract

Precipitation gauging networks have been declining in many developing countries such as South Africa. Satellite-based precipitation products (SPPs) and reanalysis data have shown vast improvements over the last 5 years and are postulated to mitigate the lack of ground-based observations. The Western Cape of South Africa has recently experienced a devastating drought (2015–18) and consecutive local floods (2023/24) but lacks sufficient precipitation observations to fully conceptualize the impacts of these climatic extremes. Using a variety of statistical metrics, we analyze the performance of six SPPs and reanalysis-driven datasets, namely, Climate Hazards Group Infrared Precipitation with Stations (CHIRPS V2), ERA5, Multi-Source Weighted-Ensemble Precipitation (MSWEP V2.8), and Integrated Multi-satellitE Retrievals for Global Precipitation Measurement (IMERG) (Early and Final Runs of V6B, Final Run of V7) compared with historical records collected from local weather bureaus from 2012 to 2021. On a daily scale, reanalysis products such as ERA5 and MSWEP outperform other SPPs concerning a dichotomous wet-/dry-day distinction, temporal rainfall pattern, and Heidke skill score. All products performed best in the wet parts of the year (JJA), and performance is influenced by climate zone and topography. All products performed poorly for extreme values (i.e., .95th percentile) and are therefore of less use for evaluating potential flood peaks. SPP performance was more related to climate type as opposed to altitude, and future improvements in SPPs should focus on local microclimatic variability to capture headwater precipitation patterns. While in the African tropics, infrared (IR)/passive microwave (PMW)-based products perform best, in the Mediterranean climate region of the Western Cape, reanalysis-driven products are superior.