Evaluation of satellite precipitation products across climatic and topographic gradients in a basin in Northern South America

Abstract

Reliably estimating precipitation remains a challenge, especially in tropical regions characterized by high climate variability, complex topography, and sparse rainfall station coverage. In these contexts, gridded precipitation satellite products (GPPs) are an important alternative for hydrological analysis, climate assessment, and risk management, which aligns with the Sustainable Development Goals, particularly SDG 6 (clean water and sanitation), SDG 11 (resilient cities and communities), and SDG 13 (climate action). This study evaluates the performance of five GPPs: CHIRPSv2, CHIRPSv3, ERA5-Land, MSWEPv2.8, and PERSIANNCDR, in the Ranchería river basin in northern Colombia. The evaluation covered the period 1991–2020 and used continuous and categorical metrics on daily, 3-day, monthly, and annual time scales. Fourteen rainfall stations from the national monitoring network were used as a reference. The study area is a complex hydroclimatic environment ranging from semi-arid lowlands to mountainous terrain influenced by the Sierra Nevada de Santa Marta (the world’s highest coastal mountain range). This configuration allows analysis of the sensitivity of satellite algorithms to topographic gradients and varying precipitation intensities. The results indicate that CHIRPSv3 has the best overall performance and the highest inter-scale consistency, followed by CHIRPSv2 and PERSIANNCDR. In general, all products improve with temporal aggregation; however, their performance degrades with increasing elevation, especially in ERA5-Land and MSWEPv2.8. The detection of daily extreme events remains limited, with moderate improvements when considering 3-day accumulations. As a general recommendation, satellite products should be selected based on local conditions, such as topography, time scale, and management objectives, to support reliable decision-making.