Spatial Variability Analysis Of Soil Heavy Metal Element On GIS And Geostatistics

Geographic information systems(GIS)has the abilities of regional synthesis and dynamic prediction. Kriging as a method for modeling, the unbiased and optimal character can give the attribute data of estimated points a best approximation to its mathematical expectation. Using GIS spatial analysis and geostatistics technology and referring to soil environment quality assessment technology and methodology, this paper analyzed the actuality and tendency of soil heavy mental contaminates, discussed the regularities and reason of spatial distribution. The main contents are as follows:(1) The statistic analysis show that Cu,Zn,Cd,Cr content are comparatively not high, but Hg and As by contraries.(2) Using regionalized variable theoretic and simulative empirical variation function, we choose theoretical formula to analysis the heavy mental in soil—Hg—the contaminated badly mental. The result shows that the content of Hg was normal distribution after logarithmic transformation and has evident spatial correlation; the content of Hg exist the southeast and northwest direction trend of second order function and has evident spatial variability. Also an obvious nugget effect is found in the content of Hg and nugget value is 0.258, and the structural variation of variable accounts for major. The content of Hg don't has significant change laws in small-scale, the spatial variability was effected mostly by human factors, such as: fertilization,crops and management level, but there is some graded distribution law in middle scale. It indicates that man-made activities affect Hg content deeply and weaken the function of space autocorrelation.(3) We made the space distribution isoline map of Hg. By the map we know that its content only behaves continuous distribution locally. In the whole research area, the distribution is relatively continuous and can be plot out to be several high-value sections approximately, which centralize in a northwest-southeast zonal region. The high-value sections in C and D area affected mainly by industry and those in a area mainly by agriculture. (4) Discussed the application of disjunctive kriging(DK).