Re-leaf
Measuring the cooling effects of trees globally
Trees are the most effective cooling systems in urban areas. In Re-leaf we have collected and analyzed thousands of thermal and RGB images globally to demonstrate that different tree genera have different cooling powers, and that they vary based on urban settings.
We also built an unsupervised AI tool that scans street-view images to infer the distribution of tree genera and calculate key diversity indices, helping cities boost the ecological and climate benefits of their green spaces.
Re-leaf was first shown in the 19th Venice Architecture Biennale.
About
Re-leaf uses computer vision, RGB and thermal imagery to investigate the cooling
powers
of trees. It investigates how different tree genera perform in different urban
settings
in Amsterdam, Boston, Dubai, Los Angeles, and Rome.
We compare the cooling power of each genus in different ambient temperatures, weather conditions, and urban settings. Re-leaf provides a data-informed evaluation tool to help cities to define their tree-planting schemes.
As most cities lack a comprehensive cadaster of their urban trees, we also developed an unsupervised AI framework that leverages unlabeled street-view imagery to infer the distribution of tree genera and recover key diversity indices such as Shannon and Simpson. This enables scalable monitoring of both abundance patterns and evenness across urban areas, providing cities with actionable insights to enhance the ecological and climate benefits of their green infrastructure.
We compare the cooling power of each genus in different ambient temperatures, weather conditions, and urban settings. Re-leaf provides a data-informed evaluation tool to help cities to define their tree-planting schemes.
As most cities lack a comprehensive cadaster of their urban trees, we also developed an unsupervised AI framework that leverages unlabeled street-view imagery to infer the distribution of tree genera and recover key diversity indices such as Shannon and Simpson. This enables scalable monitoring of both abundance patterns and evenness across urban areas, providing cities with actionable insights to enhance the ecological and climate benefits of their green infrastructure.
Venice Biennale
Explore urban
greenery
in Amsterdam, Dubai, Los Angeles and Rome. Each 500 m grid is
extruded based on its greenery index, drawing from our Green Bias
project.
Switch to Tree View to examine individual trees,
and browse the catalogue to discover each city's five most common genera. Tap a tree
to
view its
distribution; double tap to learn more about its morphology.
With the support of Dubai Future Foundation, AMS Institute and Arma dei Carabinieri
With the support of Dubai Future Foundation, AMS Institute and Arma dei Carabinieri
Press
Download the Press Material
The material on this website can be used freely in any publication provided that:
1. It is duly credited as a project by the MIT Senseable City Lab
2. A PDF copy of the publication is sent to senseable-press@mit.edu
Accessibility
The material on this website can be used freely in any publication provided that:
1. It is duly credited as a project by the MIT Senseable City Lab
2. A PDF copy of the publication is sent to senseable-press@mit.edu
Accessibility
Team
Carlo Ratti, director
Mikita Klimenka, lead
Diaa Addeen Abuhani, lead
Marco Seccaroni
Martina Mazzarello
Fabio Duarte
Simone Mora
Les Norford
Chada Elalami, web + viz*
*the 3D tree models were provided by Jae Joong Lee, from Sara Beery’s team at MIT CSAIL
Mikita Klimenka, lead
Diaa Addeen Abuhani, lead
Marco Seccaroni
Martina Mazzarello
Fabio Duarte
Simone Mora
Les Norford
Chada Elalami, web + viz*
*the 3D tree models were provided by Jae Joong Lee, from Sara Beery’s team at MIT CSAIL
Publication
Klimenka, M., Mora, S., Abuhani, D. A., Norford, L., Duarte, F., Ratti, C. (2025).
Cooling performance: Exploring the heat mitigation effect of
urban
trees with computer
vision . arXiv.
Abuhani, D. A., Seccaroni, M., Mazzarello, M., Zualkernan, I., Duarte, F., Ratti, C. (2025). Unsupervised mapping of urban tree diversity using spatially-aware visual clustering. arXiv.
Abuhani, D. A., Seccaroni, M., Mazzarello, M., Zualkernan, I., Duarte, F., Ratti, C. (2025). Unsupervised mapping of urban tree diversity using spatially-aware visual clustering. arXiv.
A project by Senseable City Dubai