区域土壤重金属空间分布驱动因子影响力比较案例分析

土壤重金属空间分布受到自然和人为驱动因子的共同影响，识别并评价各驱动因子对土壤重金属空间分布的影响力，对解析区域土壤重金属污染源、研究重金属迁移规律以及探明重金属空间分布模式具有重要意义.为研究不同自然和人类活动背景下各驱动因子对土壤重金属空间分布影响的差异性，并寻找在异质背景下的高影响力因子，采用地理探测器和随机森林模型，以贵州省安顺市作为土壤重金属高背景值典型地区、辽宁省葫芦岛市作为人类活动高强度典型地区，探讨市域尺度下土壤pH、地面累年值年值气温、坡度、海拔、地面累年值年值降水量（08:00）、夜间灯光指数和县GDP这7个驱动因子的空间分异性，及其对研究区内5种重金属（As、Cd、Cr、Hg、Pb）空间分布的单独及交互驱动力强度.结果表明:安顺市As、Cd、Cr、Hg、Pb这5种重金属的质量分数平均值均高于葫芦岛市，但w（Hg）差异不明显；在市域内采样点平均间距为10 km的尺度下，安顺市和葫芦岛市土壤重金属分布受到自然因子的影响程度均强于人为因子；在异质环境背景下，安顺市和葫芦岛市土壤pH、海拔和县GDP的驱动能力较稳定，并且是对土壤重金属分布影响能力最大的3个因子，适宜在上述地区作为土壤重金属含量多元非线性回归模型的通用参数.研究显示，基于地理探测器的因子驱动力评价方法可以应用于评价其他因子对土壤重金属含量空间分布影响力，以及用于土壤重金属含量回归方程自变量的选择. 

Comparative Case Study on the Influence of Spatial Distribution of Heavy Metals in Regional Area

The spatial distribution of soil heavy metals is influenced by natural and human driving factors. Identifying and evaluating the influence of these factors is of great significance for identifying the sources, migration and spatial distribution patterns of heavy metals in soil. To research the effects of various driving factors on the spatial distribution of heavy metal in soil under different background of natural and human activities, and to find the high-impact factors under heterogeneous backgrounds, Anshun City and Huludao City, which are respectively characterized with high background value of heavy metals and high intensity of human activity, were chosen as the study area. The driving factors including pH, annual ground temperature, annual ground precipitation (08:00), elevation, slope, night-time light index and GDP were taken into account. Methods including geodetector and random forests were adopted to analyze the individual and interactive effects of the driving factors on the spatial distribution of As, Cd, Cr, Hg and Pb in the typical regions. Results showed that the mean concentration of each heavy metal in Anshun City was higher than Huludao City, but the distinction of Pb concentration was not significant. In the average sampling density of 10 km, the influence of natural driving factors was stronger than human factors in the typical regions. Soil pH, elevation and GDP were the strongest driving factors that greatly affected the spatial distribution of heavy metals in soil among the referred 7 factors. Meanwhile, they were relatively stable in the heterogeneous backgrounds, and were suitable to be used as the general parameters in the multiple nonlinear regression model for calculating heavy metal concentrations in soils in the study area. The evaluation method which based on geodetector can be used to evaluate the influence of other factors on the spatial distribution of heavy metal concentrations in soil. Moreover, it is useful to select the independent variables in the regression equations to predict heavy metal concentrations.