【作者】 王倩；

【导师】 王杨君；

【作者基本信息】 上海大学 ， 环境工程（专业学位）， 2021， 硕士

【摘要】 二次有机气溶胶（SOA）是细颗粒物(PM_2.5)的重要组成部分。对空气质量、气候变化和人体健康等方面有重要影响。然而,由于模拟中多种前体物的缺失（比如:中等挥发性有机物,IVOC）和生成机制的不完善,空气质量模式的SOA模拟浓度与观测值通常相差较大。本研究主要通过空气质量模式WRF-CAMx考察IVOC和1.5维挥发性区间分级法（1.5D-VBS）对长三角地区SOA生成模拟的影响。首先,分别采用IVOC排放因子法和排放比值法（IVOC与一次有机气溶胶,POA的排放比值）,构建了长三角地区高时空分辨率的IVOC排放清单;然后,采用空气质量模式对不同方法建立的IVOC排放清单对长三角区域SOA生成的影响进行模拟,同时也对比了1.5D-VBS机制和CAMx模式中默认的SOAP机制对SOA生成的影响;最后,长三角地区的SOA进行来源解析模拟,其结果可为长三角地区大气细颗粒物污染的精细化防控提供科技支撑。基于两种方法分别估算的长三角地区2017年机动车IVOC排放清单及计算相应的SOA生成潜势的结果表明,由不同方法得到的机动车IVOC排放量差异较大,相应地SOA生成潜势差异也较大。基于IVOC/POA系数法构建的2017年长三角地区高时空分辨率IVOC排放清单中,IVOC的排放总量约为730 Gg,其中,机动车和工业源是主要排放源,分别占IVOC排放总量的36.1%和21.4%;使用排放因子法构建的清单中IVOC排放总量为333 Gg,约为前者的一半,其中工业源和非道路移动机械是主要排放源,占比分别为46.9%和22.4%。长三角地区2018年7月SOA生成的模拟结果表明,使用默认机制（SOAP）且不考虑IVOC排放时（基准情景）,与观测值相比,淀山湖站点SOA的模拟浓度低估了约61.0%。加入IVOC排放清单后,基于SOAP机制的淀山湖站点的SOA模拟浓度提高了7.9%～26.8%,上海市SOA平均浓度提高了8.2%～27.6%,长三角地区的SOA平均浓度提高了5.2%～26.2%;将IVOC的SOA产率增至5倍后,淀山湖站点的SOA模拟值与观测值更加吻合,其模拟浓度提高了135.8%,上海市和长三角地区SOA模拟浓度也显著增高。相较于传统的SOAP机制,采用1.5D-VBS机制后,淀山湖站点的SOA模拟浓度较基准情景（默认的SOAP机制和产率）升高了75.6%,上海市和长三角地区的SOA模拟浓度分别增加了89.7%和41.9%。使用1.5D-VBS机制后,春季、秋季和冬季的SOA模拟结果均比基准情景提高了约2倍。研究结果表明,IVOC排放和1.5D-VBS机制均能显著地提升长三角地区SOA生成的模拟效果。考虑到目前公开发布的CAMx版本中,颗粒物源解析技术（Particulate Source Apportionment Technology,PSAT）仍无法直接应用VBS机制对SOA进行来源解析,所以本研究采用基于默认的SOAP机制的PSAT技术对长三角地区2018年不同季节SOA的区域来源解析进行了模拟,并用基于1.5D-VBS的brute force方法的源解析结果进行验证,其结果发现两者吻合较好。SOA的区域溯源结果表明,安徽省、江苏省和浙江省是长三角地区SOA的主要贡献者。浙江省内排放在春季和夏季对长三角区域SOA浓度平均贡献率较高,贡献率分别为20.3%和28.4%;秋季和冬季则为江苏省排放贡献较高,贡献率分别为28.2%和26.1%。源解析结果还表明,在春季和冬季,工业源排放对长三角地区SOA浓度平均贡献率较高,贡献率分别为55.5%和65.8%;夏季天然源的贡献率较高,为36.1%;秋季的主要贡献源为溶剂使用源和工业源,贡献率分别为40.1%和34.8%。对上海市各国控站点的源解析结果表明,除夏季在背景站（淀山湖站）的天然源贡献率会高于其它站点外,本地源类对各站的SOA浓度贡献率差异不大;对合肥市、南京市和杭州市的城区站点的各自省内源类解析结果表明,春季和冬季工业源贡献较大,夏季和秋季是溶剂使用源和工业源贡献率较高。更多还原

【Abstract】 Secondary organic aerosol（SOA）is an important component of fine particulate matter(PM_2.5)with significant impacts on air quality,climate change and human health.However,due to the absence of precursor emission inventory and uncertainties in the modeling scheme,SOA is usually underestimated compared to observations.Therefore,the objective of this study was to quantify the impact of intermediate volatile organic compounds（IVOC）emissions and the 1.5D-VBS mechanism on simulated SOA over the Yangtze River Delta（YRD）region of China using an integrated meteorology and air quality model（WRF-CAMx）.IVOC is reported to be one of the key precursors of SOA but was not considered in the modeling of SOA over the YRD region.In this study,IVOC emission inventory with high spatial and temporal resolution was first developed for the YRD region based on two methods（i.e.,the emission factor method and the emission ratio method of IVOC to primary organic aerosol,POA）.Secondly,the impacts of IVOC emissions and 1.5D-VBS mechanism on the simulation of SOA over the YRD region were investigated,respectively.Finally,source apportionment of SOA in the YRD region was simulated.IVOC emissions from vehicles and their corresponding SOA formation potential based on two methods both exhibit substantial differences.Based on the IVOC/POA coefficient method,the total amount of IVOC emissions are estimated to be 730 Gg with on-road and industry sectors being the main contributors（36.1%and 21.4%,respectively）.The total IVOC emissions estimated by the emission factor method are333 Gg,which is about half of the first method,with industry and off-road sectors being the major contributors,accounting for 46.9%and 22.4%,respectively.In the base scenario,the simulated SOA concentration for July 2018 based on the default emission inventories（no IVOC emission inventory）and the default modeling scheme（Secondary Organic Aerosol Processor,SOAP）significantly underestimated observed SOA concentration by 61%at the Dianshan Lake monitoring site（DSL）.After adding the IVOC emissions,simulated SOA concentration increased by 7.9%～26.8%,8.2%～27.6%,5.2%～26.2%at DSL,Shanghai,and the YRD region,respectively.Increasing the SOA mass yields from IVOC by a factor of five leads to better agreement with observations with average simulated SOA concentration increased by 135.8%at the DSL site;simulated SOA concentrations in Shanghai and over the YRD region both increased significantly.The 1.5D-VBS approach increases simulated SOA concentration by 76%at DSL compared to the base scenario（default SOAP scheme and yields）,and the simulated SOA concentrations in Shanghai and YRD region increased by 89.7%and 41.9%,respectively.In addition,with the 1.5D-VBS scheme,the simulated SOA concentrations were nearly 2 times better than the base scenario in spring,autumn and winter.These results show that both IVOC emission and 1.5D-VBS scheme can greatly enhance the SOA formation.Considering that the VBS scheme has not been applied with source apportionment,PSAT with the default scheme（i.e.SOAP）was used to simulate the source apportionment of SOA over the YRD region in 2018,which is evaluated against the results obtained using a Brute Force method based on 1.5D-VBS scheme.Evaluation results between the two methods showed good agreements.The results of SOA regional source apportionment show that Anhui,Jiangsu and Zhejiang provinces are the main contributors of SOA in the YRD region.In spring and summer,Zhejiang province contributed 20.3%and 28.4%to SOA concentration,respectively.In autumn and winter,the contribution from Jiangsu province was 28.2%and 26.1%respectively.Industrial source emissions contribute most to SOA concentration in the YRD region in spring and winter with contribution rates of 55.5%and 65.8%,respectively,while the contribution of natural source is higher during the summer with contribution of 36.1%.In autumn,the main contribution sources were solvent use sources and industrial sources,with contribution of 40.1%and 34.8%,respectively.The results of source apportionment in national monitors in Shanghai show that the contribution of natural sources at the background site（DSL）is higher than those at other stations in summer,and the contribution of local sources at all stations show negligible differences.The results in Hefei,Nanjing and Hangzhou showed that industrial sources contributed more in spring and winter,while solvent use sources and industrial sources contributed more in summer and autumn.更多还原