Mohammed A. El-Shirbeny
Accurate and usable potential evapotranspiration (PET) estimation is important for managing water resources around the world, planning agriculture, and adapting to climate change. Complex energy balance models yield valuable insights; yet practical applications necessitate straightforward, resilient, and simply implementable long-term monitoring methodologies. This study confirms a more straightforward empirical model that estimates monthly PET only utilizing MODIS land surface temperature (LST) data of 25 years (2000–2024), addressing a deficiency in the comprehension of simple model transferability across global climatic regimes. The LST products (MOD11A1/MYD11A1) processed in Google Earth Engine to confirm the accuracy of PET predictions against the FAO-56 Penman–Monteith (FAO-PM) technique, which was based on data from 58 ground-based meteorological stations in 5 Major Köppen–Geiger climate zones. The model was very accurate (R² = 0.76, RMSE = 30.02 mm/month); however, it was completely unique in different areas because of environmental controls. The model worked well in the Continental and Mediterranean climate zones (R² = 0.93, NSE = 0.88), but it had trouble in the Tropical Wet (R² = 0.39, NSE = -6.15) and Polar (R² = 0.64, NSE = -2.64) regions because of the moisture in the air and the complicated way energy is divided. The initial comprehensive analysis of basic LST-based model constraints sets essential standards for operational implementation and underscores the necessity for climate-zone-specific parameterization in this global, long-term validation. The results enhance the comprehension of environmental influences on remote sensing-derived PET estimation and inform water resource management in a dynamic climate.