Scholarlink

<p>Estimating vehicle emissions across large urban road networks requires extensive data, particularly traffic flow data, which is essential for reliable emission estimation for the entire vehicle population. Due to the sparse distribution of data collection devices, traffic flow data is often unavailable for some roads, leading to inaccurate emission estimates. To address this, this study introduces a machine learning-based model to infer unobserved traffic flow using auxiliary urban data like population, points of interests (POIs), and taxi GPS data. After that, the complete traffic flow data are integrated with the localized Motor Vehicle Emission Simulator (MOVES) to estimate road-level emissions of carbon monoxide (CO), nitrogen oxides (NOx), volatile organic compounds (VOCs), and fine particulate matter (PM<sub>2.5</sub>). A Case study from the core areas of Guangzhou, China, reveasl significant spatial and temporal disparities in emissions. CO emissions are evenly distributed across the roads, while NOx and VOCs are concentrated on expressways. CO and PM<sub>2.5</sub> emissions peak in the morning, while NOx and VOCs peak in the evening. All four types of emissions follow power-law distributions, with a small number of heavily polluted roads accounting for most emissions, and NOx and VOCs showing the greatest spatial disparity. These findings provide a detailed measurement of large-scale urban traffic emissions and offer actionable insights for urban environmental protection and sustainable development strategies.</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

Estimating Large-Scale Urban Traffic Emissions with Incomplete Covered Traffic Flow Data—A Case of Core Areas of Guangzhou, China

  • 期刊论文

  • 发表日期：

    2025年5月5日

查看全文
  • Keyu Lu,
  • Hui Meng,
  • Xinhang Liu,
  • Dan Zou,
  • Qiuping Li

摘要

Estimating vehicle emissions across large urban road networks requires extensive data, particularly traffic flow data, which is essential for reliable emission estimation for the entire vehicle population. Due to the sparse distribution of data collection devices, traffic flow data is often unavailable for some roads, leading to inaccurate emission estimates. To address this, this study introduces a machine learning-based model to infer unobserved traffic flow using auxiliary urban data like population, points of interests (POIs), and taxi GPS data. After that, the complete traffic flow data are integrated with the localized Motor Vehicle Emission Simulator (MOVES) to estimate road-level emissions of carbon monoxide (CO), nitrogen oxides (NOx), volatile organic compounds (VOCs), and fine particulate matter (PM2.5). A Case study from the core areas of Guangzhou, China, reveasl significant spatial and temporal disparities in emissions. CO emissions are evenly distributed across the roads, while NOx and VOCs are concentrated on expressways. CO and PM2.5 emissions peak in the morning, while NOx and VOCs peak in the evening. All four types of emissions follow power-law distributions, with a small number of heavily polluted roads accounting for most emissions, and NOx and VOCs showing the greatest spatial disparity. These findings provide a detailed measurement of large-scale urban traffic emissions and offer actionable insights for urban environmental protection and sustainable development strategies.