近年来，我国O3污染问题日益突出，尤其是京津冀地区较为严重，O3污染防治迫在眉睫。为探究O3变化规律，揭示典型O3污染过程光化学污染影响因素，提出有针对性、科学性、合理性的前体物挥发性有机物（VOCs）控制策略，本文于2016年夏秋季节在北京市下风向怀柔地区开展光化学污染外场强化观测，获取夏秋季节O3及其前体物污染特征，辨识典型O3污染过程，采用相关关系分析等多种手段研究典型污染过程中气象条件、前体物等因素对O3浓度的影响，建立光化学比值法和连续反应模型法联用的VOCs光化学消耗计算方法，基于光化学消耗VOCs的臭氧生成潜势（△OFP）和SOA生成潜势（△SOAP）研究VOCs关键活性物种，并应用PMF（正交矩阵因子）模型对初始VOCs的来源进行解析。主要研究结果表明： （1）观测期间O3小时浓度最大值179.7 ppb，超过GB3095—2012二级标准限值0.9倍。NOx和VOCs等光化学前体物昼间消耗显著，呈现出与O3基本相反的日变化特征，而纯粹由大气光化学产生的二次污染物O3、PAN、H2O2等则呈现中午及午后数小时高值的特征，光化学生成对维持这些二次污染物浓度起到了强烈的支撑作用，夏秋季节怀柔地区光化学污染特征显著。 （2）观测期间O3污染过程频发，出现了夜间O3高值、O3持续高值、O3和PM2.5依次升高、PM2.5和O3污染共存的四类典型污染过程。整个观测期间不同O3污染过程的影响因素有所不同。虽然不同污染过程O3敏感特征有一定差异性，但是怀柔地区大气O3的生成敏感化学特征以VOCs控制为主。 （3）使用光化学比值法和连续反应模型法联用的方法计算VOCs初始浓度和消耗浓度，如忽略光化学消耗，VOCs浓度将被低估约22.5%。基于消耗VOCs分析不同污染过程的△OFP，结果表明异戊二烯、间/对-二甲苯、1,3,5-三甲苯、1,2,4-三甲苯、丙烯是O3生成的关键活性物种。除此之外，芳香烃在PM2.5和O3污染共存过程中对SOA的形成也起到重要作用。 （4）基于初始VOCs的PMF解析结果显示，在O3持续污染过程中机动车尾气源对初始VOCs的贡献率（23.5%）居首位，其次是溶剂使用源（18.3%）、植物排放源（18.1%）、工业过程源（17.6%）、生物质燃烧源（12.1%）和煤炭燃烧源（10.5%）。与环境VOCs来源解析结果相比，提高了相对活性物质含量更高的污染源的相对贡献。研究结果表明，在北京怀柔地区典型O3污染过程中，减少机动车尾气源、溶剂使用源、上风向工业过程源的排放是控制怀柔地区VOCs的有效措施。

In recent decades, ozone pollution in China has become increasingly serious, especially in Beijing-Tianjin-Hebei region. Therefore, the prevention and controlling of ozone pollution is urgently needed. Aim to investigate the variation of ozone and influencing factors of photochemistry in typical ozone episodes, propose scientific, rational and targeted measurements for controlling of Volatile Organic Compounds (VOCs), an important precursors of ozone, a field campaign was conducted in Huairou, the downwind area of Beijing, in summer and autumn 2016. The characteristics of ozone and its precursors as well as the typical ozone episodes were identified. In this paper, the influencing factors of ozone pollution, such as meteorological elements and effect of precursors, were analyzed by comprehensive analysis method. Meanwhile, photochemical consumption of VOCs were calculated by the combined method of photochemistry ratio method and sequential reaction model. Besides, the ozone formation potential (△OFP) and SOA formation potential (△SOAP) based on consumed VOCs were also calculated to identify the key species of VOCs. And the source of VOCs based on initial concentration was investigated by the Positive matrix Factorization (PMF) model of EPA. The main results of this paper are shown as follows: (1) The maximum of hourly O3 concentration was 179.7 ppbv, which exceeded 0.9 times the Class 2 limit values of the National Ambient Air Quality Standard (GB3095-2012). Photochemical precursors, such as NOx and VOCs, exhibited opposite diurnal variations with O3, which showed the characteristics of low values at noon. Meanwhile, the secondary pollutants produced by atmospheric photochemistry such as O3, PAN and H2O2, showed the characteristics of high values at noon and several hours later in the afternoon. It indicated that photochemistry plays a significant supporting role in maintaining the concentration of these secondary pollutants. And photochemical pollution in Huairou during summer and autumn seasons has a significant feature. (2) With frequent occurrence of ozone pollution incidents, four typical ozone episodes occurred during the campaign, including high nighttime O3, persistent high concentration of O3, O3 and PM2.5 rising in turn, and high PM2.5 and O3 at the same time. The formation mechanisms of ozone in the 4 episodes were different. Though there was a significant difference on O3 sensitivity in 4 episodes, the O3 formation in Huairou was VOCs limited in general. (3) Photochemical consumption of VOCs were calculated by the combined method of photochemistry ratio method and sequential reaction model. The average value of VOCs would be underestimated by 22.5%, if the photochemical losses were neglected. The △OFP values indicated that isoprene, m,p-xylene,1,3,5-trimethylbenzene, 1,2,4-trimethylbenzene and propylene were the key active species for ozone formation. Besides, aromatic was one of significant species in SOA for both O3 and PM2.5 high-value processes. (4) Positive Matrix Factorization (PMF) model based on initial VOCs was applied to identify the dominant emission sources and examine their fractions in total VOCs. It was found that vehicle exhaust was the major contributor to the initial VOCs (23.5%), and followed by solvent use source (18.3%), plant source (18.1%), industrial process source (17.6%), biomass combustion source (12.1%) and coal combustion source (10.5%). Compared with the results based on observed VOCs, the value of relative contribution has been improved for the sources with higher relative active VOCs species. The results also indicated that in the typical O3 pollution process, reducing the emission of vehicle exhaust, solvent source and industrial process source in windward direction was an effective measure to control VOCs in Huairou.