Hyperspectral remote sensing for precision nitrogen management in cereal crop production

Efficient nitrogen fertilizer use in cereal production is central to closing yield gaps while limiting environmental losses, yet farmers still rely largely on uniform applications that ignore within-field variability. This study evaluates hyperspectral remote sensing as a basis for precision nitrogen management in winter cereal systems. Field experiments combined highresolution hyperspectral imagery, intensive plant and soil sampling and a calibrated crop growth model to derive nitrogen status indicators and variable-rate fertilizer recommendations. A set of nitrogen-sensitive spectral indices and machinelearning inversion models was developed and tested across contrasting soil and climate conditions. Performance was assessed in terms of prediction error for plant nitrogen concentration, nitrogen uptake and yield, together with economic return and residual soil nitrate after harvest. Results show that hyperspectral-driven management zones and variable-rate nitrogen application significantly improve nitrogen use efficiency and gross margins, while reducing residual nitrogen, demonstrating the practical value of hyperspectral sensing in cereal nitrogen management.

Перевод пока недоступен