| In this study, winter wheat as the object, combined with field experiment and indoor analysis,applied the theory and method of hyperspectral remote sensing. To explore the changes of winter wheat canopy spectra at different growth stages, analyze thecorrelation of the physiological and biochemical parameters of winter wheat and canopy spectral. Consturct vegetation index and spectral characteristic index, and find out its correlation with the physiological and biochemical parameters of winter wheat. To establishestimation model of chlorophyll content, leaf area index and nitrogen content. The study can provide timely and accurate winter wheat nitrogen nutrition diagnosis, to obtain accurate winter wheat growth information, to guide field management, and then promote the development of precision agriculture with providing data and technical support. The main conclusions are as follows:(1) From tillering stage to filling stage, with the growth and development of winter wheat, winter wheat leaf chlorophyll content and leaf area index increased; by filling stage into the milky stage, because of wheat plant nutrition all goes for reproduction growthat this time, the leaf structure changed, chlorophyll pigments are gradually disintegrated, leaf chlorophyll content and leaf area index continuously decreased. From tillering stage to milky stage, with the growth of winter wheat, dry matter continued to accumulate so that the nitrogen in the plant was consumed continuously, and the nitrogen of winter wheat was gradually reduced.(2) Numerous bands of winter wheat canopy spectra in all the growth stages have high correlation with leaf chlorophyll content and leaf area index: less than 700 nm in the original spectrum, the leaf chlorophyll content and leaf area index were negatively correlated and relatively stable;higher than 700 nm in the original spectrum, the leaf chlorophyll content and leaf area index has the trend of positive correlated with the spectrum, and reachedmaximum on the shoulders of the red edge, then basically remained stable. In the visible band, the green band of the original reflectance spectrum at different growth stages had stable negative correlation with plant nitrogen content, and the other bands were negatively correlated with plant nitrogen content, but the correlation was low. The derivative spectra of winter wheat canopy were negatively correlated with chlorophyll content, leaf area index and plant nitrogen content in the green band, and showed positive correlation in red wave band.(3) Using stepwise regression method,based on sensitive bands of first derivative spectral, to establish a multiple linear regression equation in every growth stage of estimate leaf chlorophyll content, leaf area index and plant nitrogen content;Analyze the leaf chlorophyll content, leaf area index and nitrogen content with kinds of vegetation indices and spectral characteristic indices in regression method, established estimation models of the leaf chlorophyll content, leaf area index and nitrogen content at every growth stages.Through test validation, analyze various estimation model fitting effect and prediction accuracy, to select the most general applicable and accurate estimation model of each growth stage. |
