Meteorological Data Fusion Approach for Modeling Crop Water Productivity Based on Ensemble Machine Learning

Crop water productivity modeling is an increasingly popular rapid decision making tool to optimize resource management in agriculture for the makers. This work aimed model, predict, and simulate crop (CWP) grain yields of both wheat maize. Climate datasets were collected over period from 1969 2019, including: mean temperature (Tmean), maximum (Tmax), minimum (Tmin), relative humidity (H), solar radiation (SR), sunshine hours (Ssh), wind speed (WS), day length (DL). Five machine learning (ML) methods applied, including random forest (RF), support vector regression (SVM), bagged trees (BT), boosted (BoT), matern 5/2 Gaussian process (MG). Models implemented by MG, Tmean, SR, WS, DL (Model 3); Tmax, Tmin, Ssh, H, 8); SR 9), found optimal (r2 = 0.85) forecasting CWP wheat. Moreover, results maize showed that BT a combination Tmin data, achieved high correlation coefficient 0.82 compared others. The outcomes demonstrated several performance ML-based alternative estimation case limited climatic data supporting designers, developers, managers resources.