Evaluation of Chirps Tropical Rainfall Estimates Using Automatic Weather Stations for Validation in West Sumatra, Indonesia

Abstract The availability of accurate and evenly distributed rainfall data is still a significant challenge in managing water resources in West Sumatra. One of the efforts to fulfill data needs is using Automatic Weather Stations (AWS) to record data automatically. However, AWS has limitations, such as high operational costs, limited spatial coverage, and susceptibility to technical disturbances. As an alternative, Climate Hazards Group InfraRed Precipitation with Station Data (CHIRPS) satellite products offer broad spatial coverage and high temporal resolution, but still require validation against observational data. This study aims to evaluate the accuracy of CHIRPS daily rainfall data by comparing it to AWS Ambient Weather WS-2902 data installed at the Faculty of Engineering, Andalas University. The analysis was conducted from September 2023 to March 2025 using a point-to-pixel approach and statistical parameters such as Correlation Coefficient (CC), Standard Deviation, and Centered Root Mean Square Difference (cRMSD). Bias correction was applied through three methods: Linear Scaling (LS), Local Intensity Scaling (LOCI), and Empirical Quantile Mapping (EQM). Results show that CHIRPS underestimates high-intensity rainfall, but is quite accurate for light to moderate rainfall. This study can significantly improve the accuracy of bias correction using the LOCI method with CC = 0.80 and cRMSD = 30.39 mm. The detection performance evaluation shows good performance (POD = 0.90; FAR = 0.10; CSI = 0.80. These findings support using corrected CHIRPS as an alternative rainfall data source in observation-scarce areas, though uncertainties of precipitation data remain in extreme events.