Scholarlink

<p>Rapid urbanization and rising urban demands, driven by unprecedented population growth and migration, have significantly altered Land Use/Land Cover (LULC) patterns in many cities, leading to serious environmental consequences. Previous studies indicate that such transformations contribute to both short- and long-term impacts, particularly rising Land Surface Temperature (LST) and the intensification of the Urban Heat Island (UHI) effect; however, the relationship between LULC dynamics and LST requires deeper investigation. This study explores the linkage between LULC changes and their impact on LST in Moradabad District, Uttar Pradesh, India, over a 25-year period (1998, 2013, and 2023). LULC classification was conducted using supervised methods and further improved by integrating spectral indices with selected image bands, thereby enhancing classification accuracy and reducing mixed-pixel effects. To minimize data redundancy, optimal band combinations were identified using the Optimum Index Factor (OIF). Analysis of Landsat TM and OLI imagery revealed a substantial decline in vegetation cover and a doubling of built-up areas between 1998 and 2023. This transformation corresponded to increases in the minimum temperature from 18.83&#xa0;°C to 23.42&#xa0;°C and in the maximum temperature from 29.18&#xa0;°C to 34.05&#xa0;°C. The spatial distribution of LST was assessed using Moran’s I, which indicated moderately clustered temperature patterns. Key indices NDVI, MNDWI, and NDBI were computed to evaluate their relationship with LST, followed by regression analysis using Ordinary Least Squares (OLS) and Geographically Weighted Regression (GWR). The results demonstrated that GWR (R² = 0.80, 0.78, and 0.87) outperformed OLS (R² = 0.23, 0.27, and 0.36) for 1998, 2013, and 2023, respectively, highlighting the need for sustainable urban planning strategies that preserve vegetation to mitigate future increases in LST.</p> Graphical Abstract <p></p> <p>This study outlines the methodological framework and principal conclusions on land-use–land-cover (LULC) dynamics and urban thermal fluctuations. Initially, False Color Composite (FCC) imagery and several Land Use/Land Cover (LULC) indices were amalgamated, and the optimal spectral band combination was determined through the Optimum Index Factor (OIF) by assessing all conceivable band combinations. LULC maps were produced using optimal bands via machine-learning classification, with accuracy validated against ground truth from Google Earth Pro images and field surveys. The findings reveal substantial urban growth in the study area from 1998 to 2023, with vegetation cover decreasing from 45.45% to 30.57% and built-up areas increasing from 20.21% to 40.48%. The land modifications were associated with significant thermal alterations, as the mean Land Surface Temperature (LST) rose by roughly 3.8&#xa0;°C and Urban Heat Island (UHI) intensity increased by around 1.41&#xa0;°C, indicating heightened urban thermal stress. Furthermore, regression analysis was conducted between land surface temperature (LST) and various spectral indices, including NDVI, NDBI, BCI, and Tasseled Cap indices, to investigate correlations between land cover attributes and surface temperature trends. The spatial modeling results indicated that Geographically Weighted Regression (GWR) outperformed Ordinary Least Squares (OLS) in elucidating the spatial variability of Land Surface Temperature (LST), indicating a robust spatial dependence among urban heat drivers. The OLS values were − 0.23 (1998), 0.27 (2013), and 0.36 (2023), whereas the GWR values increased to − 0.80, 0.78, and 0.87, respectively.</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 Multi-Index Framework for Improved Land Use/Land Cover Mapping and Its Implications for Urban Heat Island Dynamics

  • 期刊论文

  • 发表日期：

    2026年7月6日

查看全文
  • Vishva Deep Singh,
  • Ashish Simalti,
  • Atul Kant Piyoosh,
  • Aqil Tariq

摘要

Rapid urbanization and rising urban demands, driven by unprecedented population growth and migration, have significantly altered Land Use/Land Cover (LULC) patterns in many cities, leading to serious environmental consequences. Previous studies indicate that such transformations contribute to both short- and long-term impacts, particularly rising Land Surface Temperature (LST) and the intensification of the Urban Heat Island (UHI) effect; however, the relationship between LULC dynamics and LST requires deeper investigation. This study explores the linkage between LULC changes and their impact on LST in Moradabad District, Uttar Pradesh, India, over a 25-year period (1998, 2013, and 2023). LULC classification was conducted using supervised methods and further improved by integrating spectral indices with selected image bands, thereby enhancing classification accuracy and reducing mixed-pixel effects. To minimize data redundancy, optimal band combinations were identified using the Optimum Index Factor (OIF). Analysis of Landsat TM and OLI imagery revealed a substantial decline in vegetation cover and a doubling of built-up areas between 1998 and 2023. This transformation corresponded to increases in the minimum temperature from 18.83 °C to 23.42 °C and in the maximum temperature from 29.18 °C to 34.05 °C. The spatial distribution of LST was assessed using Moran’s I, which indicated moderately clustered temperature patterns. Key indices NDVI, MNDWI, and NDBI were computed to evaluate their relationship with LST, followed by regression analysis using Ordinary Least Squares (OLS) and Geographically Weighted Regression (GWR). The results demonstrated that GWR (R² = 0.80, 0.78, and 0.87) outperformed OLS (R² = 0.23, 0.27, and 0.36) for 1998, 2013, and 2023, respectively, highlighting the need for sustainable urban planning strategies that preserve vegetation to mitigate future increases in LST.

Graphical Abstract

This study outlines the methodological framework and principal conclusions on land-use–land-cover (LULC) dynamics and urban thermal fluctuations. Initially, False Color Composite (FCC) imagery and several Land Use/Land Cover (LULC) indices were amalgamated, and the optimal spectral band combination was determined through the Optimum Index Factor (OIF) by assessing all conceivable band combinations. LULC maps were produced using optimal bands via machine-learning classification, with accuracy validated against ground truth from Google Earth Pro images and field surveys. The findings reveal substantial urban growth in the study area from 1998 to 2023, with vegetation cover decreasing from 45.45% to 30.57% and built-up areas increasing from 20.21% to 40.48%. The land modifications were associated with significant thermal alterations, as the mean Land Surface Temperature (LST) rose by roughly 3.8 °C and Urban Heat Island (UHI) intensity increased by around 1.41 °C, indicating heightened urban thermal stress. Furthermore, regression analysis was conducted between land surface temperature (LST) and various spectral indices, including NDVI, NDBI, BCI, and Tasseled Cap indices, to investigate correlations between land cover attributes and surface temperature trends. The spatial modeling results indicated that Geographically Weighted Regression (GWR) outperformed Ordinary Least Squares (OLS) in elucidating the spatial variability of Land Surface Temperature (LST), indicating a robust spatial dependence among urban heat drivers. The OLS values were − 0.23 (1998), 0.27 (2013), and 0.36 (2023), whereas the GWR values increased to − 0.80, 0.78, and 0.87, respectively.