Scholarlink

<p>Rapid urbanization is a primary driver of localized climate change and the intensification of land surface temperature (LST) in growing cities. This study aims to assess the spatiotemporal impacts of urban expansion on LST dynamics and the formation of the urban heat island (UHI) effect in the Isfahan metropolis from 2013 to 2024, offering a novel integration of multi-source data and advanced analytical methods. The dataset includes Landsat 8, Sentinel 2, and MODIS imagery, 25,166 temperature points, 37,380 thermal records, monthly meteorological station data (1981 to 2024), a digital elevation model (DEM), and vegetation, surface moisture, and radiation indices. Analytical methods employed include supervised classification, artificial neural networks, time series analysis, machine learning algorithms, and spatial statistical tools such as Moran’s I, Getis Ord Gi*, and Ripley’s K function. Findings reveal a 163% increase in built-up areas, a 75% reduction in dense vegetation, a 22% decline in surrounding agricultural lands, and a 16% rise in barren land. These changes significantly elevated LST and triggered dense thermal clusters. Maximum UHI intensities peaked at 44.2&#xa0;°C during the summer months. NDVI showed a notable seasonal decline, especially in July, while spatial correlations exceeded 0.91, underscoring a strong link between urban growth and thermal anomalies. The results highlight the urgent need for climate-sensitive urban planning. Practical recommendations include expanding urban green spaces, mitigating surface sealing, managing urban sprawl, and investing in urban cooling infrastructure to reduce heat stress and improve environmental resilience.</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

A hybrid machine learning approach to analyzing the impacts of urban development on land surface temperature and the urban heat island effect in Isfahan

  • 期刊论文

  • 发表日期：

    2025年9月17日

查看全文
  • Majid Javari

摘要

Rapid urbanization is a primary driver of localized climate change and the intensification of land surface temperature (LST) in growing cities. This study aims to assess the spatiotemporal impacts of urban expansion on LST dynamics and the formation of the urban heat island (UHI) effect in the Isfahan metropolis from 2013 to 2024, offering a novel integration of multi-source data and advanced analytical methods. The dataset includes Landsat 8, Sentinel 2, and MODIS imagery, 25,166 temperature points, 37,380 thermal records, monthly meteorological station data (1981 to 2024), a digital elevation model (DEM), and vegetation, surface moisture, and radiation indices. Analytical methods employed include supervised classification, artificial neural networks, time series analysis, machine learning algorithms, and spatial statistical tools such as Moran’s I, Getis Ord Gi*, and Ripley’s K function. Findings reveal a 163% increase in built-up areas, a 75% reduction in dense vegetation, a 22% decline in surrounding agricultural lands, and a 16% rise in barren land. These changes significantly elevated LST and triggered dense thermal clusters. Maximum UHI intensities peaked at 44.2 °C during the summer months. NDVI showed a notable seasonal decline, especially in July, while spatial correlations exceeded 0.91, underscoring a strong link between urban growth and thermal anomalies. The results highlight the urgent need for climate-sensitive urban planning. Practical recommendations include expanding urban green spaces, mitigating surface sealing, managing urban sprawl, and investing in urban cooling infrastructure to reduce heat stress and improve environmental resilience.