Extraction of Photosynthetic Parameters of Cotton Leaves under Disease Stress by Hyperspectral Remote Sensing
Chen Bing,Wang Gang,Liu Jing-de,Ma Zhan-hong,Wang Jing,Li Tian-nan
DOI: https://doi.org/10.3964/j.issn.1000-0593(2018)06-1834-05
2018-01-01
Spectroscopy and spectral analysis
Abstract:Hyperspectra remote sensing technique was applied to detect photosynthetic parameters (PP) in cotton leaf defected Verticillium wilt. The reflectance data about 207 was acquired in 350 similar to 2 500 nm bands in different dates and severity level of cotton leaves, and PP were measured by photosynthetic instrument. Analysis of variance and relationship analysis were used to process the PP character, extra spectral sensitive bands and selected character parameters of spectra with PP linear and nor-linear models were applied to product PP of cotton leaf of disease and tested. The result showed: with the disease condition increase, the data was increased to leaf net photosynthetic rate (A), transpiration rate (E), stomatal conductance (GH(2)O), however intercellular CO2 (CI) firstly decreased and then up, and the difference was significant between severity levels and PP. The relationship became better between severity level and PP, which r were -0. 97, -0. 957, -0. 886, -0. 715, respectively. New spectral parameters, R704, R706, R699, R690, FD688, FD732, FD690, FD731, FD681 were built on the base of sensitive bands together with tradition spectral parameters to established revise models of A, E, GH(2)O, CI of cotton leaves with disease. Those models as PRI[FD732, FD688]), R706, RVI[890, 670]), R690 for the independent variable had the highest accuracy to estimate A, E, GH(2)O, CI, R-2 of prediction, which were 0. 827, 0. 810, 0. 658 and 0. 573 respectively; RMSE were 5. 466, 2. 801, 109. 500 and 63. 500 respectively; RE were 0. 041, 0. 137, 0. 158 and 0. 021 respectively, which can realize the inversion of photosynthetic physiological parameters of cotton by remote sensing.