Gradient-explicit assessment of surface urban heat island intensity in subtropical megacities
Assessments of surface urban heat island intensity (SUHII) in subtropical megaregions often rely on inconsistent urban-rural definitions, limiting comparability across cities and over time. Here we propose a gradient-explicit framework that combines Local Climate Zone (LCZ)-informed urban-rural area...
Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article (Journal) |
| Language: | English |
| Published: |
1 June 2026
|
| In: |
Sustainable cities and society
Year: 2026, Volume: 143, Pages: 1-18 |
| ISSN: | 2210-6715 |
| DOI: | 10.1016/j.scs.2026.107378 |
| Online Access: | Verlag, lizenzpflichtig, Volltext: https://doi.org/10.1016/j.scs.2026.107378 Verlag, lizenzpflichtig, Volltext: https://www.sciencedirect.com/science/article/pii/S2210670726002659 
|
| Author Notes: | Zilang Cheng, Qinshan Li, Siyu Zhou, Qianyi Bei, Zhuoyan Lin, Hongzhang Lin, Jing Xie, Desheng Xue |
| Summary: | Assessments of surface urban heat island intensity (SUHII) in subtropical megaregions often rely on inconsistent urban-rural definitions, limiting comparability across cities and over time. Here we propose a gradient-explicit framework that combines Local Climate Zone (LCZ)-informed urban-rural areas (core, inner urban, suburban, and fringe areas) with pooled standard deviation (σ) normalization to quantify SUHII consistently at the urban-agglomeration scale. Using MODIS land surface temperature data for 2000-2024 in the Guangdong-Hong Kong-Macao Greater Bay Area, we mapped seasonal daytime and nighttime SUHII, estimated interannual trends and stability, and modeled SUHII variability using Random Forest with SHapley Additive exPlanations based on biophysical, built-environment, topographic, and socioeconomic predictors. A clear spatial hierarchy emerged across the gradient: daytime SUHII decreased from the core to the fringe area in all seasons and was strongest in summer, whereas nighttime gradients were weaker and peaked in autumn. Compact built LCZs (1-3) showed the highest intensities and fastest warming, while natural LCZs (A-G) remained low and stable. Attribution results showed clear seasonal and diurnal differentiation: topographic setting dominated most scenarios, surface moisture became especially important at night, particularly in autumn, and daytime SUHII was more closely associated with surface optical properties, socioeconomic intensity, and built-form-related conditions. This reproducible baseline supports area-specific heat adaptation in subtropical megaregions. |
|---|---|
| Item Description: | Online verfügbar: 6. April 2026, Artikelversion: 13. April 2026 Gesehen am 10.06.2026 |
| Physical Description: | Online Resource |
| ISSN: | 2210-6715 |
| DOI: | 10.1016/j.scs.2026.107378 |