Performance of high-resolution gridded precipitation products for determining long period return values and Standardized Precipitation Index in Colombia

Limited coverage of precipitation stations and short observational records hinder hydrological studies. Consequently, satellite precipitation data is a viable alternative for estimating rainfall amounts in regions that lack monitoring networks. However, assessing their representativeness is crucial to ensure adequate precision and accuracy in studies derived from these types of datasets. The present study assessed the representativeness of the CHIRPS v2 and MSWX precipitation datasets, along with their mean (MEAN), for daily precipitation frequency analysis and estimation of precipitation anomalies in Colombia. The performance was assessed by comparing the frequency analysis and precipitation anomalies obtained from the datasets with those derived from observational information. A total of 987 rainfall stations with daily precipitation data available between 1980 and 2022 were used. The results indicate a tendency to underestimate the daily intensities for the different return times evaluated and a poor performance in reproducing precipitation anomalies using datasets evaluated throughout Colombia.