On New York University Week: Green spaces are key for cooling cities, but not every urban area has enough.
Rumi Chunara, associate professor of computer science and engineering and biostatistics, investigates this from an urban planning perspective.
The overarching goal of Rumi Chunara’s research is to develop computational and statistical approaches for acquiring, integrating and using data to improve population-level public health. She focuses on the design and development of data mining and machine learning methods to address challenges related to data and goals of public health, as well as fairness and ethics in the design and use of data and algorithms embedded in social systems.
At NYU, Chunara leads the Center for Health Data Science, which develops computational and statistical methods across data mining, natural language processing, spatio-temporal analyses and machine learning, to study and improve population health. Previously, she was a Postdoctoral Fellow and Instructor at HealthMap and the Children’s Hospital Informatics Program at Harvard Medical School. She completed her PhD at the Harvard-MIT Division of Health Sciences and Technology and BSc at Caltech.
Mapping Parks and Greenspaces for Healthier Cities
When cities face extreme heat, green spaces become lifelines: cooling neighborhoods, filtering air pollution, and providing places for exercise, stress relief, and social connection.
But here’s the problem: cities have long struggled to accurately track their vegetation. We developed an artificial intelligence solution using high-resolution satellite imagery. By training AI algorithms to better handle variation in vegetation appearance caused by different lighting and seasonal conditions, we significantly improve the model’s generalizability in real-world environments.
Our system reached about 90% accuracy and reliability in identifying vegetation, compared to traditional methods that only achieve about 63% in those areas. This represents a substantial improvement in detection that makes all the difference for urban planning.
We tested our approach in Karachi, Pakistan’s largest city, and exposed stark environmental divides. The city averages just over 4 square meters of green space per person, less than half the World Health Organization’s recommended minimum. Some areas have over 80 square meters per person, but five neighborhoods have less than one-tenth of a square meter per capita.
While traditional methods rely on simple light wavelength measurements, our AI system learns to recognize subtle patterns that distinguish trees from grass, even in challenging urban environments. This enabled us to show that areas with more vegetation showed lower surface temperatures, demonstrating how green spaces cool cities.
This accurate mapping is essential for cities addressing climate change. Without precise data, urban planners cannot easily know which neighborhoods desperately need vegetation or develop green spaces that will deliver maximum benefits. This methodology now makes a way for heat mitigation to be enjoyed by everyone who lives, works or visits there.
Read More:
[NYU Tandon School of Engineering] – New AI system accurately maps urban green spaces, exposing environmental divides
[Center For Health Data Science] – Quantifying greenspace in Karachi, Pakistan with high-resolution satellite imagery
[ACM Digital Library] – Quantifying Greenspace with Satellite Images in Karachi, Pakistan, Using a New Data Augmentation Paradigm
