Jump to Content

Heat resilience

Using AI and satellite imagery to map rooftop reflectivity, helping cities fight rising urban temperatures.

Explore our data Our latest publication Research Blog post
Heat_Resilience_hero.original

Overview

Our high-resolution rooftop albedo (reflectivity) dataset provides the insights cities need to combat rising temperatures. By mapping building-level reflectivity, this dataset helps urban planners understand their current baseline, identify and prioritize neighborhoods for equitable cool roof interventions, and monitor changes over time to evaluate policy effectiveness.

Quick links

About Cool Roofs

Urban areas are warming at twice the global average rate, exacerbating the "urban heat island" effect. This phenomenon is driven by heat-absorbing surfaces like dark roofs and pavements, and a lack of green spaces. The consequences of extreme urban heat are severe, including:

  • Public health: Increased heat-related illnesses and mortality, disproportionately affecting vulnerable populations, including the elderly, children, and low-income communities.
  • Energy consumption: Higher demand for air conditioning leads to increased energy costs and strain on the electrical grid.
  • Quality of life: Reduced ability to be outdoors, impacting recreation, social interaction, and economic activity.
  • Environmental justice: Often, the hottest neighborhoods are those with less green space and a history of underinvestment.

Cool roofs are a proven, cost-effective solution. By reflecting more sunlight and absorbing less heat, they can lower building temperatures, reduce energy use, decrease air pollution, and mitigate the urban heat island effect. Our work provides the granular, building-level data needed to strategically implement cool roof programs.

The Cool Roofs dataset

The Cool Roofs dataset provides high-resolution, building-level albedo (reflectivity) measurements. These measurements are derived from specialized AI models applied to very high-resolution satellite and aerial imagery. The dataset is designed to provide the granular insights needed to help researchers, urban planners, and policymakers strategically implement cool roof programs and monitor changes over time.

The dataset features albedo data associated with building centroids (a single coordinate point per building, which contain the average albedo value for that roof) generated from a comprehensive, proprietary Google building outlines dataset. This version covers approximately 50 cities globally.

Explore the dataset

The dataset is fully accessible via the interactive Heat Resilience Earth Engine App. Urban planners can use the tool to pan across a city, toggle between aggregated neighborhood views and building-level details, and select individual structures to view specific albedo data for targeted cool roof planning.

Albedo data displayed for individual buildings (as centroids). A nested interface allows users to transition seamlessly from aggregated views down to individual building insights. Click here to open the app in another window

Data Download

Centroids generated using a comprehensive building outlines dataset (Google proprietary). Coverage of approximately 50 cities.

Argentina:

Australia:

Brazil:

Greece:

India:

Indonesia:

Kenya:

Mexico:

Nigeria:

Pakistan:

South Africa:

Spain:

United Kingdom:

United States:

Recent publications

Estimating high-resolution albedo for urban applications
David Fork, Elizabeth J. Wesley, Salil Banerjee, Vishal Batchu, Aniruddh Chennapragada, Kevin Crossan, Bryce Cronkite-Ratcliff, Ellie Delich, Tristan Goulden, Mansi Kansal, Jonas Kemp, Eric Mackres, Yael Mayer, Rebecca Milman, John C. Platt, Shruthi Prabhakara, Gautam Prasad, Shravya Shetty, Charlotte Stanton, Wayne Sun, and Lucy R. Hutyra
Nature Communications, June 2026

FAQ

What is albedo?

Albedo is a metric of surface reflectivity, representing the fraction of solar energy it reflects into the atmosphere. Measured on a scale from 0 to 1, a value of 0 indicates complete absorption, while a value of 1 indicates complete reflection. In urban environments, utilizing high-albedo materials for rooftops (often called "cool roofs") reduces heat absorption, lowers internal building temperatures, and mitigates the urban heat island effect.

Typical albedo values for common surfaces include:

  • Fresh asphalt / Dark roofs: 0.05 – 0.15
  • Grass / Vegetation: 0.15 – 0.25
  • Aged concrete: 0.20 – 0.30
  • White "cool" roofs: 0.60 – 0.85

Which cities are covered?

We are currently covering 50+ cities across 9 countries:

  • Argentina: Buenos Aires
  • Australia: Melbourne
  • Brazil: Belém, Belo Horizonte, Campina Grande, Campinas, Curitiba, Florianopolis, Fortaleza, Goiânia, Manaus, Porto Alegre, Recife, Rio de Janeiro, Salvador, São Paulo, Teresina, Vitoria
  • Greece: Athens
  • India: Ahmedabad, Bhopal, Chennai, Coimbatore, Kolkata, Madurai, Mumbai, Navi Mumbai
  • Indonesia: Jakarta
  • Kenya: Kisumu, Nairobi
  • Mexico: Hermosillo, Mexico City, Monterrey
  • Nigeria: Aba City, Benin City, Damaturu, Etim Ekpo, Ika, Kaduna, Kanos, Lagos, Orukanam, Ukanafun, Yola
  • Pakistan: Karachi
  • South Africa: Cape Town, Durban, Johannesburg
  • Spain: Barcelona
  • United Kingdom: London
  • United States: Austin, Baltimore, Boston, Boulder, Colorado Springs, Dallas, Fort Worth, Los Angeles, Miami-Dade, Nashville, New York, Oklahoma City, Phoenix, San Antonio, Stockton, Tempe, Washington

What is the plan for covering additional cities?

We are continuously working to expand our coverage.

Will this data be refreshed in the future?

Yes, we aim to provide periodic refreshes to help cities monitor changes over time, subject to imagery availability and processing capabilities.

Can I use the data directly in my tools?

Yes, the downloadable data is provided in standard formats suitable for use in GIS software and other analytical tools. Ensure you correctly mention Google and Overture's contribution.

What is the difference between this albedo data and the data available on Environmental Insights Explorer and on Google Earth?

This dataset provides high-resolution, building-level albedo measurements derived from specialized AI models applied to very high-resolution imagery. While Environmental Insights Explorer (EIE) also features cool roof insights, our platform provides more granular downloadable data and the interactive Earth Engine App for detailed exploration. Building on Google Earth image capabilities, this dataset introduces quantitative albedo data at this scale.

How should I correctly cite the dataset and the underlying building outlines?

Proper citation depends on the dataset version you are utilizing:

  1. Source

  2. Attribution:

    © OpenStreetMap contributors. Available under the Open Database License.

    Esri Community Maps contributors. Available under CC BY 4.0.

    Global ML Building Footprints. Licensed by Microsoft under the Open Database License.

    Google Open Buildings. Available under CC BY 4.0.

    USGS 3D Elevation Program Digital Elevation Program.

    Qian Shi, et al. A First High-quality Vector Data of Buildings in East Asian Countries Based on a Comprehensive Large-scale Mapping Framework. Zenodo, 22 July 2023, doi:10.5281/zenodo.8174931. Available under CC BY 4.0.

    Work derived from BTN 2024 ign.es. Available under CC BY 4.0

×