Introduction

Extreme heat is emerging as one of the most dangerous and inequitable consequences of climate change, with heatwaves increasing in frequency, duration, and intensity. These events drive excess mortality, infrastructure failures, and economic losses in cities worldwide. With over half of the global population currently residing in urban areas, and projections indicating this will rise to two-thirds by 2050, cities are becoming critical hotspots of climate impact. Urban populations face compound risks due to anthropogenic warming, land cover modification, and socio-economic disparities. Despite this, most global assessments remain exposure-centric, relying on proxies like land surface temperature or heatwave frequency. This study addresses these limitations by developing a globally harmonised, component-resolved heat risk framework for 205 large cities with populations exceeding one million.

Materials and methods

The research adapts the INFORM Risk Index framework to assess heat risk at the city scale. Risk is conceptualised as a composite function of three interrelated dimensions: hazard exposure, vulnerability, and lack of coping capacity. Hazard exposure is characterised using population-weighted Universal Thermal Climate Index (UTCI)-based Cooling Degree Days (CDD) and land surface temperature (LST) anomalies. Vulnerability integrates demographic age structure (proportions of children under four and adults over 65), economic capacity (inverse GDP per capita), and inverse air-conditioning (AC) penetration. Lack of coping capacity is measured through the fraction of vegetation cover, expressed via the Normalised Difference Vegetation Index (NDVI) and Fractional Tree Cover (FTC), alongside electricity prices, which act as potential barriers to cooling access. The composite risk score is calculated as the geometric mean of these three dimensions to ensure a balanced view, acknowledging that high values in any one component can significantly amplify overall risk.

Results

The global assessment reveals that over 95% of cities in the 90th percentile of composite heat risk are concentrated in South and Southeast Asia and Sub-Saharan Africa. India, Pakistan, Nigeria, and Ghana host the largest number of these high-risk cities. A key finding is that high hazard exposure does not always translate to high risk. For instance, cities such as Kuala Lumpur, Bangkok, and Jeddah rank above the 90th percentile for hazard exposure but fall outside the top decile for composite risk because their relatively strong coping capacity, such as greater urban greening or lower energy costs, offsets thermal stress. Conversely, cities like Karachi, Faisalabad, and Kaduna face severe risk despite moderate exposure due to compounding social and infrastructural vulnerabilities. Specifically, Al Basrah in Iraq ranks as the highest-risk city with a composite score of 0.83, followed by Ahmadabad in India (0.79) and Bamako in Mali (0.78). Other high-risk cities include Madurai (0.73), Ibadan (0.73), and Kano, which exhibits extreme vulnerability (1.0) and lack of coping capacity (0.71).

Discussion

The results suggest that global heat risk planning must move beyond hazard-centric approaches toward component-targeted strategies that explicitly address exposure, vulnerability, and coping capacity. The study highlights the “sustainable cooling hierarchy,” which prioritises passive cooling and nature-based strategies like urban greening and shading, especially in rapidly growing and resource-constrained cities. In cities where coping capacity is high, there is a risk of maladaptation if there is an over-reliance on energy-intensive air conditioning, which can raise greenhouse gas emissions and waste heat. The analysis reveals that vulnerability-driven cities require stronger public health and social protection measures, including heat-health early warning systems. In contrast, cities with limited coping capacity should prioritise investments in affordable cooling, electricity resilience, and urban greening to build near-term resilience. Component-specific risk profiles can help align interventions with dominant local risk drivers, such as shading for exposure-driven cities or economic support for vulnerability-driven ones.

Conclusion

This study presents a globally consistent, component-resolved assessment of heat risk that integrates hazard, vulnerability, and lack of coping capacity. It demonstrates that vulnerability and coping deficits can substantially amplify risk even under moderate exposure conditions. The framework reveals distinct pathways through which heat risk emerges, even within the same country, highlighting the necessity for targeted, context-specific adaptation strategies. By disaggregating risk into its constituent drivers, urban planners can better align interventions with dominant local risk factors, whether that involves reducing exposure, alleviating vulnerability, or building infrastructural capacity. The findings provide a scalable basis for diagnosing heat risk in a rapidly warming world and support more effective, socially contextualised adaptation planning to advance urban heat resilience.