Monitoring Leaf Nitrogen Concentration in Wheat Based on Spectral Parameters from HJ-CCD Data with Stepwise Regression Method

Abstract

Nitrogen is a crucial parameter in maintaining crop health and predicting crop yield. It is possible that leaf nitrogen concentration (LNC) was quickly and non-destructively estimated by remote sensing method. The objective of this experiment was to develop a sensitive spectral index for monitoring LNC in wheat based on HJ-CCD data. In this study, we assessed several common spectral indices based on stepwise regression methods with the determination coefficient (R2) and root mean square error (RMSE). The results indicated that compared to other spectral index, the spectral index of 1.65SIPI-1.35PSRI was the most positively related to LNC (R2 = 0.6328, p<0.01), and the model for monitoring LNC based on 1.65SIPI-1.35PSRI was with the RMSE of 0.685 g.m-2 (p<0.01), and was significantly better than the models adopting other spectral indices. In conclusion, this study confirmed the feasibility of utilizing 1.65SIPI-1.35PSRI derived from airborne remotely sensed data to monitor LNC in wheat.