Strategic planning for water resources management in the face of climate uncertainties requires accurate prediction of changes in evapotranspiration processes. The present study forecasts changes in reference evapotranspiration in Fars province under common socio-economic path (SSPs) scenarios and CanESM5 climate model outputs. To this end, daily minimum temperature, maximum temperature, and precipitation data from 1995 to 2025 were obtained from three stations: Abadeh, Shiraz, and Lar in Fars Province. Subsequently, the LARS-WG model was employed to downscale temperature and precipitation for the 2025–2045 horizon after rigorous validation. The climate model outputs were then used as inputs for the Hargreaves–Samani empirical model to calculate reference evapotranspiration. The results demonstrated that the LARS-WG model simulated minimum and maximum temperatures with high accuracy (R2 > 0.85) and RMSE < 0.4, while exhibiting moderate accuracy in precipitation simulation (0.62<R2<0.74). The findings indicate an upward warming trend at all stations, with the intensification of scenarios from SSP126 to SSP585, with the temperature increase in some regions and seasons reaching more than 2.4 °C. While the temperature in all scenarios and stations has an increasing trend and is in line with the intensity of the emission scenarios, precipitation faces severe seasonal fluctuations and heterogeneous behavior (including decreases in warm seasons and fluctuations in cold seasons). The findings of this study indicate that the increase in the reference evapotranspiration trend indicates an intensification of the evaporative demand of the atmosphere in Fars Province. This change, as one of the main driving factors, can increase the potential for water stress and water resource constraints in the future.