<![CDATA[The availability of accurate high-resolution rainfall data is crucial for water resources management and hydrometeorological disaster mitigation in the Upper Citanduy Watershed, yet constrained by sparse rain gauge networks. This study evaluates satellite precipitation products (GPM IMERG and CHIRPS) and reanalysis datasets (ERA5 and NASA POWER) against 12 station observations during 2020–2023 using multi-scale evaluation encompassing quantitative, categorical, and volumetric metrics at point and regional scales. Results show station elevation (200–1000 masl) has no significant effect (p > 0.05) on estimation accuracy; performance variability is predominantly determined by algorithm characteristics and spatial resolution. GPM IMERG demonstrates superior performance with Nash-Sutcliffe Efficiency of 0.279, highest correlation (r = 0.674), and lowest RMSE (8.914 mm/day) at watershed scale, excelling in volumetric metrics with VCSI of 0.55–0.77 and low VFAR (0.10–0.30). Conversely, ERA5 and NASA POWER exhibit extreme positive bias from coarse resolution triggering spatial smoothing effects and high FAR (0.25–0.71). CHIRPS shows moderate capability in replicating spatial patterns but overestimates cumulative volumes. GPM microwave sensors prove more effective in capturing tropical convective rainfall compared to infrared estimation and model-based reanalysis. GPM IMERG is recommended as primary data source for hydrological modeling in the Upper Citanduy Watershed, while reanalysis products require bias correction before operational application.]]>
