The model for grain wheat yield prediction at high spatial resolution based on physical-geographical properties and satellite vegetation indices

Precision agriculture is promising approach for improving agricul-tural production, especially nowadays when the population is rap-idly increasing. For that, crop yield estimation provides valuableinformation. The main research focus was to predict within-fieldgrain yield and detect its drivers. The Random Forest regressionmodel on data from diverse sources at the 10-meter spatial reso-lution was developed. The study was conducted in the Vojvodinaregion (Serbia) for eight wheat-planted fields, having precise grainyield data. Open-source data including 15 vegetation indices (VIs)was calculated from Sentinel-2 satellite bands, physical-geograph-ical features obtained from the digital elevation model and soilproperties. The model succeeded in predicting the wheat grainyield with the RMSE of 0.66 t/ha (average yield of 0.09 t/ha) andthe best predictors were VIs considering chlorophyll and moisturecontent in plants, while physical-geographical properties man-aged to explain within-field variability. This methodology can beapplied to other crops (maize, soybean).