Evaporation is a significant process in the water cycle and poses a considerable challenge to water reserves globally, particularly in arid countries. This study aims to address the quantification of evaporation, which is crucial for effective water resource management, lake water balance studies, and hydrological cycle prediction. To achieve this, we propose a multiple linear regression (MLR) model that combines two models and relies on three key parameters: solar radiation, average air temperature, and daylight. The performance of the model is assessed using six statistical indexes: Nash–Sutcliffe efficiency (0.99), root mean square error (11.57), mean absolute error (9.75), coefficient of determination (0.98), root mean square ratio (0.14), and Willmott index (0.995). The model demonstrates promising accuracy, with an annual error of evaporation rate equal to 0% (Ep = 2574.44 mm; EMLR = 2582.70 mm). Monthly errors of evaporation range from −8% (Ep = 105.33 mm; EMLR = 124.30 mm) to 9% (Ep = 164.86 mm; EMLR = 150.10 mm). The multi-linear regression model (MLR) and evaporation pan (Ep) numerical results highlight their comparative performance across the months. Based on our numerical and statistical results, we conclude that the proposed empirical model can accurately estimate the evaporation rate from the Djorf-Torba reservoir dam in the arid region.