Surface urban heat island intensity across 29 global cities: Disentangling the roles of climate zone, socioeconomic development, and urban land cover using landsat and google earth engine

Understanding surface urban heat island (SUHI) intensity variation across cities in diverse climate zones and socio-economic development conditions is critical for targeted urban heat mitigation. However, no prior study has jointly disentangled the effects of climate zone and socioeconomic development using a single, reproducible global protocol. This study quantifies daytime SUHI intensity across 29 cities present in three Köppen climate zones (Tropical, Arid, Temperate) and categorized into two Human Development Index (HDI) groups. SUHI intensity is assessed as a climatological, peak-season composite urban–rural LST contrast rather than an instantaneous value, is computed using Landsat 8/9 land surface temperature (LST) data accessed through Google Earth Engine. Urban and rural areas were delineated using the GHS-SMOD Degree of Urbanisation grids, and peak heat periods were identified from MODIS LST climatology (assessed for the period 2003-2023) for each city. Results indicate that SUHI intensity significantly varied across climate zones: tropical cities exhibited a mean SUHI of 3.38°C, temperate cities 2.46°C, while arid cities showed a surface urban cool island effect with −2.20°C. Additionally, HDI has strong correlation with SUHI intensity (partial r = 0.545, controlling for climate zone). High-HDI cities showed consistently higher SUHI with mean SUHI of +3.62°C compared to low-HDI cities with a mean of −0.84°C. Random Forest analysis indicates vegetation indices (urban and rural NDVI) significantly control the SUHI intensity, explaining 72% of the variance, followed by NDBI (13%) and bare surface fraction (6%). The thermal gradient analysis across radial LST transect profiles for 29 cities confirms diversity in SUHI patterns for different climate zones. Tropical and temperate cities exhibited declining LST gradients with distance from the urban core (−0.11 to −0.50°C/km), whereas arid cities showed increasing LST (0.06 to 0.29°C/km). Additionally, an NDVI threshold of 0.240 was identified, below which cities tend to exhibit urban cool islands. Temporal analysis of SUHI indicates, between 2014–2016 and 2022–2024, SUHI intensity increased significantly in tropical (0.80°C, p = 0.033) and temperate (0.60°C, p = 0.009) cities, but not in arid cities. These findings offer a scalable framework for SUHI assessment and emphasize that SUHI intensity varies significantly across global cities based on prevailing climate and socio-economic conditions.