2017—2019年太原南部城区夏季O3特征及其影响因子

利用2017—2019年夏季（6—8月）太原市污染物浓度和气象逐小时数据，分析了太原南部城区O₃浓度及其影响因子的变化特征，通过神经网络构建了O₃与其影响因子的关系模型，并进行了检验。结果表明：2017—2019年夏季太原南部城区O₃浓度超标天数分别为55 d、39 d、59 d，超标主要集中在6月和7月；O₃浓度日变化特征呈单峰型，每日06时前后达到最低，15时前后达到峰值。高温、强辐射、低湿、低压、西南风容易引起太原南部城区O₃浓度升高，西北风有利于O₃扩散；NO₂、CO与O₃浓度表现为负相关关系，但NO₂对其影响更加显著。利用神经网络构建O₃浓度与影响因子的关系模型，相关系数达0.96，均方根误差、平均绝对误差分别为8 μg·m^-3、6%，TS评分为0.95，神经网络模型具有较高预报能力，可为太原地区O₃预报提供参考价值。

Analysis of the characteristics of O3 concentration and its influencing factors in summer from 2017 to 2019 in the southern urban area of Taiyuan

Based on the hourly data of pollutant concentration and relative meteorological factors during summer (June to August) from 2017-2019, the distribution characteristics of O₃ concentration and its influencing factors in Taiyuan were analyzed using the neural network method.The results show that, during the summer from 2017-2019, the numbers of days in which the O₃ concentration exceeds the limit in Taiyuan are 55 d, 39 d, 59 d, respectively.The cases that the O₃ concentration exceeds the limit mostly occur in June and July.The diurnal variation of O₃ concentration is unimodal, with the lowest around 06:00, and the peak around 15:00.The conditions such as high temperature, strong radiation, low humidity, low pressure, and southwest wind can easily lead to an increase of O₃ concentration in Taiyuan urban areas.The NW wind is beneficial to the diffusion of O₃ concentration.The relationships of NO₂ and CO with O₃ concentration are negative, and the influence of NO₂ is more significant.The selected cases show that the O₃ concentration fluctuates with the influencing factors.The relationship between O₃ concentration and the influencing factors is constructed using the neural network method, with the correlation coefficient of 0.96, the mean square root and average absolute error of 8 μg·m^-3 and 6%, respectively, and TS score of 0.95.The neural network model is valuable for the O₃ concentration prediction and ozone pollution control in the Taiyuan area.