基于受体—后轨迹耦合模型的大气颗粒物PM2.5区域源方位解析方法研究

Deterioration of urban air quality has become a widespread concern both in the developed and developing countries. PM2.5 (particulate matter less than 2.5 microns in diameter), an important pollutant in urban ambient, strong associates to the prevalence of mortality and morbidity. So, there are numerous studies focus on the PM2.5 level in urban ambient. .According to several studies, urban PM2.5 receptor is not only attributed from the local source categories, but also from the regional source categories. The regional source is transported from different location areas. So the contribution of each regional source is the sum of the attributions from different location areas. Thus, for PM2.5 regional pollution, we should know the local and regional source categories and their contribution; what more, for each regional source, the attributions from different locations should be quantitative calculated (In this study, we define it as "Location Apportionment"). .However, existing source apportionment technology has been fallen short of PM2.5 regional pollution. Firstly, there was no sophisticated method to discriminate the local and regional sources; secondly, it is difficult to location apportion the regional sources..In this project, we want to develop a Location Apportionment Method" for PM2.5 regional source, in order to quantify the attributions from different locations. The main research aspects of this project are listed as follows: 1) Using the receptor model to identify the possible source categories and quantify their contributions; 2) Using the back trajectories to analysis the potential origin of each source, in order to discriminating the local source categories and regional source categories; 3) For each regional source category, combining the receptor model with the back trajectory model as well as meteorologic data, to quantify the attributions from different locations. 4) Establishing a combined receptor- back trajectory model; 5) developing the Location Apportionment Method for PM2.5 regional source; 6) analyzing on reasonability and practicability of the results of this method..To sum up, the aim of this project is to develop a new Location Apportionment Method for PM2.5 regional source, which can provide detailed information of sources for urban PM2.5 controlling.

PM2.5已经成为城市环境空气中的重要污染物，区域性污染是PM2.5的主要污染特征之一。PM2.5区域污染防控首先需要识别本地源类和区域源类型，还需进一步定量解析区域源不同来源方位上的贡献率。而现有源解析技术尚未不能达到这一目的，已无法满足城市PM2.5区域污染的防控需求。申请人在前期工作中已经对PM2.5本地源和区域源类识别方法进行了研究，更针对源解析过程中实际问题建立了新的复合受体模型。本项目拟在现有源解析技术和前期工作的基础上，综合运用理论研究、试验分析及模型模拟等手段，通过设计合理的模型耦合技术路线、构建受体-后轨迹耦合模型，来建立本地源和区域源识别方法、区域源不同来源方位贡献率的定量解析方法，最终建立适合我国PM2.5区域污染特征的、易于操作的区域源类方位解析方法。为PM2.5污染的区域防控提供科学依据和理论指导，具有重要的科学意义和实用价值。

研究表明，PM2.5易形成区域性污染。PM2.5区域性污染的研究，已成为当前国际上研究热点之一。如何针对PM2.5的区域污染特征，从源头上进行防控，是城市的空气环境保护工作的首要任务。对区域源类进行方位解析，为PM2.5污染防控提供依据，具有重要的科学意义。受体模型和气象模型是大气颗粒物来源解析技术中两种重要的工具，长期以来，在国内外的源解析研究工作中得到深入的研究和广泛的应用。本研究利用受体模型和气象模型的特点，建立了受体—后轨迹耦合模型，识别区域、本地源类并定量解析其贡献。先采用受体模型识别PM2.5污染源类型、定量解析源贡献；再结合后轨迹分析，区分本地源类和区域源类；最后将受体模型解析结果连同后轨迹模型、气象参数耦合使用，在定性分析区域源类的可能来源方位的基础上，定量解析区域源不同方位上的贡献。耦合模型具体方法包含三个步骤：（1）源类识别和源贡献时间序列定量：利用PMF模型识别污染源类，并定量计算污染源时间序列的贡献值；（2）来向识别：利用后轨迹模型模拟气团来向，识别污染源的可能来向，并根据气团轨迹，划分本地、外来区域；（3）污染源来向贡献定量：将第一步受体解析出的各源类贡献时间序列与第二步识别出的主要来向进行一一对应耦合计算不同方位的源贡献。本研究采集了天津市PM2.5样品，利用构建的耦合模型，分析天津市污染源并对各方位来源进行解析。分析了颗粒物中无机元素、水溶性离子、碳组分等化学组成，构建了受体数据，纳入PMF模型进行识别，识别出扬尘、燃煤尘、机动车、二次硫酸盐&SOC和二次硝酸盐五个因子。其中二次硫酸盐&SOC的贡献最高（25%），其他依次为扬尘（24%），燃煤尘（21%），机动车（15%），二次硝酸盐（15%）。根据后轨迹分析，初步分析气团的可能来向，按方位将污染源来向分成下列方位：R1：中部；R2：西北；R3：东北：R4：东南：R5：西南；R6：西北（区域）；R7：东北（区域）；R8：东南（区域）：R9：西南（区域）。最终解析得到各污染源在不同方位的贡献，并讨论了季节性差异。最后利用结果诊断技术评估了结果，表明模型结果是可接受的。