A Unified Physically Based Method for Monitoring Grassland Nitrogen Concentration with Landsat 7, Landsat 8, and Sentinel-2 Satellite Data

The increasing number of satellite missions provides vast opportunities for continuous vegetation monitoring, crucial precision agriculture and environmental sustainability. However, accurately estimating traits, such as nitrogen concentration (N%), from Landsat 7 (L7), 8 (L8), Sentinel-2 (S2) data is challenging due to the diverse sensor configurations complex atmospheric interactions. To address these limitations, we developed a unified physically based method that combines soil–plant–atmosphere radiative transfer (SPART) model with bottom-of-atmosphere (BOA) spectral bidirectional reflectance distribution function. This approach enables us assess effect rugged terrain, viewing angles, illumination geometry on multiple sensors. Our methodology involves inverting variables using numerical optimization estimate N% creating hybrid model. We used Gaussian process regression (GPR) incorporate inverted into prediction, resulting in estimation across different shows validation accuracy 0.35 (RMSE %N), mean prediction interval width (MPIW) 0.35, an R2 0.50, independent sensors collected between 2016 2019. promising solution L7, L8, S2 data, overcoming limitations posed by