Estimation Precipitation Data Model with Observation Rainfall in Medan based on Validation of CHIRPS and GPCC Data

Abstract

Precipitation is Indonesia's dominant and most studied weather and climate factor. This study validated the CHIRPS and GPCC rainfall as global satellite estimation data with observation rainfall data at the Central Meteorology, Climatology, and Geophysics Region I Medan. Estimation data used is CHIRPS daily estimation data version 2.0 with a spatial resolution of 0.05° and GPCC with a spatial resolution of 1.0°, and validator data is daily observation data of Central MKG Region I Medan for the period 2017-2019. In this study, we calculated Pearson correlation, accuracy, bias, mean error (ME), mean absolute error (MAE), and root means square error 
(RMSE) values and used simple linear regression to see overestimated or underestimated estimation data on observation data. This study's results indicated that CHIRPS's performance was superior to GPCC with a higher correlation value, smaller MAE, and RMSE. Simple linear regression showed that CHIRPS and GPCC estimation was underestimated to the observed value. Overall, CHIRPS and GPCC cannot be used to forecast daily rainfall in Medan City.