This article combined multivariate linear regression method with Geostatistics method tostudy the spatial distribution and pollution by analysis the elements content of heavy metals inHainan island surface soil: Cr, Pb, Cu and Zn. Established multiple linear regression models:the independent variables consist of longitude and latitude, altitude and slope, the content ofheavy metals are the dependent variables; Add simulation points; The original data with thesimulation point values combination interpolation, comparison the precision of the improvedmethod and the traditional method (the original data direct interpolation), the results showedthat:1ã€In the original data, the average content of Cr is62.486ug/g, Pb is30.411ug/g, Cu is54.611ug/g, Zn is69.579ug/g. The content of Cr, Pb, Zn are obviously less than the primarystandard of Soil environment quality standard in China(GB15618-1995); The content of Cu islarger than the second standard of Soil environment quality standard inChina(GB15618-1995); In general, the quality of soil is good. In the improved data, theaverage content of Cr is57.330ug/g, Pb is Pb为34.401ug/g, Cu is51.062ug/g, Zn is65.726ug/g. The content of Cr, Pb, Zn are obviously less than the primary standard of Soilenvironment quality standard in China(GB15618-1995); The content of Cu is a few morethan the second standard of Soil environment quality standard in China(GB15618-1995).2ã€In traditional method, Among the heavy metals in the surface soil of Hainan island, Crbelongs to the Spherical model. Pb and Zn belong to the Gaussian model. And Cu suits forExponential model. In terms of spatial correlation, from small to large in the order: Cr(0.1%)>Zn(28.2%)>Pb(29.7%)>Cu(35.2%); Cr is characterized by strong spatial correlation; Zn, Pband Cu were characterized by moderate spatial correlation. In proved method, the content ofCr, Pb, Cu and Zn suit for Exponential model. In terms of spatial correlation, the coefficientof spatial variation from small to large in the order: Zn(2.7%)>Pb(8.4%)>Cu(34%)>Cr(49.9%); Pb and Zn is characterized by strong spatial correlation; Cr and Cu werecharacterized by moderate spatial correlation.3〠Except Pb, the high values of the content of Cr, Cu and Zn are located inHaikou-Laocheng-Ligao-Yangpu zone and Sanya in surface soil of hainan island. There is adecreasing trend from east to west in the content of Cr and Zn; Cu on the contrary; Thecontent of Pb shows the tendency of decrease from the central hainan island to the north andsouth. Improved method is better than the traditional method on local smooth; Distribution of four heavy metals are plaques in improved method; The high values of the content of Pb arelocated in the central mountain; The high values of the content of three other elements arelocated in Haikou-Laocheng-Ligao-Yangpu zone. Except Pb,illustrate that the distribution ofindustry have great influence on the content of heavy metals in surface soil in hainan island.Resampling validation,the error of Cr in the traditional kriging interpolation method isminimum, precision is the best; The error of Pb in the traditional interpolation method isminimum, precision is the best; The error of Cu and Zn in the improved interpolation methodis minimum, precision is the best.4ã€Evaluating the original data through the single factor evaluation method. The resultindicates that: Cr pollution appeared in the Surface soil in northern hainan island, mainly inhaikou, Ligao and Yangpu, especially Haikou contaminated area is the largest and serious, thecoefficient of Single factor of Cr is larger than2in surface soil of Haikou; Pb, Cu and Znpollution do not appeared in the Surface soil in Hainan island; However, in some places, thecoefficient of Single factor of Cu and Zn are more than0.7, and measures should be taken toprevent its deterioration. Evaluation the original data among the different terrain units byNemero comprehensive pollution index method, the result shows that the level of thecomprehensive pollution of heavy metals in surface soil in Hainan island is Hilly> MesaPlains> Central Mountain. |