Predictive Modelling of Reference Evapotranspiration Using Machine Learning Models Coupled with Grey Wolf Optimizer

Mismanagement of fresh water is a primary concern that negatively impacts agricultural productivity. Judicious use in agriculture possible by estimating the optimal requirement. The present practice crop requirements using reference evapotranspiration (ET0) values, which considered standard method. Hence, predicting ET0 vital allocating and managing available resources. In this study, different machine learning (ML) algorithms, namely random forests (RF), extreme gradient boosting (XGB), light (LGB), were optimized naturally inspired grey wolf optimizer (GWO) viz. GWORF, GWOXGB, GWOLGB. daily meteorological data 10 locations falling under humid sub-humid regions India for cross-validation stages employed, eighteen input scenarios. Besides, empirical models also compared with ML models. hybrid found superior accurately at all stations than conventional reduction root mean square error (RMSE) from 0.919 to 0.812 mm/day region 1.253 1.154 was seen least accurate model hyperparameter tuning. RF have improved their accuracies substantially GWO LGB XGB