Predictive Modelling of Maize Yield Using Sentinel 2 NDVI

Accurate yield prediction is essential for precision agriculture as it enables farmers to optimize their inputs and manage their resources more efficiently, ultimately leading to higher profitability and sustainable farming practices. The aim of this research is to verify whether satellite images can be a valid tool for predicting yield in summer maize (Zea mays L.). The study was conducted in 2022 in two adjacent fields in Umbria, Italy. The approach adopted is based on the use of NDVI (Normalized Difference Vegetation Index) data from Sentinel 2 (S2) satellites and yield data from combine harvesters to estimate the grain yield of maize. The NDVI data, derived from S2 images acquired at different growth cycle stages, were used for a Principal Component Analysis (PCA). Subsequently, a multiple linear regression (MLR) model was created. The results show a significant correlation between NDVI and grain yield, suggesting that the central stages of the maize phenological cycle are more indicative for this purpose. Further experiments are necessary to confirm the effectiveness of the proposed approach, using more specific vegetation indices or including soil and climate data in the procedure, thus obtaining more comprehensive responses.