Under the SSP5-8.5 scenario, the study projects that the number of major cities exceeding 29 °C average annual temperature would increase from 17 in 2011-2040 to 217 in 2071-2100. This leap concentrates risk in cities with varying response capacities.
The Essentials
- Between 2071 and 2100, 320 million residents will live in cities exceeding the 29 °C average annual temperature threshold, compared to 17 cities affected in 2011-2040 (NCBI/Nature Climate Change).
- Risk is not distributed uniformly: the cities experiencing the highest temperatures are also those with the weakest adaptive capacity.
- Where insurance companies withdraw and real estate prices collapse, leaving ceases to be a choice and becomes an economic constraint.
- Infrastructure decisions made between 2025 and 2040 will determine who can stay: according to the United Nations, 1.8 trillion dollars in adaptation investment would prevent 7.1 trillion dollars in future costs.
- The governance of climate migration remains a blind spot in international law.
217 Cities, But Not the Same Problems
The figure strikes with brutal force. Seventeen cities exceeded 29 °C average annual temperature during 2011-2040. Between 2071 and 2100, the Scientific Reports study projects 217 cities beyond 29 °C under SSP5-8.5. That is twelve times more. Up to 320 million people could live in 2100 in cities exceeding 29 °C average annual temperature in the SSP5-8.5 scenario.
But the global figure masks a geography. These 217 cities are not Los Angeles and Toronto. According to the study, projected populations are concentrated primarily in cities in Asia, Africa, and to a lesser extent South America. Regions where baseline heat is already high, where urban infrastructure remains fragile, and where social safety nets are thin. The coincidence between intensity of risk and weakness of adaptive capacity is not a statistical accident.
It is the product of centuries of differentiated investment.
The study primarily crosses temperature projections, population, and urban morphology; it does not construct a temperature/adaptive-capacity ratio. Adaptation and displacement result from interactions between climatic hazards, inequalities, infrastructure, institutions, income, public policies, and mobility options.
Heat as an Economic Solvent
To understand why heat displaces populations, one must first understand how heat destroys a local economy.
The first mechanism is productivity. A wet-bulb temperature close to 35 °C constitutes an extreme danger for thermoregulation, but there is no single universal threshold beyond which all heat dissipation becomes impossible. At very high levels of heat and humidity, outdoor work becomes extremely dangerous and requires reduced effort, breaks, shade, water, and sometimes work stoppage; the threshold is not universally set at 35 °C. Agriculture, construction, port logistics: entire sectors shut down. In cities already close to this threshold, a few additional degrees do not shift the productivity curve—they make it collapse.
The second mechanism is insurance. Insurers calculate their exposure based on historical risk distributions. When these distributions shift permanently, premiums rise, then actors withdraw. We already observe this today in American and Australian coastal cities struck by repeated climate events. In Florida, several major insurers have ceased renewing residential contracts.
This is a precursor signal of what awaits the most exposed zones: when insurance disappears, mortgage credit follows. And when mortgage credit disappears, real estate prices collapse.
The third mechanism is real estate. A property in a zone that insurers are fleeing and that banks refuse to finance loses its value in fits and starts. Households that leave first still recover some fraction of their capital. Those who remain, lacking resources to leave, see their wealth evaporate without being able to reposition elsewhere.
The structuring role of real estate capital in household economies makes this mechanism particularly destructive for the middle classes of developing countries, whose accumulated wealth is largely in this form.
Populations leaving these zones depart from a place that has become too expensive to live in, in the literal sense, without necessarily reaching somewhere better.
Adaptation: Effects and Beneficiaries
Adaptation to urban climate exists. It takes concrete forms: green roofs and public spaces to reduce heat island effects, reflective surface coatings, collective air conditioning networks powered by renewable energy, reorganization of work schedules, early warning systems, universal access to cooled spaces. Singapore has invested in these measures for thirty years. Some Japanese cities have rebuilt their roads with absorbent materials. Dense urban forest pilot projects exist in Medellin, Melbourne, and Montpellier.
These solutions work. But they cost. And their cost is precisely beyond the reach of the cities that need them most.
The World Economic Forum estimates a potential 50% loss in value of non-adapted infrastructure assets by 2050 in high climate risk zones. UNEP cites cumulative investment of 1.8 trillion USD in adaptation over 2020-2030; it does not specify 1.8 trillion USD per year as the investment necessary to achieve 7.1 trillion USD in benefits. The equation is favorable. But it presupposes States capable of long-term investment, public or private capital markets willing to finance thirty-year assets, and sufficiently stable local governance to execute multiyear plans.
These conditions exist in Copenhagen, Seoul, Dubai. They are more fragile in Dhaka, Lagos, Karachi. Several cities have thin fiscal margins facing growing climate risks. The relationship between needs and capacities is inverted.
This imbalance also runs through the interior of wealthy cities. In Global North metropolises, the most intense heat islands systematically overlap with the poorest neighborhoods: less vegetation, more concrete, poorly insulated housing, limited access to air conditioning. Heat amplifies preexisting inequalities within cities themselves that can globally adapt.
Decisions Between 2025-2040 Will Draw the Map of 2100
The window between today and 2040 is the one that matters. Infrastructure built today will strongly influence long-term habitability and the possibility of remaining, among other major determinants.
A residential building constructed today has a lifespan of fifty to seventy years. An urban transport system, sixty years. A water network, more than a century. Cities that invest now in infrastructure designed for a climate at +3 or +4 °C preserve their future habitability. Those that build by the old standard, for lack of financing or vision, create stranded assets: equipment that will become unusable before the end of their theoretical life.
International climate policy meets development policy here. Financing adaptation in vulnerable countries is a decision about the human geography of the coming century. The commitments made in climate negotiations, notably the loss and damage fund established at COP27 and the long-undercapitalized adaptation fund, aim precisely to provide these cities with the capacity to invest in time.
The signals are mixed. Adaptation financing remains largely below identified needs. Developed countries were slow to reach the goal of 100 billion USD annually in climate finance for developing countries; this goal covered both mitigation and adaptation, not mitigation alone. Innovative financing mechanisms—guarantee funds, climate bonds, public-private partnerships oriented toward resilience—exist but still operate at too small a scale.
The Governance of Migration Facing Climate Risk
The geography of differentiated risk poses a central political question: that of mobility. Tens of millions of people could no longer be able to stay where they live. Rules for reception, associated rights, and destinations remain to be defined.
International refugee law, built after 1945, does not recognize environmental causes as grounds for protection. A person displaced exclusively by thermal and economic degradation does not automatically benefit from refugee status under the 1951 Convention, but other protections may sometimes apply depending on context. The Geneva Convention protects those fleeing persecution. It is silent on those whom structural heat progressively expels from their place of residence.
A few countries are beginning to fill this gap through bilateral means. New Zealand offers a specific residence pathway to certain Pacific nationals via the Pacific Access Category, but this is not a climate protection or relocation visa due to submersion. The African Union is working on regional frameworks for climate-related internal mobility. These initiatives are useful. They remain sporadic and without guarantee of rights.
The institutional question runs deeper. Scientific, health, and risk management frameworks already allow for assessment of components of urban uninhabitability. Rules for distributing financing for climate displacement remain fragmented and insufficiently binding. Conditions for receiving persons displaced by climate are covered partially and fragmentarily by existing frameworks; there is no unified specific international regime.
The way societies organize their institutions determines their capacity to absorb shocks: large-scale climate migration will test these institutions as few events have.
The stakes exceed the humanitarian plane alone. Unanticipated and ungoverned migration flows can produce political tensions capable of weakening climate policies. Proactive governance of climate mobility, built before flows are massive, is both more effective and less costly than crisis management.
Cities Experimenting with Large-Scale Adaptation
Several cities in high-risk zones are not content to wait. They experiment, often with limited resources, with solutions that merit attention.
Chennai, in India, confronted with increasingly intense heat waves and severe water stress, has deployed a network of free public cooling centers in popular neighborhoods, coupled with an SMS alert system. The direct impact on mortality during heat peaks is documented, even if infrastructure remains fragile.
Medellín, in Colombia, has reduced average temperature in certain urban corridors by up to 4 °C through “green corridors” linking parks and green roofs. A municipal program, partly financed by international cooperation, has made it possible to scale up solutions that, individually, exist everywhere in the world but never reach scale.
Rwanda already had a national building code from 2019 incorporating passive design and thermal comfort requirements; a technical standard from 2021 addresses thermal actions on buildings. The additional construction cost is estimated at 5-8%; the energy cost in air conditioning avoided over the building’s lifetime is several times higher.
These examples share a characteristic: they rest on strong local political will, often supported by targeted international financing. They show that low-income adaptation is possible. They also show that it remains conditional on resources and governance that many affected cities do not yet have.
Mobility as a Right, Not as Fate
The image of 217 cities at 29 °C in 2100 can discourage. It should not. Most of these cities still have fifty years to make decisions that will alter their trajectory. The window is closing, but it remains open.
The central question is not whether populations in the most exposed zones will need to move, but whether this movement will be organized or imposed, anticipated or endured, financed publicly or paid for through private impoverishment. A city that plans its infrastructure today, trains its municipal technicians in bioclimatic design, negotiates regional mobility agreements, and accesses international adaptation funds can substantially reduce the number of its inhabitants forced to leave.
What remains to be built, at the international scale, is the framework that allows these cities not to face alone an inequality they did not create. Rich economies have contributed very significantly to cumulative historical emissions and generally have higher adaptive capacities, but attribution of majority cumulative future emissions under the SSP5-8.5 scenario requires separate quantitative analysis. Financing adaptation in vulnerable cities is an essential component, along with emissions reduction, planning, public services, and equitable governance, of maintaining habitability.
Sources
- NCBI / Nature Climate Change – Study on cities exceeding 29 °C average annual temperature between 2071 and 2100: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12457604/
- United Nations – Report on climate adaptation financing (1.8 trillion / 7.1 trillion in avoided costs): United Nations Environment Programme, Adaptation Gap Report, UNEP, 2023 edition
- World Economic Forum – Estimate of value loss of non-adapted infrastructure assets by 2050: WEF, The Global Risks Report 2025
- UNHCR – 1951 Geneva Convention relating to the status of refugees: https://www.unhcr.org/en/