Inversion Of Heavy Metals In Rice Land By Rice Canopy Spectrum

Applying remote sensing technology to identify and monitor crops heavy metal contamination is of great significance to agricultural production,food security and human survival environment.As the spectral characteristic of crops under heavy metal pollution were weak and unstable in the natural environment,the credibility of directly using the original spectrum to evaluate the ecological status of heavy metal pollution in farmland is not high.So,how to enhance the weak spectral characteristic information under heavy metal contamination stress,and how to establish the remote sensing calculation model of farmland ecological pollution stress levels,are the foundations of achieving the rapid and accurate monitoring for the farmland ecological heavy metal pollution remote sensing,but also the problem of quantitative development of remote sensing needs to besolved.On the basis of obtaining a good inversion model in the previous study of a heavily polluted arable land,here we select the low-polluted rice plantation area in HuaRong as the experimental study area,collect the field hyperspectral data,and detect the heavy metal content.Then using the spectral analysis to obtain the rice spectroscopy part of vegetation index and spectral parameters,establish the hyperspectral inversion model,such as Pb?Cu?Cd?Hg in soil by partial least squares regression method.The research results mainly include:1)The pollution degree of heavy metals such as Pb,Cu,Cd and Hg in the study farmlands:Cd to moderate pollution;Hg for the slight pollution,and less pollution points.2)The heavy metal content data is transformed by index to obtain the transformation data,the transformation data and the spectral data can achieve a high correlation.3)Applying the heavy metal content index in rice to the partial least squares method can improve the accuracy of the model.4)In the study area,the content of heavy metals such as Pb?Cu and Hg in the whole area or most of the area is lower than the local background value,it is difficult to get a better fit model;Cd pollution is more severe in the vegetation,it is easily captured by spectral information and get a better inversion model.As the research shows that combine the indexed transformation of heavy metal content from soil with spectral data able to obtain quantitative information on heavy metals that have reached pollution levels in soil,which provides a new idea and new method for remote sensing inversion of heavy metals in arable land,it has laid a technical foundation for the quantitative inversion from the measured spectrum to the inversion of the spectrum,and has a certain reference value for realizing large-scale monitoring of heavy metals in soil.