区域突发性大气污染事件风险评价方法及其应用

A Method of Regional Scale Environmental Risk Assessment for Abrupt Atmospheric Pollution Events

大气污染事件风险评价对区域环境风险防控管理具有重要作用.已有研究多限于小尺度(化工厂或工业园区)单一项目或单一风险评价,较少考虑多源多受体的复杂影响及跨界、邻域影响,为此针对性提出区域尺度的大气污染事件风险评价方法.通过借鉴风险分级思想,在相对风险评估框架下利用矩阵法、模型模拟手段(多烟团大气扩散模式)、Arc GIS工具等评估"源-受体-影响(暴露效应)"风险链的相对风险值,并以风险受体为核心划分评价单元,以此综合区域的风险值.在沿海某城市的案例研究表明,该市14个风险评价单元中,高风险区2个,中风险区4个,低风险区8个.风险值跨度大(0—808),各评价单元大气环境风险空间差异明显(如风险源危害性、受体易损性、跨界影响).风险源主要包括石油加工、炼焦业和化学原料及化学品制造等工业企业,居民区和自然保护区受大气污染胁迫大.为验证模型的有效性和稳定性,采用蒙特卡洛模拟法进行不确定性和敏感性分析.结果表明,模型模拟分布均值与评估值相近,风险源对不确定性的影响最大.

Abrupt atmospheric pollution risk assessment provides a helpful tool in decision-making of risk prevention and mitigation in a region. However,previous assessments paid more attentions to a local scale( e. g. a chemical factory or an industry park) and few considered complex risks from multi-stressors to multi-receptors or even transboundary risks. Herein,a Regional-scale Accidental Atmospheric Pollution Risk Assessment( RAAPRA) method is proposed which adopts a risk ranking procedure to characterize multiple complex risk routes of ' stressor-receptorimpact' using quantitative tools( e. g. the multi-puff model). The matrix technique,instantaneous atmospheric quality model,and Arc GIS tools are used to refine the risk ranking procedure within the framework of a relative risk model of the regional-scale ecological risk assessment. Using the RAAPRA method,a risk map for a coastal city was generated. The results show that there are two high risk regions,four medium risk regions and eight high risk regions.The range of risk scores is relatively wide( 0-808) and all risk zones differ obviously in spatial pattern. Major risk sources were identified as enterprises within petroleum processing,coking and chemical industries,which mainly pose abrupt atmospheric pollution threats to residential areas and nature reserves. The uncertainty analysis and sensitivity analysis were completed by Monte Carlo method to verify the robustness and stability of the new method. The results show that the mean values of the distributions are similar to the risk forecast,and risk sources are the main contribution to the parameters uncertainty of the model.