基于随机森林的北京城区臭氧敏感性分析

明确臭氧（O₃）与前体物的非线性关系是O₃防控措施制定的基础和关键.基于北京城区站点2020年4~9月O₃、挥发性有机物（VOCs）、氮氧化物（NO_x）和气象要素在线观测,分析了O₃及其前体物污染特征,利用随机森林（RF）模型结合SHAP值探究了影响O₃的关键因素,并通过多情景分析探讨了O₃-VOCs-NO_x敏感性.相关性分析结果显示O₃小时浓度与温度（T）呈显著正相关,与TVOCs和NO_x呈显著负相关；但从每日结果来看,O₃与T、TVOCs和NO_x均呈显著正相关.RF模型模拟的O₃浓度与实测值吻合较好,进一步计算了各个特征变量的SHAP值,结果显示T和NO_x对O₃影响最高,但前者是正向影响,而后者是负向影响.以观测期间O₃污染天的NO_x和VOCs平均值为基础情景,设置了对应不同NO_x和VOCs的多种情景,并利用RF模型计算不同情景下的O₃,得到O₃等值线（EKMA曲线）,结果显示北京城区O₃-VOCs-NO_x敏感性处于VOCs控制区,与基于观测的盒子模型（OBM）得到的结果一致,这说明RF模型可以用作O₃-VOCs-NO_x敏感性分析的一种补充方法.

Ozone Sensitivity Analysis in Urban Beijing Based on Random Forest

The basis and key step to developing ozone (O₃) prevention and control measures is determining the non-linear relationship between O₃ and its precursors. Based on online observations of O₃, volatile organic compounds (VOCs), nitrogen oxides (NO_x), and meteorological elements from April to September 2020 at an urban site in Beijing, we analyzed the pollution characteristics of O₃ and its precursors, explored key factors affecting O₃ using the random forest (RF) model combined with SHAP values, and explored the O₃-VOCs-NO_x sensitivity through a multi-scenarios analysis. The results of correlation analysis showed that the hourly concentration of O₃ was significantly positively correlated with temperature (T) and negatively correlated with TVOCs and NO_x. However, in terms of the daily values, O₃ was significantly positively correlated with T, TVOCs, and NO_x. The simulated O₃ values by the RF model agreed with the measured values. The SHAP values of each characteristic variable were further calculated. The results suggested that T and NO_x showed the two highest effects on O₃, with positive and negative values, respectively. Based on the average NO_x and VOCs on O₃ pollution days during the observation period (the base scenario), multi-scenarios with different NO_x and VOCs were set up. The RF model was used to calculateO₃ under different scenarios and obtain the O₃ isopleth (EKMA curve). The results showed that the O₃-VOCs-NO_x sensitivity in urban areas of Beijing was in the VOCs-limited regime, which was consistent with the results obtained from the observation-based box model(OBM). This indicated that the RF model could be used as a complementary method for O₃-VOCs-NO_x sensitivity analysis.