Study On Predicting Model Of Dry Matter Accumulation In Soybean By Leaf Reflectance Spectra

The accumulation of soybean shoot dry matter is an important indicator of soybean net photosynthesis, and the ultimate form of the material basis for economic output. For a long time, the conventional method on measuring dry matter in soybean was indoor testing species after sampling. The results were relatively accurate, but time-consuming and destructive. In recent years, spectral measurement was widely used in plant physiology, ecology and other aspects because of its simple, fast, non-destructive characteristics. There were many researches on hyperspectral remote sensing of the main crops at home and abroad, including study on physiological parameters of spectral monitoring in soybean. Some diagnostic models, such as leaf area, aboveground biomass, leaf nitrogen nutrition, were built by spectrometer built, but no report on dry weight.In this paper, on the basis of actual measuring the leaf reflectance spectra and dry matter accumulation in above-ground of soybean in experimental field, the sensitive wavelength of soybean shoot dry matter was established by analyzing spectra reflectance of near infrared light and visible light. The vegetation index was computed, and the prediction model on the shoot dry matter accumulation was established in soybean. The main results were as follows.1. Among the wavelength in visible light, there were highly significant correlations between spectral reflectivity of 510 and 680nm and the shoot dry matter accumulation of Henong60.2. Among the wavelength of near infrared light, there were highly significant difference between spectral reflectivity of 800nm and 900nm and shoot dry matter accumulation of Henong60, and significant difference between 1005nm and shoot dry matter accumulation.3. RVI showed the best relativity with shoot dry matter accumulation among those four vegetation indices. The corresponding prediction model established by vegetation indices of RVI was Y= 4.0216×RVI2 (900.680)-99.106×RVI((900,680)+625.36, which could be accurate to predict the shoot dry matter accumulation in soybean. The results provided the basis for the early estimates on soybean yield.