Researchers at USC have developed a low-cost artificial intelligence tool that uses free aerial imagery to help cities map urban tree canopies and identify where trees could provide the greatest benefit in reducing extreme heat, the university announced Wednesday.
The tool combines publicly available aerial photographs with AI to create detailed tree canopy maps without relying on costly surveys or commercial satellite imagery that many existing systems require, according to the researchers.
“Trees provide a wide range of benefits, including helping reduce the health risks associated with rising temperatures in cities,” John Wilson, founding director of USC’s Spatial Sciences Institute and the study’s corresponding author, said in a statement. “But to plant trees where they’ll make the biggest difference, cities first need a clear picture of their existing tree canopy.”
The study, published in the journal Remote Sensing, comes as cities across the country grapple with more frequent periods of extreme heat and look for ways to improve climate resilience.
Researchers said the system uses aerial photographs collected every two to three years through the U.S. Department of Agriculture’s National Agriculture Imagery Program, allowing communities to create detailed tree canopy maps using free data.
The tool was developed and tested in Boyle Heights and City Terrace, neighborhoods east of downtown Los Angeles that historically have had less tree coverage than wealthier areas of the city, researchers said.
To test whether the technology could be used elsewhere, researchers applied the model to neighborhoods in San Francisco and Phoenix without additional training and reported similarly strong results despite differences in climate and urban design.
According to the research team, the ArcGIS deep learning package developed through the project has been downloaded more than 12,900 times during the last six months from Esri’s Living Atlas platform.
“Our work shows how we can use free, publicly accessible data to map tree canopy over time — data cities can use to guide planting plans at every scale, from a single street block to an entire county,” Wilson said.
The research, code and ArcGIS model have been made publicly available for use by cities and organizations that may not have in-house machine learning expertise.
Wilson said researchers next plan to combine the AI tool with publicly available data to estimate how much shade existing trees provide and how much additional shade future plantings could create.