This study examines how changing climate and LULC in southern India's semi-arid Chitravathi basin affect evapotranspiration (ET) and potential evapotranspiration (PET) in the near future (NF) (2015–2030). The future LULC map is produced using cellular automata and artificial neural networks (CA-ANN). ET and PET were estimated using the hydrological model SWAT (Soil and Water Assessment Tool). This study used the meteorological data (precipitation and temperatures) with a resolution of 0.25° × 0.25° from ten gauging stations across the basin. Based on the Coupled Model Intercomparison Project 6 (CMIP6), Global Climate Model (GCM) MPI-ESM1-2-LR, the NF climate under the socioeconomic pathway SSP2-4.5 is projected. The SWAT model is calibrated for discharge data at a gauging station. SUFI-2 algorithm was used in the SWAT-CUP for the automation calibration of the SWAT model. Statistical measures were used to assess model performance, including coefficient of determination (R2) and Nash Sutcliffe efficiency (NSE). The model demonstrates an overall accuracy with R2 = 0.83 and NSE = 0.81. The estimated future parameters were compared with the baseline (1985–2014) parameters. Results revealed that ET is 332 mm during the NF, and PET is 1683 mm. Compared to baseline values, 13.5% decreased ET and 37.5% increased PET were observed. This research will help design and implement adaptation measures to mitigate the effects of land cover and climate change on Chitravathi basin's water resources.