Key points are not available for this paper at this time.
The Upper Blue Nile Basin (UBNB) is a critical water resource, facing challenges from population growth, climate variability, and transboundary water management. For any type of water resources assessment, the availability and reliability of rainfall data is of fundamental importance. Global data sets, derived from satellite information and/or climate models and/or assimilated products have been applied in recent years more frequently, but their reliability for particular regions needs to be evaluated. This study therefore evaluates the performance of four rainfall datasets-CHIRPS, IMERG, ERA5, and ERA5-Land-against ground-based rainfall data from 75 stations across the UBNB from 2001 to 2019. Metrics including RMSE, MAE, ME, RE, R, and NSE were used to assess the datasets at daily and monthly scales. CHIRPS, ERA5, and IMERG demonstrated the best agreement with observed data at the daily scale, with mean RE values of 2.68, 3.14, and 4.56% respectively, compared to 66.06% for ERA5-Land. On the monthly scale, IMERG, CHIRPS, and ERA5 outperformed ERA5-Land, achieving NSE values of 0.83, 0.80, and 0.76, respectively, while ERA5-Land averaged −0.84. The annual rainfall averages reveal that CHIRPS, ERA5, and IMERG align closely with station measurements, whereas ERA5-Land consistently overestimates rainfall, with several areas in the UBNB showing average annual values exceeding 5,000 mm. Scatter plot analysis at the dekadal scale further demonstrates that CHIRPS, IMERG, and ERA5 achieve higher correlation coefficients compared to ERA5-Land, which shows a significant overestimation of heavy rainfall events. Using the WASA-SED model to simulate the streamflow of the UBNB. CHIRPS and IMERG yielded NSE values of 0.81 and 0.80 at daily scales during validation, confirming their suitability for hydrological modeling. Conversely, ERA5-Land exhibited poor performance due to substantial rainfall overestimation. IMERG particularly excelled with daily RE values as low as −0.94%, making it highly reliable for water resource management. This study highlights the potential of CHIRPS, IMERG, and ERA5 for hydrological applications in complex, data-scarce regions like the UBNB while emphasizing the need for improved rainfall representation in ERA5-Land. These findings support informed decision-making for sustainable water management and reinforce the applicability of evaluated accordingly adjusted satellite-derived rainfall data for streamflow prediction in data-scarce tropical regions.
Zargar et al. (Mon,) studied this question.