当前我国面临严重的大气臭氧（Ozone, O₃）污染问题，尤其在京津冀及周边地区（以下简称“2+26”城市）和汾渭平原。对于这两个区域O₃污染成因的研究主要集中在其中一个地区上。作为相邻的大气污染防治重点区域，“2+26”城市和汾渭平原常发生同步、大范围的O₃污染事件。然而，对于这类污染事件的成因研究仍相对不足。本研究立足该背景，利用外场观测和WRF-Chem（Weather Research and Forecasting model with Chemistry）数值模拟，并结合卫星数据和气象再分析资料等多种方法，从城市和区域两个尺度对“2+26”城市和汾渭平原的O₃污染成因进行了研究。得出以下主要研究结论：

（1）2020年北京总挥发性有机物（Total Volatile Organic Compounds, TVOCs）的浓度为23.4 ppbv，低于运城（38.3 ppbv）。两城市全年氮氧化物（Nitrogen Oxide, NO_x）浓度相当，北京为24.5 ppbv，运城为24.3 ppbv。两城市间的气象因子差异主要体现在：运城的气温（Temperature, T）和相对湿度（Relative Humidity, RH）均较高，而北京的日间风速（Wind Speed, WS）较高，夜间WS较低。源解析结果显示2020年7月对两城市挥发性有机物（Volatile Organic Compounds, VOCs）贡献最大的来源均为机动车排放。根据我们开发的区域传输贡献量化方法，结果显示，在10月份，区域传输对北京和运城的VOCs贡献达到最高水平，分别为37.6%和50.4%。具体而言，西南方向对北京VOCs的传输贡献最大，而对于运城来说，西方传输贡献最大。基于观测的VOCs/NO_x比值和基于卫星的HCHO/NO₂比值均显示，两城市在7月份的O₃敏感性存在差异。

（2）2020年7月北京的VOCs和NO_x浓度均高于运城。探究O₃前体物日变化特征，发现两城市OVOCs存在显著差异。进一步采用观测数据计算两城市的OH暴露量，发现北京的OH暴露量（6.3×10¹⁰ molecule·cm^-3·s）高于运城（4.1×10¹⁰ molecule·cm^-3·s），表明两城市的光化学反应存在差异。此外，WRF-Chem的速率分析结果显示，北京的最大O₃净化学生成速率是运城的2.5倍，也证明了光化学反应并非影响两城市O₃污染水平的唯一因素。综合观测结果和模型模拟，我们提出了两城市的O₃污染生成机制。在运城，较高的夜间O₃背景浓度和较低的O₃净生成速率（6.2 ppbv/h）的特点使得其与具有较高的O₃净生成速率（8.4 ppbv/h）和较低的夜间O₃背景浓度的北京在日间O₃峰值上达到相似水平。减排情景模拟结果显示，活性VOCs的减排效果与所有人为源VOCs的减排效果相似，此外，根据模型结果，北京和运城应分别采取VOCs控制和NO_x控制策略，以实现对于O₃的有效管控。

（3）对比2018-2020年暖季（4月-9月）O₃污染天和优良天在“2+26”城市和汾渭平原的气象因素差异，发现O₃污染天相较于优良天，两区域南部地区的T和地面短波辐射（surface net short-wave radiation flux, SSR）显著升高，且O₃污染天主导风向（Wind Direction, WD）为东南风。O₃污染天的散度呈现“辐散-辐合-辐散”的垂直分布，这种分布特征导致近地面大气上升运动和高空大气下沉运动，相反的垂直运动有利于O₃污染在高空积累。分时段探究物理化学过程对日间O₃的影响，发现在O₃浓度上升时段（6:00-14:00），O₃污染天的光化学反应速率显著升高，而在O₃浓度下降时段（14:00-18:00），O₃污染天的O₃消耗能力相对较弱，进一步加剧了O₃的积累。总结2018-2020年暖季两区域O₃污染事件中的O₃传输特征，发现存在两条从“2+26”城市向汾渭平原的O₃传输通道，分别为焦作-晋城-临汾和郑州-洛阳-运城，证明了区域间联防联控的必要性。

（4）聚焦2018-2020年6月份两区域发生的三次同步、大范围、长时间O₃污染事件，探究O₃污染的演变机制。发现在O₃污染发生期，“2+26”城市南部区域和汾渭平原表现出高T、SSR以及低地面WS的特征，这些气象条件可加速VOC的OH消耗速率，从而促使初期O₃污染的形成。在O₃污染持续期，主导气象因素为两区域南部地区的高T和低RH。然而，O₃浓度与地面10 m处的v向风速（10m v-component of wind, V10）的相关关系在两区域呈现一定的差异。在“2+26”城市，O₃与V10呈正相关，而在汾渭平原的西安、咸阳和渭南等地则呈负相关，表明风速对O₃浓度影响较复杂。此外，在三次O₃污染事件中，大气垂直运动特征有助于O₃污染在近地面高空的积累。进一步通过对比O₃污染持续期和发生期的物理化学作用的变化，发现在O₃污染持续期，地面的垂直混合（Vertical Mixing, VMIX）作用增加，而近地面高空的VMIX作用减弱，表明通过湍流混合作用，高空积累的O₃可以被输送回地表，加剧地表的O₃污染。为了有效应对持续的O₃污染，建议“2+26”城市和汾渭平原的西安地区采取针对VOCs的同步联防联控措施，以有效降低两区域的日间O₃浓度。在O₃污染清除期，尽管O₃生成速率没有显著降低，但T的降低和RH的增加提高了O₃消耗速率，从而降低了O₃净化学生成速率，减轻了地表O₃污染。

Recently, severe summertime ozone (O₃) pollution has swept across most areas of China, especially in the two key regions of the Beijing-Tianjin-Hebei and surrounding areas (referred to as “2+26” cities) and Fenwei Plain. However, previous studies on the causes of O₃ pollution in these two regions mainly focused on one of the areas. The “2+26” cities and the Fenwei Plain, two adjacent regions, frequently experience synchronous and large-scale O₃ pollution events. However, research on the causes of such pollution events remains relatively limited. In this study, we used online observations and WRF-Chem simulations, combined with satellite data and meteorological reanalysis data, to reveal the causes of O₃ pollution at both urban and regional scales in these two regions. The main conclusions of this study are as follows:

(1) In 2020, the concentration of the total volatile organic compounds (TVOCs) in Beijing was 23.4 ppbv, lower than that in Yuncheng (38.3 ppbv). The annual NO_x concentrations in the two cities were comparable, with Beijing at 24.5 ppbv and Yuncheng at 24.3 ppbv. The differences in meteorological factors were mainly due to the fact that Yuncheng had higher temperature (T) and relative humidity (RH) than Beijing. However, daytime wind speed (WS) in Beijing was higher than that in Yuncheng, while nighttime WS was lower than that in Yuncheng. The source apportionment analysis of volatile organic compounds (VOCs) revealed that in July 2020, the largest contribution to VOCs in both cities was from vehicle emissions. Based on the self-developed method, it was found that in October, regional transport contributed the highest levels of VOCs to Beijing and Yuncheng, accounting for 37.6% and 50.4%, respectively. Specifically, the southwest direction contributed the most to VOCs transport in Beijing, while for Yuncheng, the highest contribution came from the west. Both the observed VOCs/NO_x ratio and the satellite-based HCHO/NO₂ ratio indicated differences in O₃ sensitivity between the two cities in July.

(2) In July 2020, the concentrations of VOCs and NO_x in Beijing were higher than those in Yuncheng. There were significant differences in the diurnal variation of OVOCs between the two cities. The results obtained using observed data indicated that the OH exposure in Beijing (6.3×10¹⁰ molecule·cm^-3·s) was higher than that in Yuncheng (4.1×10¹⁰ molecule·cm^-3·s), indicating there are differences in photochemical reactions between the two cities. The rate analysis of the model showed that the maximum O₃ net chemical production rate in Beijing was about 2.5 times higher than that in Yuncheng, further proving that photochemical reactions are not the only factor contributing to the level of O₃ pollution in the two cities. Based on the comprehensive analysis of observation results and model simulations, we propose the O₃ pollution formation mechanisms in the two cities. In Yuncheng, the combination of higher nighttime O₃ background concentration and lower net O₃ generation rate (6.2 ppbv/h) leads to similar daytime O₃ peak levels compared to Beijing, which has a higher net O₃ generation rate (8.4 ppbv/h) and lower nighttime O₃ background concentration. The simulation results of emission reduction scenarios indicated that the reduction in reactive VOC emissions had a similar effect to the reduction of all anthropogenic VOC emissions. Additionally, according to the model results, Beijing and Yuncheng should adopt VOC control and NO_x control strategies, respectively, to achieve effective control of O₃ pollution.

(3) By comparing the meteorological factors on the O₃-polluted days and O₃-clean-days in the warm seasons from 2018 to 2020, it was found that higher T and SSR were witnessed in the southern regions of the two areas on the O₃-polluted days, and the dominant wind direction on the O₃-polluted days was southeast. In the O₃-polluted days, the divergence exhibited a vertical distribution pattern of “divergence-convergence-divergence”. This distribution characteristic led to the upward movement of the near-surface atmosphere and the downward trend of the upper atmosphere. These opposite vertical atmospheric movements contributed to the accumulation of O₃ pollution at higher altitudes. By investigating the physical and chemical processes during different time periods, it was found that during the O₃-rising period (6:00-14:00), the photochemical reaction rates significantly increased, while during the O₃-descending period (14:00-18:00), O₃ consumption rates were relatively weak, resulting in intensified O₃ accumulation on the O₃-polluted days. Based on the transmission characteristics of O₃ between the two regions, two O₃ transmission channels from the “2+26” cities to the Fenwei Plain, namely the Jiaozuo-Jincheng-Linfen and Zhengzhou-Luoyang-Yuncheng channels, were proposed for the first time, demonstrating the necessity of joint prevention and control of O₃ between the two regions.

(4) By focusing on three synchronized, long-term and large-scale O₃ pollution events in the “2+26” cities and Fenwei Plain in June from 2018 to 2020, we explored the mechanisms of O₃ pollution evolution. The results showed that during the occurrence stage of O₃ pollution, the southern regions of “2+26” cities and the Fenwei Plain exhibited high T and SSR, as well as low WS. These meteorological conditions accelerated the OH consumption rate of VOCs, thereby promoting the formation of initial O₃ pollution. During the persistence stage of O₃ pollution, the dominant meteorological factors were mainly high T and low RH in the southern areas of the two regions. Notably, we found that the relationships between O₃ and the 10m v-component of wind (V10) were significantly different between the two regions. In the “2+26” cities, O₃ was positively correlated with V10, while in the cities such as Xi’an, Xianyang, and Weinan in the Fenwei Plain, it was negatively correlated, indicating a more complex influence of wind speed on O₃ concentration. In addition, the atmospheric vertical motion characteristics during the three O₃ pollution events contributed to the accumulation of O₃ at higher altitudes. Furthermore, it was found that during the persistence stage of O₃ pollution, vertical mixing (VMIX) near the surface increased while VMIX in the upper levels decreased compared to the occurrence stage of O₃ pollution. This indicated that the O₃ accumulated in the higher altitudes could be transported back to the surface through turbulent mixing, exacerbating surface O₃ pollution. To effectively tackle the persistent O₃ pollution, it is recommended that the “2+26” cities and Xi’an in the Fenwei Plain should implement coordinated prevention and control measures targeting VOCs to effectively reduce daytime O₃ concentrations in both regions. During the clearance stage of O₃ pollution, although the O₃ generation rate did not decrease, the decrease in T and the increase in RH increased the O₃ consumption rate, resulting in a decrease in the net chemical generation rate of O₃ and further a reduction in surface O₃ concentration.