Intra-Plot Variable N Fertilization in Winter Wheat through Machine Learning and Farmer Knowledge

The variable fertilization rate (VFR) technique has demonstrated its ability to reduce nutrient losses by adapting the fertilizer dose crop needs. However, transferring this technology farms is not easy. This study aimed make a map in commercial plot where there no data from yield monitor, combining machine learning techniques and farmer’s knowledge. In addition normalized difference vegetation index (NDVI) obtained Sentinel-2 digital elevation model (DEM), information captured monitor 2019 was used train validate models. Among 15 algorithms trained, best result random forest (RF), with an RMSE of 496 R2 0.90. Using “leave one out” technique, capacity predict entire tested. Finally, RF algorithm tested on 12-hectare wheat were available. novelty work lies collaborative developed between farmers researchers implement VRF plots precise do exist validation. collaboration scientists resulted very positive exchange that allowed farmer change strategy whole farm better understand how soil properties history affect yield.