随着我国社会经济快速发展，土壤重金属污染问题日益严重。重金属可在土壤环境中长期累积滞留，通过多介质、多途径直接或间接给人体健康或生态系统带来污染风险。尤其是近年来，与土壤相关的重金属污染事件频发，对土壤重金属污染进行风险分析、评价和调控愈发紧迫。目前，对于建设用地土壤污染的人体健康风险评价相对成熟，但对生态系统风险则多停留在动植物毒理学研究层面上，较少涉及区域尺度的定量化、概率化风险表征。为此，论文针对区域土壤重金属污染风险“多源、多受体、多途径、多效应终点”的特点，建立区域土壤重金属污染风险综合评价框架；采用经验贝叶斯克里金模型和径向基函数获取区域土壤重金属污染分布，分别进行人体健康风险评价和概率生态风险评价，通过风险等级组合实现每个风险小区的综合评价。选取长江重要节点城市——铜陵市长江南岸地区开展案例研究，主要的研究内容和成果如下：

（1）建立区域尺度土壤重金属污染风险的数据库，完成插值预测和污染分析

将采样监测、资料收集获取的土壤重金属浓度，文献计量学获取的毒性数据和实地调研获取的风险源、风险受体信息等整合为研究区土壤重金属污染风险数据库。选取经验贝叶斯克里金模型和径向基函数对土壤重金属初始浓度进行“去聚加密”，插值预测的结果弥补了数据的自身缺陷，不仅让区域土壤重金属污染分布可视化，也为后续风险评价打下基础。采用地累积指数和潜在生态风险指数法对插值后的浓度数据进行污染分析，得到的结果显示土壤存在一定程度的Cu、Cd、Hg污染，且钟鸣镇、天门镇、西湖镇和大通镇的污染程度较为严重，有必要进一步开展区域土壤重金属污染风险评价。

（2）从土地利用类型分类出发，完善区域土壤重金属污染人体健康风险评价过程

采用不同角度划分风险小区，在已有的“风险识别—暴露评估—毒性效应评估—风险表达”框架的基础上，由单一建设用地或农用地的人体健康风险评价扩充到两大地类分组进行评价，选取适用于不同地类的人体健康风险评价模型计算两种风险小区划分方式下的人体健康风险值。再将不同用地类型面积占比作为权重，结合人口分布情况，整合乡镇单元下区域土壤重金属污染人体健康风险分组评价的结果。研究结果表明：两大用地类型中，不同土壤重金属污染的人体健康风险评价结果有差异，Cu和Hg在农用地的非致癌风险大于建设用地的，其余重金属则是在建设用地的风险较高；人体健康风险的优先管理单元为天门镇、顺安镇和钟鸣镇。完善后的评价过程在保证结果和分析更科学的同时，便于进行风险防控与管理。

（3）改进区域土壤重金属污染生态风险评价方法，实现风险定量化、概率化表达

以研究区土壤重金属污染风险数据库为基础，将暴露浓度和毒性数据作为两个随机变量，前者经过对数变换进行累积分布，即暴露浓度分布；后者在六种分布函数Log-normal、Log-logistic、Triangular、Burr III、Weibull及Gumbel中选择最优拟合曲线表示物种敏感性分布，即毒性效应分布。运用联合概率的思想计算两种分布的重叠面积，即单一重金属的预期生态风险。结果表明，研究区土壤Zn和As的生态风险较高，分别为41.99%和31.86%。将两个随机分布联合的过程可以概率化、定量化地进行生态风险评价。

（4）搭建区域土壤重金属污染风险综合评价框架，开展区域尺度案例研究证实框架的可行性

基于不同角度划分风险小区，选取具有科学性的风险等级组合，对单一重金属污染风险进行综合评价。案例研究结果表明：从人体健康风险角度，Ni、As、Cd是对受体造成较大威胁的风险物质，研究区范围内38.8%的相关企业，71.4%的矿山，37.0%的尾矿库及一块遗留场地位于高致癌风险小区；从生态风险角度，Zn、As、Hg、Ni的危害性较大，受影响的物种比例分别为41.99%、31.86%、27.86%、20.68%。两种风险综合评价的结果显示：RR10和RR14的Ni污染，RR1和RR6的As污染要予以高度重视；以天门镇、钟鸣镇和顺安镇为代表，要对以上两种土壤重金属污染开展修复工作。

With the rapid development of my country's society and economy, the problem of heavy metal pollution in soil has become increasingly serious. Heavy metals can accumulate and stay in the soil environment for a long time, and bring pollution risks to human health or ecosystems through multiple media and multiple channels directly or indirectly. Especially in recent years, heavy metal pollution incidents related to soil have occurred frequently, and risk analysis, evaluation and regulation of soil heavy metal pollution have become more and more urgent. At present, the human health risk assessment of soil pollution in construction land is relatively mature, but the ecological system risk mostly stays at the research level of animal and plant toxicology, and rarely involves quantitative and probabilistic risk characterization at the regional scale. To this end, the paper establishes a comprehensive evaluation framework for regional soil heavy metal pollution risks based on the characteristics of "multi-source, multi-receptor, multi-path, and multi-effect end point" of regional soil heavy metal pollution risk; adopts Empirical Bayesian Kriging model and Radial Basis Function to obtain the distribution of heavy metal pollution in the regional soil, carries out human health risk assessment and probabilistic ecological risk assessment respectively, and realizes a comprehensive assessment of each risky community through a combination of risk levels. Select the important city of the Yangtze River—the area on the south bank of the Yangtze River in Tongling City to conduct a case study. The main research contents and results are as follows:

(1) Establish a regional-scale soil heavy metal pollution risk database, complete interpolation prediction and pollution analysis

The soil heavy metal concentration obtained from sampling monitoring and data collection, the toxicity data obtained from bibliometrics, and the risk source and risk receptor information obtained from field research are integrated into the soil heavy metal pollution risk database in the study area. The Empirical Bayes Kriging model and Radial Basis Function are selected to "de-aggregate and encrypt" the initial concentration of heavy metals in the soil. The results of interpolation prediction make up for the data's own shortcomings, not only visualizing the distribution of heavy metal pollution in the regional soil, but also laying the foundation for subsequent risks evaluation. Pollution analysis of the interpolated concentration data using the geo-accumulation index and the potential ecological risk index methods shows that there is a certain degree of Cu, Cd, and Hg pollution in the soil, and the pollution levels in Zhongming Town, Tianmen Town, Xihu Town and Datong Town are relatively serious. It is necessary to further carry out regional soil heavy metal pollution risk assessment.

 (2) From the perspective of different types of land, improve the regional soil heavy metal pollution human health risk assessment process

Based on the existing "risk identification-exposure assessment-toxic effect assessment-risk expression" framework, divide risk communities from different perspectives. the human health risk assessment of a single construction land or agricultural land has been expanded to two major land groups for evaluation. Select human health risk assessment models suitable for different geographical categories to calculate human health risk values under the two risk community division methods. Then take the proportion of different land types as weights, and combine the population distribution to integrate the results of the group evaluation of the regional soil heavy metal pollution human health risk under the township unit. The results of the study show that the human health risk assessment results of heavy metal pollution in different soils are different in the two major types of land. The non-carcinogenic risk of Cu and Hg in agricultural land is greater than that in construction land, and the risk of other heavy metals in construction land is higher; The priority management units for human health risks are Tianmen Town, Shun'an Town and Zhongming Town. The improved evaluation process ensures that the results and analysis are more scientific, and at the same time facilitates risk prevention, control and management.

(3) Improve regional soil heavy metal pollution ecological risk assessment methods to achieve quantitative and probabilistic expression of risks

Based on the soil heavy metal pollution risk database in the study area, the exposure concentration and toxicity data are taken as two random variables. The former undergoes a logarithmic transformation for cumulative distribution, that is, the exposure concentration distribution; the latter is based on the six distribution functions Log-normal, Log-logistic, Triangular, Burr III, Weibull and Gumbel to represent the species sensitivity distribution, that is, the distribution of toxic effects. The idea of joint probability is used to calculate the overlapping area of the two distributions, that is, the expected ecological risk of a single heavy metal. The results show that the ecological risks of Zn and As in the study area are relatively high, 41.99% and 31.86%, respectively. The process of combining two random distributions can carry out ecological risk assessment in a probabilistic and quantitative manner.

(4) Build a comprehensive regional soil heavy metal pollution risk assessment framework, and carry out regional-scale case studies to verify the feasibility of the framework

Divide risk communities based on different perspectives, select a scientific risk level combination, and conduct a comprehensive evaluation of a single heavy metal pollution risk. The results of the case study show that from the perspective of human health risks, Ni, As, and Cd are risk substances that pose a greater threat to the receptor. 38.8% of the relevant enterprises in the study area, 71.4% of the mines, 37.0% of the tailings ponds and a legacy site are located in a high carcinogenic risk area; from the perspective of ecological risk, Zn, As, Hg, and Ni are more harmful, and the proportions of affected species are 41.99%, 31.86%, 27.86%, and 20.68%, respectively. The results of the comprehensive evaluation of the two risks show that: Ni of RR10 and RR14, As of RR1 and RR6 should be highly valued; With Tianmen Town, Zhongming Town and Shun'an Town as representatives, remediation work should be carried out on the heavy metal pollution of the above two types of soil pollution.