A Research On Winter Wheat Growth Monitoring Indicators With Remote Sensing

The crop growth monitoring system using remote sensing technology has some shortcomings. For instance, the system use single remote sensing index, and the remote sensing index cannot quantitatively reflect crop growth condition. These problems limit the accuracy of monitoring. Analyzing of those factors of the quantitative relationship between remote sensing Indicators and growth parameters and using mathematical or physical models to quantify the growth parameters is very important for improving the accuracy of crop growth monitoring. In this paper, remote sensing indicators of wheat growth condition has been study. The research coverage is as follows:1. Using correlation analysis, the author got Correlation coefficients between remote sensing indicators and growth parameters in each stage. Analyzing of the correlation coefficients canopy remote sensing indicators and LAI, chlorophyll, yield respectively, the author studied the factors that affect their relationship.2. Using ground survey data and the corresponding HJ-1 multi-spectral remote sensing images, the author studied and verified the general method of retrieving LAI of winter wheat in different periods.3. By two up-scaled methods, the author got 240m, 480m and 960m resolution vegetation indices and LAI from the 30m resolution multi-spectral surface reflectance. Coefficients between vegetation indices of different pixel scale and LAI respective were got by correlation analysis in different periods.The main results are as follows:1. As LAI increasing,the effect of soil background decrease,and the value of correlation coefficient between remote sensing indicators and LAI increase. Soil-adjust remote sensing indices using different algorithm to reduce effect of soil background have limitation respective. Remote sensing indices can not reflect leaf chlorophyll content well as LAI increasing.2. HJ-1 multi-spectral remote sensing images have atmospheric effects. The Spectral feature curve had been change well after the atmospheric correction by FLAASH. The data after the atmospheric correction was used to calculate NDVI, and then NDVI was used to retrieve LAI of winter wheat in different stages by a nonlinear regression model. The accuracy of inversion LAI method was effect by error of the measured values.3. The relationship of the vegetation indices of Pixel scale and LAI depend on standard deviation of LAI in the pixel and the pixel scale. Within the same scale as per standard deviation of LAI higher, the correlation coefficient of vegetation index and LAI lower. With the larger pixel scale, standard deviation of LAI increased, but the correlation coefficients of vegetation index and LAI are not reduced.The research shows those factors such as soil background, atmospheric effect, scale effect all effect the quantitative relationship between remote sensing indicators and winter wheat growth parameters, and different factors have different ways. In this study, correlation and regression analysis reflect the qualitative semi-quantitative relationship of remote sensing indicators and the growing parameters. Future research would focus on the quantitative models of remote sensing indicators and Growth Parameters by mathematical or physical models, and quantitatively eliminate the effect of soil, air and scale according to different factors.