Scholarlink

<p>Heatwaves are intensifying globally due to climate change, and while this poses challenges for cities worldwide, urban centers in Sub-Saharan Africa are particularly vulnerable to the Urban Heat Island (UHI) effect. Although a growing body of studies has documented rising land surface temperatures (LST) in the region, analyses that use predictive modeling to relate these trends to land cover dynamics remain limited. As part of efforts to address this limitation, this study applies a comparative machine learning approach to quantify the influence of land cover changes on heatwave patterns in Ibadan, the second-largest city in Africa, between 2001 and 2020. Trends in LST and land cover were analyzed using Moderate Resolution Imaging Spectroradiometer (MODIS) data and Landsat-derived Normalized Difference Vegetation Index (NDVI). The relationship was then modeled using random forest, gradient boosting, and support vector regression algorithms. The results showed a 34.8% increase in built-up area and a proportionate decline in forest cover. Urban areas reached a peak LST of 47.72&#xa0;°C, compared to 39.42&#xa0;°C in more vegetated areas. NDVI values below 0.4 were associated with LST increases of up to 12&#xa0;°C. The random forest algorithm demonstrated the highest predictive accuracy (R² up to 0.86), indicating strong non-linear relationships between vegetation cover and surface temperature. These findings emphasize the important role of urban greenery in heat mitigation and offer a scalable, data-driven approach to help urban planners and policymakers develop strategies to reduce heat exposure.</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

Modeling heatwave trends from land cover dynamics using satellite observations and machine learning in Ibadan, Nigeria

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年12月24日

查看全文
  • Femi Emmanuel Ikuemonisan,
  • Yusuf Olanrewaju Kayode,
  • Samuel Toluwalope Ogunjo,
  • Opeyemi Bilikisu Odubote,
  • Sakiru Abiodun Okedeyi

摘要

Heatwaves are intensifying globally due to climate change, and while this poses challenges for cities worldwide, urban centers in Sub-Saharan Africa are particularly vulnerable to the Urban Heat Island (UHI) effect. Although a growing body of studies has documented rising land surface temperatures (LST) in the region, analyses that use predictive modeling to relate these trends to land cover dynamics remain limited. As part of efforts to address this limitation, this study applies a comparative machine learning approach to quantify the influence of land cover changes on heatwave patterns in Ibadan, the second-largest city in Africa, between 2001 and 2020. Trends in LST and land cover were analyzed using Moderate Resolution Imaging Spectroradiometer (MODIS) data and Landsat-derived Normalized Difference Vegetation Index (NDVI). The relationship was then modeled using random forest, gradient boosting, and support vector regression algorithms. The results showed a 34.8% increase in built-up area and a proportionate decline in forest cover. Urban areas reached a peak LST of 47.72 °C, compared to 39.42 °C in more vegetated areas. NDVI values below 0.4 were associated with LST increases of up to 12 °C. The random forest algorithm demonstrated the highest predictive accuracy (R² up to 0.86), indicating strong non-linear relationships between vegetation cover and surface temperature. These findings emphasize the important role of urban greenery in heat mitigation and offer a scalable, data-driven approach to help urban planners and policymakers develop strategies to reduce heat exposure.