Accurate precipitation measurement is essential for climate, hydrological, and agronomic studies. However, in regions such as the Amazon, the scarcity of rain gauges and frequent gaps in historical series pose a significant challenge for long-term analyses. This study evaluated the performance of satellite and gridded precipitation estimates for gap-filling daily rainfall data recorded between 2019 and 2024. The observed dataset was obtained from a micrometeorological tower installed in an oil palm-based Agroforestry System (AFS) in the Eastern Amazon. The evaluation employed widely recognized statistical metrics, such as the coefficient of determination (R2), root mean square error (RMSE), mean absolute error (MAE), percent bias (PBIAS), Nash-Sutcliffe efficiency (NSE), and Willmott’s index of agreement (d). Additionally, cumulative precipitation curves from different databases were compared with the observed series to identify over- or underestimation trends. The results showed that, among the tested databases, NASA Power (NP) exhibited the best performance in terms of consistency and lower bias, making it the most suitable for filling gaps in the observed series. The analyses highlighted the importance of a careful selection of alternative databases to ensure data continuity and quality in remote tropical regions, an essential aspect for hydrological modeling studies.
Souza et al. (Mon,) studied this question.