New vegetation indices for satellite monitoring of the nitrogen nutrient index in corn

LAPAZ, ADRIÁN; CARCIOCHI, WALTER D.; SAINZ ROZAS, HERNAN R.; BALZARINI, M.; CASTRO-FRANCO, M.; TOVAR HERNÁNDEZ, SERGIO; ÁVILA, OSCAR; LARREA, GASTÓN; MARIA PIA RODRIGUEZ; REUSSI CALVO, NAHUEL

Congreso; XXVIII Congreso Argentino de la ciencia del suelo; 2022

The nitrogen (N) nutrition index (NNI) is a reliable indicator of in-season N status in corn (Zea mays L.). However, NNI determination is expensive and time-consuming. Therefore, integrating information from satellite remote sensing tools with soil data could estimate the Spatio-temporal variation of NNI. The goal was to develop vegetation indices by integrating soil and satellite information to predict NNI. Eleven field experiments in corn were conducted in the Argentine Pampas, applying five N rates (0, 60, 120, 180, and 240 kg N ha-1) at sowing. Available N (Nav) was the sum of the inorganic N in the soil at sowing (0-60 cm) plus the N contained in the fertilizer (N rate). Plants collected at stages sixth, tenth, and fourteenth developed leaves and flowering of corn (V6, V10, V14, and R1, respectively) were used to determine the NNI. Sentinel-1 and Sentinel-2 satellite observations were recorded in the areas of the collected plants for each sampling date. With these observations, the vegetation indices were calculated by integrating Nav, spectral bands (Sentinel-2) and C-band backscatter from synthetic aperture radar (C-SAR, Sentinel-1). Simple regression models relating NNI to vegetation indices were calibrated and validated. During calibration, the coefficient of determination (R2) ranged from 0.61 to 0.78 for different models. During validation, the NNI prediction had high accuracy, with the mean absolute percentage error (MAPE) ranging from 6 to 12%. Local incidence angle corrected C-SAR backscatter and reflectance in the red spectrum were essential for calculating the three vegetation indices that were selected. In summary, it was possible to predict NNI in corn by integrating satellite platforms and pre-plant soil information, which then would facilitate N monitoring during the corn growing season.