Scholarlink

<p>Ground-level ozone (O<sub>3</sub>) is a widespread global threat to human health and ecosystems, with substantial impacts in China. This study investigated the spatiotemporal patterns and driving mechanisms of O₃ pollution in the Middle Reaches of the Yangtze River Urban Agglomeration (MRYR) from 2015 to 2024. Utilizing a Light Gradient Boosting Machine (LightGBM) framework, we performed meteorological normalization to disentangle contributions of emissions and meteorology to daily maximum 8-hour average (O<sub>3</sub>-8&#xa0;h) concentrations; emission influences were inferred from observed concentrations using ambient NO<sub>2</sub> and CO concentrations together with temporal and spatial variables (year, month, day, latitude, and longitude) as integrated proxy features for precursor emissions. SHapley Additive exPlanations (SHAP), perturbation analysis, and the control variable method were employed to quantify individual meteorological impacts, while Geographically Weighted Regression (GWR) was applied to elucidate the spatially heterogeneous drivers of O<sub>3</sub>-8&#xa0;h. Results show O<sub>3</sub>-8&#xa0;h increased from 37.4 ppb in 2015 to 48.8 ppb in 2024, with higher levels in eastern and central urban cores. Meteorological normalization indicated that meteorology contributed 70.3% of interannual variability (2015–2023) versus 29.7% from emissions; emission contributions rose to 45.3% in 2024, signaling a shifting driver regime. SHAP analysis identified 2-meter temperature and surface solar radiation as primary promoters, while dewpoint and soil temperature inhibited formation. Given its ability to capture nonlinear interactions, SHAP is recommended over perturbation or control variable methods for future diagnostic applications. This integrated approach clarifies evolving O₃ drivers, supporting targeted mitigation in rapidly developing urban agglomerations.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Meteorological and emission drivers of ozone in the Middle Reaches of the Yangtze River urban agglomeration, China

  • 期刊论文

  • 发表日期：

    2026年7月3日

查看全文
  • Xiaoyong Liu,
  • Liuming Wang,
  • Junhui Yan,
  • Yi Shen,
  • Xiangchao Cui

摘要

Ground-level ozone (O3) is a widespread global threat to human health and ecosystems, with substantial impacts in China. This study investigated the spatiotemporal patterns and driving mechanisms of O₃ pollution in the Middle Reaches of the Yangtze River Urban Agglomeration (MRYR) from 2015 to 2024. Utilizing a Light Gradient Boosting Machine (LightGBM) framework, we performed meteorological normalization to disentangle contributions of emissions and meteorology to daily maximum 8-hour average (O3-8 h) concentrations; emission influences were inferred from observed concentrations using ambient NO2 and CO concentrations together with temporal and spatial variables (year, month, day, latitude, and longitude) as integrated proxy features for precursor emissions. SHapley Additive exPlanations (SHAP), perturbation analysis, and the control variable method were employed to quantify individual meteorological impacts, while Geographically Weighted Regression (GWR) was applied to elucidate the spatially heterogeneous drivers of O3-8 h. Results show O3-8 h increased from 37.4 ppb in 2015 to 48.8 ppb in 2024, with higher levels in eastern and central urban cores. Meteorological normalization indicated that meteorology contributed 70.3% of interannual variability (2015–2023) versus 29.7% from emissions; emission contributions rose to 45.3% in 2024, signaling a shifting driver regime. SHAP analysis identified 2-meter temperature and surface solar radiation as primary promoters, while dewpoint and soil temperature inhibited formation. Given its ability to capture nonlinear interactions, SHAP is recommended over perturbation or control variable methods for future diagnostic applications. This integrated approach clarifies evolving O₃ drivers, supporting targeted mitigation in rapidly developing urban agglomerations.