Research On Remote Sensing Monitoring Of Sowing Time Based On Preliminary Spectral Information Of Winter Wheat

The sowing time informations of winter wheat is important to guide the classification management, the space data stable obtain of high spatial and high temporal resolution for early sowing of winter wheat offers the possibility. This study research on data mining and choose time from massive amounts of remote sensing data, and how to use the spectrum or vegetation index construction sowing monitor model and analyze it. the main conclusions are follows:Firstly, the imitation of the reflectivity of canopy spectrum by way of coupling model. The author chooses the coupling model combined with CERES-Wheat and PROSAIL, uses the experimental data and parameters based on the data from Xiaotangshan by way of CERES-Wheat model which includes the experimental data of LAI of winter wheat in 2006-2007 and takes advantage of the imitation of new meteorological and soil parameters to get corresponding LAI which regards other parameters as those of the coupling model in order to imitate and analyze the differences in mechanism of canopy spectral response of winter wheat and variations of rules. The maximal mechanism of the differences ofspectral response can choose some sensitive bands which are the basis of the choice of vegetation index.Secondly, the optimal phase for the choice of sowing monitoring use the data of the model. Getting different reflectivity of growth stage of winter wheat in different sowing time to analyze their rules by imitating physical model and choosing the most sensitive wave by way of taking advantage of its separability in data of seeding. First of all, preliminarily evaluate separability of sowing timein different times by accounting the distance of J-M in order to choose some good phases. Then, analyze the sowing time by way of discriminant analysis to judge the category of the unknown sample. Choosing the optimal phase of sowing monitoring according to the accurate category of precision to choose the biggest difference of phase in different sowing time-the best phase of seperability.Thirdly, the choice of vegetation index and the foundation of the model of sowing monitoring.Taking advantage of different spectrum of seeding time which has the biggest sensitive wave to build a variety of vegetation index-RVI, NDVI, ARVI, EVI, SAVI and MSAVI etc. and making full use of the relativity analysis between vegetation index and sowing time of winter wheat to build relative monitoring model of winter wheat. Finally, comparing and taking advantage of the relative coefficient build from six vegetations to get the maximal coefficient which is regarded as the optimal model for sowing monitoring of the winter wheat and used as the guidance for agricultural production.