Spatial Distribution And Influencing Factors Of Soil Heavy Metals Based On Moran's I And Geographically Weighted Regression

Due to its unique geographical location,the multiple districts of the suburb have made its status increasingly important in the process of rapid urbanization.At the same time,the pollution of farmland soil has also become increasingly serious in the compounded area.Heavy metals have caused serious damage to farmland soil,crops,and even human health because of the characteristics of hard-to-degrade and easy-to-enrich and have become the primary research object of soil pollution in the suburb.In this paper,taking composite contaminated soil of the industrial and sewage irrigation area as the research area in the eastern suburb of Kaifeng City.246 soil samples of the farmland were collected and the total amount of As,Hg,Cr,Ni,Cu,Zn,Pb,and Cd in the soil were detected.The available content of Cu,Zn,Pb and Cd,as well as the physical and chemical indicators of soil pH,organic matter,CaCO₃ and available phosphorus were also detected.Furthermore,using GIS spatial measurement tools to obtain the distance data between the sample points and the nearest villages,rivers,highways,railways and factories.Based on the Moran's I and Kriging spatial interpolation,the spatial heterogeneity of heavy metals in farmland soils of the suburb was studied.Meanwhile,the correlations between the four effective levels of Cu,Zn,Pb,Cd and the influencing factors were compared based on Ordinary Least Square and Geographically Weighted Regression.The main results are as follows:?1?The content of various heavy metals had different levels of excessive standards,mainly caused by the dust falling from the factories and the sewage irrigation.The average levels of Cd,Hg,Zn,Pb,and Cu were 12.5,7.4,2.5,1.9,and 1.4 times higher than the background values of fluvo-aquic soils,respectively,and the As,Cr and Ni seldom exceeded the background values.In general,the soil was most seriously contaminated by Cd,and there was cumulative pollution of Hg,Zn,Pb,Cu,and As in different degrees in the study area.the pollution of Cr and Ni was not obvious.In addition,the average effectiveness of heavy metal Cd do far exceeding the average effectiveness of the other three heavy metals.It shows that the Cd in farmland soil is more easily absorbed by plants,thus having a great threaten to human health.?2?Through Global Moran's I analysis,it was found that there was spatial autocorrelation of heavy metals.The global autocorrelation size is:T-As>A-Cd>T-Cd>T-Zn>A-Zn>A-Cu>T-Pb>A-Pb>T-Cu>T-Cr>T-Ni>T-Hg.Local Moran's I analysis shows that there were a high-value accumulation area of Cu,Zn,Cd,and Pb in the east of the Zinc Smelting Factory,the south of the Chemical Fertilizer Factory and the intersection of the Fertilizer River and the Longhai Railway.And the"high-high"accumulation area of A-Cu,A-Zn,A-Cd,and A-Pb were basically in accordance with the distribution of their corresponding total amounts,thus indicating that both the total amount of heavy metals and the contents of their effective state have a certain correlation.The"high-high"spatial accumulation of heavy metals indicates that high soil heavy metal content areas are surrounded by other areas with high soil heavy metal content.The heavy metal content is relatively high in this area and there is a potential pollution risk,which should be the focused area for the prevention of heavy metal pollution in soil.?3?Kriging spatial distribution interpolation was used to visually display the pollution distribution and pollution degree of various heavy metals in the study area.The order of pollution severity is:Cd>Zn>Hg>Cu>Pb>As>Cr>Ni.In the spatial interpolation pictures,the content of Cd and Zn exceed the national agricultural soil standard value by two times in the mid-western part of the study area,which is a key area for preventing and controling soil environmental pollution.The pollution scope of heavy metal As,Hg,Cu and Pb is relatively small,mainly distributed in the vicinity of Zinc Smelting Plant and Chemical Fertilizer Plant;However,there was no significant accumulation phenomenon for Cr and Ni in the study area.?4?The fitting results of the OLS model and the GWR model showed that the available levels of the four heavy metals have a significant positive correlation with the corresponding heavy metals.It was negatively correlated with pH and CaCO₃,indicating that when the pH value increased,the activity and content of heavy metals in the soil was reduced.At the same time,CaCO₃ inhibits the increase of the content of the heavy metal effective state by increasing soil pH and promoting the conversion of heavy metals to carbonate bound state,thus presenting a negative correlation.A positive correlation exists between the available contents of the four heavy metals and the available phosphorus,which was mainly caused by the side-seepage of sewage and the application of phosphorus fertilizer containing heavy metals.Due to the complexity of the mechanism of organic matter,the type of impact on the available state of heavy metals is not completely consistent,with the negative correlation being the main one.The distances from villages,rivers,highways,railways and factories to samples also have a certain effect on the content of soil available heavy metals.As the distance increases,the effect will gradually weaken.In addition,by comparing the adjust R² of model residuals of the OLS model and the GWR model,the GWR model solves the spatial autocorrelation problem of model residuals that can not be solved by the OLS model,which indicats that the local GWR model is superior to the global OLS model in interpreting the spatial heterogeneity of heavy metal availability and influencing factors in farmland soil.