Study on the biomass change derived from the hyperspectral data of cotton leaves in canopy under moisture stress

Abstract

In this study, an ASD spectrometer made in USA is used to derive the hyperspectral data of cotton leaves in canopy in north Xinjiang. The integral area variables of red edge are used to estimate the total nitrogen (TN) content in cotton leaves in canopy. The first order differential of reflective spectra is carried out. The analysis methods based on the spectral position variables are derived from the first order differential spectral data. The analysis on the correlation between the integral area of red edge (SDr, regarded as the independent variable) and the TN content (regarded as the function) is carried out, and the correlation mathematical model between the integral area of red edge and the TN content in canopy leaves of cotton varieties named as Xinluzao No.6 is developed. The analysis on the correlation between the chlorophyll content and the TN content in single cotton leaves in canopy growing under the irrigation with different water volumes is carried out. The results show that there is a significant positive correlation between the chlorophyll content and the TN content (R = 0.8723, n = 39), and the data of chlorophyll content can be used to estimate the TN content in single cotton leaves; the correlation between the integral area variable of red edge and the TN content in cotton leaves in canopy is significant, their correlation coefficient is 0.7394 (n = 40), the TN contents in canopy leaves of cotton varieties named as Xinluzao No. 6 and No. 8 can be accurately estimated by using the developed model, and their RMSE values are 0.3859 and 0.4272 respectively. After researching, it is considered that there is an applying potential to use the integral area variables of red edge for estimating the TN content in cotton leaves in canopy, and the mathematical model developed with the third-side area variables has a high applying value in deriving the TN content in crop canopy. It is also considered that it is feasible to recognize the moisture stress suffered by cotton plants by researching the shift and change extent of red edge, and the key is to develop the corresponding rational recognition index system.