Evaluation of Global Gridded Precipitation and Temperature Datasets against Gauged Observations over the Upper Tekeze River Basin, Ethiopia

The availability of satellite and reanalysis climate datasets and their applicability have been greatly promoted in hydro-climatic studies. However, such climatic products are still subject to considerable uncertainties and an evaluation of the products is necessary for applications in specific regions. This study aims to evaluate the reliability of nine gridded precipitation and temperature datasets against ground-based observations in the upper Tekeze River basin (UTB) of Ethiopia from 1982 to 2016. Precipitation, maximum temperature (T_max), minimum temperature (T_min), and mean temperature (T_mean) were evaluated at daily and monthly timescales. The results show that the best estimates of precipitation are from the EartH2Observe, WFDEI, and ERA-Interim reanalysis data Merged and Bias-corrected for the Inter-Sectoral Impact Model Intercomparison Project (EWEMBI), and the Climate Hazards Group InfraRed Precipitation with Station data (CHIRPS) datasets. The percentage biases and correlation coefficients (CCs) are within ± 15% and > 0.5, respectively, for both EWEMBI and CHIRPS at the two timescales. All products underestimate the drought conditions indicated by the standardized precipitation index (SPI), while the EWEMBI and CHIRPS datasets show higher agreement with the observations than other datasets. The T_mean estimates produced by the ECMWF Re-Analysis version 5 (ERA5) and the Climate Hazards Group InfraRed Temperature with Station data (CHIRTS) are the closest to the observations, with CCs of 0.65 and 0.55, respectively, at the daily timescale. The CHIRTS and EWEMBI datasets show better representations of T_max (T_min), with CCs of 0.69 (0.72) and 0.62 (0.68), respectively, at the monthly timescale. The temperature extremes are better captured by the ERA5 (T_mean), CHIRTS (T_max), and EWEMBI (T_min) datasets. The findings of this study provide useful information to select the most appropriate dataset for hydrometeorological studies in the UTB and could help to improve the regional representation of global datasets.