Scholarlink

<p>As urbanization is consolidated, increased land surface temperatures (LST) have also become an important concern for environmental quality and public health. While many studies examine the impacts of land use, vegetation, and built form on urban heat, the structural configuration of road networks remains poorly understood. The present study undertakes exploration of how road-network characteristics in Riyadh, such as intersection density, street connectivity, and the continuity of road segments, shape spatial variations in LST. Using MODIS satellite data from 2015 to 2024 within a 1&#xa0;km² grid, this research analysed eight road types and described their structural forms with intuitive indicators, including average distance between intersections and prevalence of looped over linear street patterns. A Random Forest (RF) model was fitted with SHAP (Shapley Additive Explanations) interpretation to establish which feature impacted the thermal outcomes most distinctly. Results reveal a pronounced thermal gradient across the city: rapidly expanding outer districts, including Al-Manar and Al-Malqa, reached frequently over 42&#xa0;°C, while central neighbourhoods cooled down to near 30&#xa0;°C by 2024, following targeted greening and shading. Across most road types, shorter distances between intersections indicative of dense and more compact street networks were associated with an LST reduction of approximately 1.2–2.5&#xa0;°C. In contrast, districts characterized by long and sparsely connected segments face up to 2.3&#xa0;°C temperature rises. Spatial SHAP mapping has shown that looped and connected layouts dissipate heat compared to peripheral areas dominated by sparse and linear roads. In demonstrating that road-network structure meaningfully shapes urban heat patterns, this paper highlights opportunities for climate-responsive street planning.</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

Road network topology and machine learning integration to model urban heat patterns in Riyadh

  • 期刊论文

  • 发表日期：

    2025年12月8日

查看全文
  • Saeed Alqadhi,
  • Javed Mallick

摘要

As urbanization is consolidated, increased land surface temperatures (LST) have also become an important concern for environmental quality and public health. While many studies examine the impacts of land use, vegetation, and built form on urban heat, the structural configuration of road networks remains poorly understood. The present study undertakes exploration of how road-network characteristics in Riyadh, such as intersection density, street connectivity, and the continuity of road segments, shape spatial variations in LST. Using MODIS satellite data from 2015 to 2024 within a 1 km² grid, this research analysed eight road types and described their structural forms with intuitive indicators, including average distance between intersections and prevalence of looped over linear street patterns. A Random Forest (RF) model was fitted with SHAP (Shapley Additive Explanations) interpretation to establish which feature impacted the thermal outcomes most distinctly. Results reveal a pronounced thermal gradient across the city: rapidly expanding outer districts, including Al-Manar and Al-Malqa, reached frequently over 42 °C, while central neighbourhoods cooled down to near 30 °C by 2024, following targeted greening and shading. Across most road types, shorter distances between intersections indicative of dense and more compact street networks were associated with an LST reduction of approximately 1.2–2.5 °C. In contrast, districts characterized by long and sparsely connected segments face up to 2.3 °C temperature rises. Spatial SHAP mapping has shown that looped and connected layouts dissipate heat compared to peripheral areas dominated by sparse and linear roads. In demonstrating that road-network structure meaningfully shapes urban heat patterns, this paper highlights opportunities for climate-responsive street planning.