Evaluasi Keakuratan Produk Satelit (CHRIPS, GPM-IMERG, PERSIANN-CCS, dan GSMaP) Dalam Mengidentifikasi Curah Hujan di Sumatera Barat

This study evaluates the accuracy of satellite rainfall products against observational data from three stations in West Sumatra, averaged to represent the region's general conditions. The analysis uses eyeball and statistical methods on daily, ten-day, monthly, and seasonal scales. The satellite products assessed include GPM-IMERG, CHIRPS, GSMaP, and PERSIANN-CCS. On the daily scale, CHIRPS shows the best accuracy with a correlation coefficient of 0.17, RMSE of 27.97 mm, NRMSE of 0.097, and MAE of 14.19 mm. On the ten-day, monthly, and seasonal scales, GPM-IMERG outperforms the other products, achieving the highest correlation coefficient of 0.84. Additional analysis using contingency table metrics (POD, FAR, CSI, FOH, HSS) confirms GPM-IMERG as the most accurate product. These findings indicate that GPM-IMERG is the most reliable satellite product for rainfall estimation in West Sumatra, especially for longer temporal analyses.