Enhancing Daily Rainfall Data Completeness Using Satellite Rainfall Estimates

Abstract The availability of complete daily rainfall data is crucial for various hydrological, climatological, and meteorological studies. However, observed data often suffer from gaps due to equipment failures or station relocations. This study aims to improve the completeness of daily rainfall records by utilizing satellite rainfall estimates from TRMM, GPM-IMERG, and GSMaP products. To estimate the missing data, both linear regression and regional weighting methods were employed. A specific case study was conducted in the Kuranji watershed, Padang, Indonesia, using data spanning 2014 to 2016. The results indicate that the linear regression method based on satellite data, particularly GPM-IMERG, exhibits a higher correlation with observed data compared to the regional weighting method at most locations and across the study years. The results show that the regional weighting method was superior in 2014, while linear regression with GPM-IMERG and GSMaP was better in 2015 and 2016 in terms of accuracy, although regional weighting showed a lower relative bias. Nevertheless, the regional weighting method demonstrated a lower relative bias. This study highlights the potential of satellite-based rainfall data as a viable alternative for filling missing rainfall records. However, further calibration and validation are necessary to address spatial variability and inaccuracies during high-intensity rainfall events.