Cities around the world are getting hotter, driven by climate change and urbanization, yet many lack the data and analytical tools to plan effective responses. This technical note from the World Resources Institute presents a methodology for helping cities understand what heat-resilient infrastructure they could realistically implement—and where.

At the heart of the framework is OpenUrban, a new high-resolution (1 m) urban land-use and land-cover (LULC) dataset built from free, globally available sources including OpenStreetMap, Overture Maps, and ESA WorldCover. Validated at 93% accuracy in the United States and 83% globally across 16 cities and eight world regions, OpenUrban maps the urban surfaces most relevant to heat—roads, buildings, parking lots, and public open spaces—at a high level of detail.

Paired with remotely sensed surface characteristics including albedo (reflectivity), tree canopy height, fractional vegetation cover, land surface temperature, and modeled shade, OpenUrban provides a spatially explicit baseline of existing conditions. From this baseline, the framework generates three nested scenario types for four key passive cooling interventions: trees, cool roofs, reflective pavements, and shade structures.

The three scenario levels—technical (maximum physically feasible), achievable (benchmarked to the city's own highest-performing areas), and program (constrained by specific policy objectives)—allow cities to compare what is theoretically possible against what is locally realistic and what a particular program or policy could deliver.

Each scenario produces two complementary outputs: a potential, which is an area-wide indicator of how much a surface characteristic (tree cover, albedo, or shade) would change under full implementation; and a scenario map, which spatially simulates where new infrastructure could be placed. Together, these outputs create a common evidence base that supports uses spanning community engagement, policy development, planning processes, and heat-modeling analyses.

The framework is designed to be city-agnostic and low-burden: it requires no local data inputs, runs on globally available open datasets, and is available as open-source code on GitHub. It has been developed in close collaboration with city officials and urban planners across more than 20 cities in Africa, Asia, Europe, Latin America, and North America, and is integrated into WRI's Cool Cities Lab.

Key Findings:

  • OpenUrban, the new 1 m resolution urban land-cover dataset introduced here, outperforms existing global products in both accuracy and urban class detail, achieving 93% accuracy in the US and 83% globally, with every validated city exceeding the 85% threshold considered reliable for land-cover datasets.
  • The methods are open-source, globally scalable, and integrated into the Cool Cities Lab, enabling cities with limited technical capacity to generate meaningful, policy-ready outputs without local data collection.

Executive Summary:

Cities worldwide face rising heat from climate change and urbanization. Passive, heat-resilient infrastructure—like trees, cool roofs, reflective pavements, and shade structures—are low-energy solutions, yet many cities lack data and tools to plan implementation. We present a globally scalable, open-source framework for generating infrastructure implementation scenarios. We introduce OpenUrban—a high-resolution urban land-use/ land-cover dataset validated at 93 percent accuracy in the United States and 83 percent globally—and pair it with remotely sensed surface characteristics (albedo, tree canopy, fractional vegetation, land-surface temperature). Using these inputs, we construct three nested scenario levels: technical (maximum feasible), achievable (benchmarked to high-performing areas), and program (policy-driven). We then produce scenario maps—spatially explicit simulations of infrastructure—and quantify them with potentials—indicators of area-wide changes in surface characteristics under full implementation. Together, these outputs provide a practical entry point for evidence-based action, lowering data barriers and supporting more resilient futures. The framework serves both technical and nontechnical practitioners, translating complex data into actionable takeaways.