Modeling and mapping reference evapotranspiration under limited climate data conditions in the Awash River Basin of Ethiopia

Reference evapotranspiration is essential for effective water resource planning and management. The Penman-Monteith (PM) method, though widely accepted, requires several meteorological variables that are often unavailable in data-scarce regions. This study aimed to develop and calibrate simplified temperature-based models to estimate ETo in the Awash basin, Ethiopia. Using multiple linear regression and optimization of the Hargreaves-Samani (HS) model through modifying both exponent and coefficients of the original model, new locally calibrated models were developed for the upper, middle, and lower parts of the basin. The calibrated coefficients were 0.0025, 0.0022, and 0.0055, respectively. Compared with the original HS model, the calibrated models substantially improved performance, achieving an average coefficient of determination (R 2 ) of 0.60, Nash-Sutcliffe efficiency (NSE) of 0.60, and index of agreement (d r ) of 0.67 for daily ETo. The root mean squared error (RMSE) was reduced by 17.6, 3.7, and 30.4% in the upper, middle, and lower regions, respectively. Spatial mapping of ETo indicated higher values in the lower and middle plains and lower values in the upper, eastern, and western highlands, reflecting topographic and climatic variability. Overall, the newly calibrated models provide reliable ETo estimation when only temperature data are available, supporting water resource planning in data-limited environments.