Inversion Of Soil Properties With Hyperspectralreflectance In Well-facilitied Capital Farmland Construction Areas Of Xinzheng City

Well-facilitied capital farmland construction is an important measure to enhance food security ability and accelerate new-style the modernization agriculture.In the process of Well-facilitied capital farmland construction,obtain to soil basic information quickly and accurately,that is the basis of Well-facilitied capital farmland construction,and the key to the modernization agriculture development.As the frontier technology in the field of remote sensing,Hyperspectral remote sensing has the advantages of multi-band and high resolution,which can be used to detect the slight difference of soil properties,and provide technical support for obtaining soil basic information quickly and efficiently.Therefore,the research about soil properties inversion by using hyperspectral remote sensing technical,can provide data basis and technical support for realization Well-facilitied capital farmland construction,and provide a reference for exploring the optimization of Well-facilitied capital farmland construction areas,and promote the implementation of Well-facilitied capital farmland construction.In this dissertation,taking the soil properties of Well-facilitied capital farmland construction areas of Xinzheng City as research object,combining field sampling and indoor hyperspectral spectroscopy(350~2500nm),the best spectral characteristic bands of six spectral transformations were selected,hyperspectral inversion model of single soil propertie was established in Well-facilitied capital farmland construction areas of Xinzheng City based on Partial Least Square Regress;The best spectral characteristic bands shared of six spectral transformations were selected,hyperspectral composite inversion model of soil properties was established in Well-facilitied capital farmland construction areas of Xinzheng City based on the Panel data model.By comparing the accuracy of the model,the best model was selected;Recommended standard was proposed for Well-facilitied capital farmland construction of Xinzheng City,and Well-facilitied capital farmland construction areas of Xinzheng City is optimized.The main results are as follows:(1)The best spectral characteristic bands of six spectral transformations were selected.The original spectral reflectance was smoothed by the Savitzky-Golay convolution smoothing method,then the spectral transformations such as Multiplicative scatter correction,Reciprocal logarithm,First order differential reflectance,Second order differential reflectance and Continuum removal were carried out.Based on the correlation analysis between six spectral indexes and soil organic matter,Available N,Available P,Available K,PH,Fe,Cr,Cd,Zn,Cu and Pb,the hyperspectral characteristic bands on the significance test of the P=0.01 level were screened.At the same time,comparative analysis of the correlation coefficient curves were given,the best spectral characteristic bands shared were selected by the fuzzy clustering maximum tree method.(2)Compared with the partial least squares model,comprehensive inversion model of soil properties was constructed in Well-facilitied capital farmland construction areas of Xinzheng City,that had high accuracy.Based on Partial Least Square Regress,with the best spectral characteristic bands of six spectral indexes as independent variable respectively,the inversion model of single soil properties was constructed.Based on the Panel data model,the best spectral characteristic bands shared of six spectral transformations as independent variable respectively,hyperspectral composite inversion model of soil properties was established.Compared with the prediction results and model precision,the prediction of the two models has reached a high accuracy,but Panel data model can inverted many soil properties at the same time through a model,and had higher precision.(3)Using the recommended standard of Well-facilitied capital farmland construction of Xinzheng City,optimization of Well-facilitied capital farmland construction area of Xinzheng City was studied.The recommended standard of Well-facilitied capital farmland construction of Xinzheng City were proposed,according to national standards and local policy standards.Based on the best hyperspectral inversion model,spatial interpolation was performed by the spatial Bayesian Kriging interpolation method;On the basis of comprehensive evaluation model of soil properties constructed by fuzzy Linear Membership Function and Principal Component Analysis,the comprehensive distribution of the study area was obtained,it showed that characteristics was high in south and low in north.Well-facilitied capital farmland construction areas of Xinzheng City was divided intoⅠLevel for the priority construction zone,II Level for conditional construction zone,Ⅲlevel area for restricted construction zone.The I Level area is 16953.7253hm~2,accounting for 45.72%of the basic farmland,theⅡLevel area is 15231.73378hm~2,accounting for 41.07%,theⅢLevel area is 4899.7844 hm~2,accounting for 13.21%.Compared to the measured values partition results,the results show that the hyperspectral inversion model can be used for optimization of Well-facilitied capital farmland construction of Xinzheng City.The main restriction factors of well-facilitied capital farmland of Xinzheng City were PH,AK and soil heavy metal pollution,to put forward the countermeasure about Well-facilitied capital farmland construction of Xinzheng City accordingly.