Mapping trees in New York City for management and environmental applications
Project Overview
Mapping trees in New York City for management and environmental applications
This project addresses the need for more comprehensive urban tree maps to support better environmental management and public health decisions. Traditional tree-mapping methods are limited in scope, so the researchers used high-frequency satellite imagery, a machine learning model and existing street tree data to classify over 1.8 million trees in New York City by genus.
The resulting dataset enables a range of practical applications, including modeling pollen exposure for allergy sufferers, assessing how trees reduce urban heat, and improving strategies for tree planting and maintenance. Ultimately, the work provides a valuable tool for city planners and researchers, helping to enhance urban resilience, reduce heat-related health risks, and improve the management of urban forests in New York City and beyond.
Maps of urban trees are important for informing management decisions, such as combating the spread of invasive pests or pathogens or quantifying the ecosystem services provided by trees. Traditional manual mapping methods can only capture a small fraction of urban trees but incorporating data from satellites could allow the creation of comprehensive maps that would support tree-related decision making on a larger scale.
Using a combination of remote sensing datasets and a street tree dataset, we classified canopy trees to the genus level in New York City. To do so, we extracted seven years of vegetation indices from 766 daily image composites from Planet Labs’ high-temporal resolution satellites. From these, we derived information on the timing of leaf-out and other life cycle events for individual trees. The features from these images and from other remote sensing data were fed into a machine learning algorithm, which then classified over 1.8 million trees with an overall accuracy of 82%.
The Impacts
This tree classification project serves as the basis for several future projects that will benefit New Yorkers. One project includes the development of operational pollen models in NYC for allergy sufferers. Our tree classification data is also being used by a multi-disciplinary, multi-institutional team of researchers to quantify the cooling effects of trees in New York City, and assess how tree planting and maintenance can best be leveraged to reduce heat-related morbidity and mortality during heat waves. Our work also serves as the foundation for efforts to quantify tree mortality in New York City and improve efficiency and success of future tree planting and maintenance efforts. We’ve shared our findings with multiple partners and stakeholders, including The Nature Conservancy and the NYC Department of Parks and Recreations and will be sharing our results with the Forests For All NYC coalition, the ReLeaf conference for urban forest professionals in New York state, and the public through a Zenodo data repository and manuscript.
Our comprehensive map provides a better understanding of NYC’s urban trees, which is key to effectively managing them and enhancing the public benefits these trees provide. Our research will help cities, especially New York City, make better tree management decisions that will result in cooler cities and fewer heat-related illnesses and deaths during extreme heat events. The tree map will allow us to better understand how New Yorkers are exposed to allergenic pollen. We also expect the methods we developed here to be applicable to many other cities in the United States and globally.
Principal Investigator
Project Details
- Funding Source: McIntire-Stennis
- Statement Year: 2025
- Status: Completed Project
- Topics: Urban, trees, parks, New York City, mapping