Using Ensembles of Machine Learning Techniques to Predict Reference Evapotranspiration (ET0) Using Limited Meteorological Data

To maximize crop production, reference evapotranspiration (ET0) measurement is crucial for managing water resources and planning needs. The FAO-PM56 method recommended globally estimating ET0 evaluating alternative methods due to its extensive theoretical foundation. Numerous meteorological parameters, needed estimation, are difficult obtain in developing countries. Therefore, ways estimate using fewer climatic data of critical importance. with methods, difference parameters temperatures, relative humidity (maximum minimum), sunshine hours, wind speed a period 20 years from 1996 2015 were used the study. recorded by 11 observatories situated various regions Pakistan. significance was evaluated sensitivity analysis. machine learning techniques single decision tree (SDT), boost (TB) forest (DTF) perform outcomes indicated that DTF-based models estimated higher accuracy variables as compared other ML DTF technique, Model 15 input, outperformed most part performance metrics (i.e., NSE = 0.93, R2 0.96 RMSE 0.48 mm/month). results mean humidity, minimum temperature could accurately techniques. Additionally, non-linear ensemble (NLE) further best input combination 15). It seen applied approach enhanced modelling stand-alone application (R2 Multan 0.97, Skardu 0.99, ISB 0.98, Bahawalpur 0.98 etc.). study affirmed use an model estimation suggest applying it parts world validate performance.