Study On Spectral Feature Of Soil Water And Rice Nitrogen Status

Soil water content is important for studying soil humidity, irrigation management, agricultural ecological planning, evaluation of crop growth potential and simulation of global land cover change affected by climate changes. Soil water content is one of the important parameters of indicating soil moisture, decision of the time of spring ploughing, compensation irrigation management and yield evaluation. The survey of soil water by experiments is expensive and time-consuming. Moreover, it is probably unfeasible with the extension of research area and spatial changes of soil.It is easy and feasible to construct soil water content evaluation models by soil spectral experiments on the conditions of different water contents. In this study, the sensitive bands to disdingusing soil moisture were in 760-1220nm and the model equations between soil spectral reflectance and soil water content were built. It is of great importance to evaluate soil water content , to appraise the duration of drought of soil and instruct agricultural irrigation. It is possible to transform from calculating soil water content by ground spectral reflectance to calculating soil water content by remote sensing images because of the similarity between ground spectrum and satellite remote sensing. This paper discussed the possibility of the transformation from calculating soil water content by ground spectral reflectance to calculating soil water content by remote sensing images. Furthermore , this paper made some research on the feasibility of monitoring soil water content of yellow brown region using TM remote sensing images.Rice is the most important staple crop in China and is of utmost importance to China's grain safety. But there is not a simple, quick and satisfying method to test the level of nitrogen in rice. Untill now Chemical analysis method is the major method to test the level of nitrogen in rice. While Chemical analysis method exists such shortcomings as time-consuming, fussy and destructive. It is hard to instruct fertilizering when Chemical analysis method is used to test the level of nitrogen in rice in paddy. Though the fast survey in the field is quick , the precision is not high.The spectral reflectance of plant has obvious features. Different plants , the same plant of different growth stages, plants of normal growth, and plants invaded by disease or insect ,have different spectral reflectance. In addition, the difference of irrigation and fertilizering can cause the change of the spetral reflectance of plants. With the deepening ofremote sensing application in agriculture, people have found that the spectral ferture of plants is affected by its physiological characteristic which decided the absorption, transmission and reflectance of light, and the physiological characteristic of plants reflect the growing status of crops. Therefore, we can monitor the growing status of crops by its spectral difference. Crop spectral analysis cosumes little time, doesn't cost chemical matters. The process of Crop spectral analysis is simple and the level of automation is high.In this study the canopy reflectance spectra during the whole growing period on rice field plots treated with different nitrogen levels were studied, the canopy spectral reflectance was sensitive to nitrogen levels in 560-610nm and 710-760nm before milkfilling. Furthermore, model equations between rice canopy spectral reflectance and vegetation index and the agronomy parameters (LAI and chlorophyll content and chlorophyll density), the estimating models of LAI chlorophyll content and chlorophyll density were built.This paper uses Visual Basic to integrate the software and hardware system, on the one hand Multi-spectral Radiometer (MSR-16R) can gathers the soil and rice canopy spectrum in real time , GPS can provide the precise spatial attributes for spectrum data, and realizes integration in spectrum data gathering, orientation and storage.On the other hand spectrum data can be analysed and processed combined with models built by experiments research of differen...