Accurate estimation of reference evapotranspiration (ETo) is crucial for irrigation scheduling and regional water resources management. The FAO-56 Penman-Monteith or the CIMIS Penman is recommended as the reference model to predict ETo but is often limited by high cost and lack of complete weather data in many regions. Air temperature data is broadly available, and its use for ETo estimations is widespread.
This study evaluated eight temperature-based empirical reference evapotranspiration models and eight machine learning models against the standard CIMIS ETo. The performance indices used for evaluation include root mean squared error (RMSE), mean absolute error (MAE), mean biased error (MBE), and determination coefficient (R2)