Scholarlink

<p>With rapid urbanization and socioeconomic development, residents’ mobility demand has surged, leading to increased private car use and associated emissions. This study analyzes emissions from private vehicles across 51 Japanese cities and examines how urban form, climate, topography, land use, socioeconomic factors, and public transport infrastructure jointly shape emission patterns. We develop an integrated framework combining regression and machine learning models to evaluate predictive performance and interpretability. The Random Forest model achieves the highest accuracy, and railway track and station densities are associated with lower emissions, while complex land use patterns and higher elevations are associated with higher emissions. We further use clustering to classify cities into five clusters, providing a foundation for tailored mitigation strategies. A nonmonotonic association between elderly population share and emissions is also identified, with slight increases at moderate levels followed by declines in highly aged cities, which reflects how demographic and infrastructural factors interact. Together, these findings highlight actionable directions for cities to transition to more sustainable, low carbon mobility and lifestyles.</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

Exploring carbon emission patterns from private vehicle use in Japanese cities

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年7月3日

查看全文
  • Liqiao Huang,
  • Xiaoyan Xu,
  • Zhiheng Chen,
  • Kenji Yamada,
  • Yoshikuni Yoshida,
  • Yin Long

摘要

With rapid urbanization and socioeconomic development, residents’ mobility demand has surged, leading to increased private car use and associated emissions. This study analyzes emissions from private vehicles across 51 Japanese cities and examines how urban form, climate, topography, land use, socioeconomic factors, and public transport infrastructure jointly shape emission patterns. We develop an integrated framework combining regression and machine learning models to evaluate predictive performance and interpretability. The Random Forest model achieves the highest accuracy, and railway track and station densities are associated with lower emissions, while complex land use patterns and higher elevations are associated with higher emissions. We further use clustering to classify cities into five clusters, providing a foundation for tailored mitigation strategies. A nonmonotonic association between elderly population share and emissions is also identified, with slight increases at moderate levels followed by declines in highly aged cities, which reflects how demographic and infrastructural factors interact. Together, these findings highlight actionable directions for cities to transition to more sustainable, low carbon mobility and lifestyles.