Based On The Spectral Characteristics Of Soil Samples And Heavy Metal Inversion In Urban Samples Of Shanghai

Nowadays,heavy metal pollution is one of the hot issues of society.Compared with suburban area,the soil in urban area contains more pollution in respect of heavy metal,which can exist for a long time.Through water,atmosphere and the food we eat,the heavy metal enters our body and do harm to our health.The metropolis Shanghai boosting the high economic progress while has a large quantity of industries in different sector.All of this lead to the high consumption of energy and soil and affects greatly the environment.The soil pollution is highly concerned.Conducting the survey,evaluation and amending is important for the city environment protection.The traditional inspection way to the heavy mental was primarily the chemical detection,which characterized with accuracy but limited range.That method can only get the instant information which will waste time and do destructive damage to the soil.Compared with suburban area,the soil in urban area contains more pollution in respect of heavy metal,which can exist for a long time.Through water,atmosphere and the food we eat,the heavy metal enters our body and do harm to our health.The metropolis Shanghai boosting the high economic progress while has a large quantity of industries in different sector.All of this lead to the high consumption of energy and soil and affects greatly the environment.The soil pollution is highly concerned.Conducting the survey,evaluation and amending is important for the city environment protection.The traditional inspection way to the heavy mental was primarily the chemical detection,which characterized with accuracy but limited range.That method can only get the instant information which will waste time and do destructive damage to the soil.This paper chooses Shanghai as the area with experimental analysis,determination of soil reflectance spectra and soil heavy metals data,analysis of the characteristics of soil spectral curve of the city,through multiple stepwise regression to determine the sensitive band inversion of soil heavy metals,by partial least squares method to establish spectral model of soil heavy metal sensitive bands based on the best inversion.The main conclusions are as follows:(1)The results of heavy metal data showed that the heavy metals in Shanghai urban soils exceeded the background values of soil heavy metals in Shanghai.Heavy metals in soil have obvious spatial differentiation characteristics,which are mainly characterized by the combination of island and strip.The results showed that most ofthe soil samples were polluted by more than two kinds of heavy metals,and the soil Zn,Pb and Cu were homologous.(2)The spectral reflectance of the soil in the study area showed an upward trend,which belonged to the slow oblique type.Most of the spectral curves were basically the same.According to the analysis and comparison of the characteristics of the spectral curve,we can see that Xuhui is greatly influenced by human beings,and the spectral reflectance curves are more dispersed.The spectral reflectance curve of Minhang District is small,but the dispersion degree of Fengxian District spectral reflectance curve is moderate.The spectral characteristics of different land use types were analyzed,and the spectral reflectance curve of residential land was the largest in agricultural land,industrial land,residential land and green land.Under the same land use type,the trend of the soil spectral reflectance curve is the same,and the variation range of reflectance is smaller.The spatial distribution of spectral reflectance was zonal distribution,and the middle area of the study area changed to the East and west.(3)The correlation analysis between the data of the soil heavy metal and the original spectrum is made.The correlation between the Cu and the square root,the first order transform data and the data of the two order transformation is higher than that of the.The correlation between the Cr and the continuum removal,the first-order transform data and the two order transform data is higher.The correlation between the Mn and the continuum removal,the first-order transform data and the two order transform data is higher.The correlation between the Pb and the continuum removal,the first-order transform data and the two order transform data is higher.Zn has a high correlation with the reciprocal and reciprocal logarithm,the first order transform data and the two order transform data.The results of correlation analysis show that the correlation between the spectra and the heavy metal is improved to a great extent by the first order transform data and the two order transform data.It is reasonable to use the first order transform data and the two order data to carry on the principal component analysis and the establishment of the heavy metal content estimation model.(4)It is necessary to understand the sensitive bands of the spectrum by means of the differential transform technique.Principal component analysis was used to extract the 11 principal components representing the dominant factors of the first order spectral data transformation.The 14 principal components are extracted to represent the dominant factors in the spectrum of the two order data.The extracted informationcan truly reflect the characteristics and rules of the objects,simplify the data and eliminate redundant data,so it is necessary to analyze the data in the relatively low dimensional space.(5)Fitting multiple stepwise regression and partial least squares model,by R-and RMS contrast analysis for the five kinds of heavy metals,except Mn,four kinds of heavy metal elements Cu,Cr,Zn,Pb model forecast better is the partial least squares regression model.Therefore,it can improve the prediction effect of the model and the inversion precision of the spectral data to the heavy metals in soil.After the establishment of the model and the comparison of the parameters,the partial least squares fitting is better than the multiple linear regression.