Scholarlink

<p>Rapid urbanization and climate change create significant challenges for land cover dynamics and demographic patterns in growing cities. This study employs advanced geospatial methods and machine learning algorithms to examine the complex relationships between seasonal land surface temperature (LST) variations and land use/land cover (LULC) changes in urban and rural areas. Using Landsat satellite imagery from 1998 to 2022, we quantified how urban growth, and environmental changes affect seasonal temperature patterns in Dhaka, Bangladesh. Our analysis reveals dramatic landscape transformations: 98% increase in urban built-up areas, accompanied by substantial losses of vegetation (48%) and water bodies (19%). These changes have intensified urban heat island effects and altered local climate patterns. Rural areas experienced similarly rapid development, with a 120% increase in built-up zones and an estimated 30% decrease in water bodies. Our seasonal analysis demonstrates significant temperature increases during summer months, exceeding 36.8&#xa0;°C, directly linked to LULC changes, providing deeper insights into land cover-temperature relationships. Geographically Weighted Regression (GWR) analysis revealed that urban development positively correlates with rising temperatures, particularly in densely built areas, while vegetated regions show negative correlations with temperature. These relationships vary spatially, highlighting the importance of accounting for spatial heterogeneity in urban climate studies. The GWR model significantly outperformed ordinary least squares regression, with R² improving from 0.578 to 0.901, demonstrating the value of spatially explicit modeling approaches. This research enhances our understanding of urbanization’s environmental impacts and provides valuable insights for policymakers and urban planners. The findings support the development of climate-resilient urban and rural environments capable of addressing ongoing global challenges including climate change and rapid urbanization. Our methodology offers a replicable framework for analyzing urban climate dynamics in other rapidly expanding cities worldwide.</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

Geostatistical assessment of Spatial climate dynamics using mono window machine learning algorithm for decoding land cover and demographic shifts influence on thermal environment

  • 期刊论文

  • 发表日期：

    2025年9月30日

查看全文
  • Md Tanvir Miah,
  • Jannatun Nahar Fariha,
  • Abdulla Al Kafy,
  • Mirza Md Tasnim Mukarram,
  • Hamad Ahmed Altuwaijri,
  • Pankaj Kanti Jodder,
  • Niloy Biswas,
  • Opelele Omeno Michel,
  • Md Mahmudul Hasan

摘要

Rapid urbanization and climate change create significant challenges for land cover dynamics and demographic patterns in growing cities. This study employs advanced geospatial methods and machine learning algorithms to examine the complex relationships between seasonal land surface temperature (LST) variations and land use/land cover (LULC) changes in urban and rural areas. Using Landsat satellite imagery from 1998 to 2022, we quantified how urban growth, and environmental changes affect seasonal temperature patterns in Dhaka, Bangladesh. Our analysis reveals dramatic landscape transformations: 98% increase in urban built-up areas, accompanied by substantial losses of vegetation (48%) and water bodies (19%). These changes have intensified urban heat island effects and altered local climate patterns. Rural areas experienced similarly rapid development, with a 120% increase in built-up zones and an estimated 30% decrease in water bodies. Our seasonal analysis demonstrates significant temperature increases during summer months, exceeding 36.8 °C, directly linked to LULC changes, providing deeper insights into land cover-temperature relationships. Geographically Weighted Regression (GWR) analysis revealed that urban development positively correlates with rising temperatures, particularly in densely built areas, while vegetated regions show negative correlations with temperature. These relationships vary spatially, highlighting the importance of accounting for spatial heterogeneity in urban climate studies. The GWR model significantly outperformed ordinary least squares regression, with R² improving from 0.578 to 0.901, demonstrating the value of spatially explicit modeling approaches. This research enhances our understanding of urbanization’s environmental impacts and provides valuable insights for policymakers and urban planners. The findings support the development of climate-resilient urban and rural environments capable of addressing ongoing global challenges including climate change and rapid urbanization. Our methodology offers a replicable framework for analyzing urban climate dynamics in other rapidly expanding cities worldwide.