基于响应面法研究植被对城市街谷内汽车尾气暴露的影响

合理的城市植被设计已被证明能改善街道峡谷内的空气状况，但目前缺乏对带有城市植被的街道峡谷内多个因素影响程度的系统研究，文章以减少街道峡谷内的行人以及道路沿线居民的污染暴露为目标，将响应面方法(response surface methodology, RSM)与计算流体动力学(computational fluid dynamics, CFD)结合起来，研究不同因素间的交互影响，并给出街道峡谷高宽比H/W、风向角θ和绿化带叶面积密度 $ D_{LA} $对街道峡谷内汽车尾气扩散的影响程度排序。通过对建立的个人吸入分数(personal intake fraction, P-IF)的二次回归预测模型进行方差分析并评估性能，表明回归预测模型可以很好地预测各因素与目标之间的关系，最小P-IF的最佳条件是 $ H/W=1.50, D_{LA}=0.50\ m^{-1}, \theta=0^{\circ} $。该研究结果可为城市规划者和政策制定者制定城市环境控制策略、减少道路沿线人类的暴露提供依据。

Reasonable urban vegetation design has been proven to improve air quality within street canyons. However, current research lacks a systematic understanding of the extent to which multiple factors affect street canyon environments with urban vegetation. This study combines response surface methodology (RSM) with computational fluid dynamics (CFD) with the goal of reducing pollution exposure for pedestrians within the canyon and residents along the roadway. It investigates the interactive effects of different factors and ranks the influence of canyon aspect ratio $ (H/W) $, wind direction $ (\theta) $, and leaf area density $ (D_{LA}) $ of green belts on vehicular exhaust dispersion within street canyons. Performance evaluation of the quadratic regression predictive model of personal intake fraction $ (P_{\text{-IF}}) $ through analysis of variance indicates that the regression predictive model can effectively predict the relationship between various factors and objectives. The optimal conditions for minimizing $ P_{-IF} $ are found to be an H/W of 1.50, a $ D_{LA} $ of 0.50 m $ ^{-1} $, and a $ \theta $ of 0°. The study results can inform urban planners and policymakers in formulating urban environmental control strategies to reduce human exposure along roadways.