Urban heat kills 489,000 people annually, according to the World Resources Institute. More than nearly all ongoing armed conflicts. And unlike those conflicts, the victims do not all die in the same place: they die in specific neighborhoods, on specific heat islands, in streets where concrete accumulates and releases heat all night, depriving bodies of the recovery that only a cool night can provide.

The mapping of these neighborhoods now exists. The technical solution is known, inexpensive, and has no energy externality. The question that remains open is not scientific. It is political.

The Essential Points

  • The WRI launched the Cool Cities Lab platform in March 2026, which models the effect of trees, white roofs, and shade structures neighborhood by neighborhood in 25 cities across five continents.
  • 489,000 deaths per year are attributable to extreme urban heat according to the WRI; WRI and WHO projections show a 50% increase in heat-related deaths by 2050, with concentration in Global South metropolises.
  • Google simultaneously published an open dataset on roof reflectivity for 50 cities, making data accessible to municipalities without the means to produce it themselves.
  • The neighborhoods most exposed to heat are also the poorest, the least tree-covered, and the last to benefit from municipal interventions when resources are allocated without spatial justice constraints.

When Concrete Releases Heat While the Rich Sleep Under Air Conditioning

To understand what passive cooling changes, we must first understand what an urban heat island is and why it kills more in some neighborhoods than others.

A city heats up more than a neighboring rural area for simple reasons: impervious surfaces absorb solar energy during the day and release it at night, the absence of vegetation deprives streets of evapotranspiration that naturally cools, and the density of buildings traps hot air. The difference can reach 7 to 10°C between a densely built neighborhood without trees and a tree-covered neighborhood in the same city. This is not marginal discomfort: at 35°C at night, the human body does not recover. The elderly, infants, physical workers without access to air conditioning die of hyperthermia.

Air conditioning solves the problem for those who can afford it. It worsens it collectively: each outdoor unit releases heat into the street, contributing to the heat island that affects the neighbor who cannot afford to install one. This is a perfect externality, invisible in individual calculations and catastrophic at the block level.

Passive cooling works in reverse. A tree planted in the right location reduces the perceived temperature by 2 to 8°C within its immediate shadow radius, depending on the species and built environment. A roof painted white reduces the heat absorbed by the building and lowers the surrounding ambient temperature. A shade structure on a sidewalk changes movement behavior and reduces exposure. These interventions have no electrical consumption. Their benefit is shared by all nearby residents, not just those with the means to invest.

The problem, until now, lay in precision. Planting trees “in hot neighborhoods” is a general political decision. Knowing that the block at the intersection of two avenues without vegetation, with dark roofs and a population density of 450 inhabitants per hectare, would gain 3.2°C in nighttime temperature with five strategically positioned trees and two sections of roof repainted white: this is an operational decision. The difference between the two is not rhetorical. It determines whether municipal budgets produce measurable impact or diffuse spending.

Cool Cities Lab: Modeling Cooling Block by Block

This is precisely the gap that the WRI sought to fill with the Cool Cities Lab platform, launched in March 2026. The tool covers 25 cities across five continents, from sub-Saharan African metropolises to Southeast Asian capitals to agglomerations in Latin America and a few cities in the Global North included as reference points.

The platform does not merely map existing heat. It models the effects of interventions. For each block in a covered city, it becomes possible to simulate what would result from a 10% increase in canopy, reflective paint on 30% of roofs, or the installation of shade structures on main pedestrian axes. The thermal gain is expressed in degrees, spatially distributed, and considered against the population density exposed.

This is a change in the scale of decision-making. Previously, studies on the effects of passive cooling existed, but they were isolated, expensive to produce, and difficult to compare between cities. A consulting firm could commission a simulation for a specific neighborhood; no city in the Global South could afford systematic modeling of its entire territory.

Google complemented the initiative by releasing a public dataset on roof reflectivity for 50 cities. These data, produced by satellite analysis and machine learning, make it possible to know precisely which roofs absorb the most heat and where interventions would have the greatest impact. For a municipality that has neither the tools nor the budget to produce this data itself, this is a direct lever.

The combination of the two tools creates something new: the capacity for thermal planning accessible to cities that could never have financed an equivalent study on their own.

Nairobi, Jakarta, Phoenix: Different Contexts, the Same Oversight

The 25 cities covered by Cool Cities Lab do not face the same constraints. Phoenix, Arizona, has significant municipal resources and has already launched a tree-planting program with a budget of 60 million dollars over five years, managed by its extreme heat office. Nairobi and Jakarta do not have these margins. Yet the platform is designed for them, precisely because the least available data are also what poor cities need most.

Researchers working on urban climate adaptation note a recurring paradox: the neighborhoods most exposed to heat are also those where land is cheapest, where density is highest, where roofs are darkest, and where vegetation is rarest. This is not coincidence. It is the product of urban history in which zones exposed to risks were ceded to populations without bargaining power, and where public investment followed the gradient of property value rather than need.

The platform can map these inequalities with unprecedented precision. It can show that a neighborhood concentrates 40% of the population exposed to nighttime temperatures above 32°C and benefits from only 5% of tree cover. What the tool cannot do is decide to prioritize that neighborhood when resources are constrained and other actors have greater political influence.

This is where the question shifts from technical to governmental.

489,000 Deaths in 2026, Probably More Than 700,000 by 2050

The figure of 489,000 annual deaths linked to extreme urban heat provided by the WRI is itself probably underestimated. Epidemiological studies on excess heat-related mortality are based on statistically measurable excess deaths during heat waves. They do not capture diffuse deaths linked to weeks of moderate but continuous heat, which progressively exhaust fragile organisms without producing an identifiable peak.

According to WRI and WHO projections, this figure could increase by 50% by 2050, reaching approximately 733,000 annual deaths. The concentration of these additional deaths will be in Global South metropolises, where demographic growth, densification, and lack of adaptation resources will cumulate.

Approximately 733,000 annual deaths from urban heat by 2050 is a reality that cities equipping themselves today with thermal planning tools choose to prevent. Cities that do not do so are not making a neutral decision: they allow a vulnerability to form that climate physics will make increasingly costly to correct.

We must understand the dimension of partial irreversibility of what is happening now. Trees planted in 2026 will have their full effect in ten to fifteen years. Roofs renovated today will last thirty years. Construction standards that integrate material reflectivity into urban codes structure the city of 2060. Each year of delay pushes back by that amount the maturity of tree cover, at the precise moment when temperatures will continue to rise.

Thermal Equity Is Not Decreed, It Is Planned

The formulation “passive cooling as a right of universal access” may sound like an abstract political injunction. It actually describes a very concrete operational decision: municipalities that have thermal modeling tools can choose to integrate a criterion of spatial justice in their budget arbitrations, or choose not to.

The experience of cities that have advanced most on this subject shows that the problem is not purely budgetary. Medellín, Colombia, developed from 2016 “green corridors” in its most densely built and disadvantaged neighborhoods, combining tree planting, renovation of public spaces, and community participation in maintenance. The documented thermal effect exceeds 3°C of reduction in the affected streets. Singapore integrated greening constraints into its building codes from 2008, when these requirements became mandatory, and applies tree replacement rules with ratios greater than 1:1. These policies did not arise spontaneously: they resulted from political decisions that recognized heat as a public health priority, not as a meteorological inconvenience.

What Cool Cities Lab brings to this equation is the capacity to make these arbitrations visible and contestable. When modeling shows that two neighborhoods have comparable exposure levels but cooling investments systematically go to the wealthier neighborhood, this distortion becomes documentable. It can be raised by associations, integrated into local planning plans, cited in budget advocacy. The tool does not guarantee just decisions, but it removes the excuse of ignorance.

The question of data sharing remains entirely open. Open platforms like those of the WRI and Google present a model different from proprietary solutions, where a city purchases modeling from a commercial provider and remains dependent on it for updates. Data openness allows local researchers, NGOs, data journalists to work on the same bases as municipal decision-makers. This is a condition of reclamation.

What 25 Cities Will Teach the World

The value of Cool Cities Lab is not solely in the 25 cities covered at its launch. It is in what these 25 cities will produce in terms of feedback. The intervention decisions made based on the models can be compared to temperatures actually measured after plantings. The documented effects will allow models to be refined, local variables that matter most to be identified (air humidity, wind, soil type), tree species most effective in different climatic contexts to be distinguished.

A shared knowledge infrastructure on urban cooling is beginning to be built, at a global scale and in very different contexts. It did not exist five years ago.

For Global South cities that do not yet have the means to develop their own tools, this growing corpus represents something precious: the possibility of not reinventing what has already been tested elsewhere. A secondary Sahel city that wants to rethink its street tree policy does not need to commission a study from scratch if the effects in Medellín or Freetown are documented, comparable, and accessible.

The WRI plans to expand the platform. The pace of this expansion, and especially the criteria for prioritizing cities to be integrated, will be one of the indicators of the seriousness of the approach. If the next cities added are Global North metropolises with strong institutional capacity and already high adaptation budgets, the promise of a tool serving the most vulnerable cities will remain theoretical. If the next phases cover Dhaka, Lagos, or Karachi, the signal will be different.

Heat kills silently, without the images of an earthquake or flood. It kills more in neighborhoods where people have no political say in planning decisions. The precision of thermal mapping does not change this balance of power. It simply makes it harder to ignore.

Sources

  1. World Resources Institute, Cool Cities Lab launch press release, March 2026: https://www.wri.org/news/release-new-global-platform-maps-urban-heat-risks-block-block-and-shows-cities-how-cool-them
  2. IPCC, Sixth Assessment Report (AR6), Working Group II, 2022 (impacts, adaptation and vulnerability), available on ipcc.ch
  3. City of Phoenix, Heat Relief program and Heat Action Plan strategy, Phoenix Office of Heat Response and Mitigation: https://www.phoenix.gov/administration/departments/heat/tree-shade-programs/tree-grant-programs.html
  4. Study on Medellín’s green corridors, Universidad Nacional de Colombia and Alcaldía de Medellín, 2020-2022: https://ap-plat.nies.go.jp/inas/goodpractices/development/5.html
  5. Google Environmental Insights Explorer, roof reflectivity data, available on insights.sustainability.google: https://research.google/blog/expanding-our-heat-resilience-data-to-50-global-cities/
  6. WRI - Urban Heat & Passive Cooling (489,000 figure and WHO projections): https://www.wri.org/initiatives/urban-heat-passive-cooling
  7. WRI - Cooling Potential of Urban Trees: https://www.wri.org/insights/urban-trees-cooling-potential
  8. BCA Singapore / Baker McKenzie - Green Mark since 2005, mandatory from 2008: https://resourcehub.bakermckenzie.com/en/resources/global-sustainable-buildings/asia-pacific/singapore/topics/green-certification