Differences in magnitude and spatial distribution of urban forest pollution deposition rates, air pollution emissions, and ambient neighborhood air quality in New York City

• Variation exists in the pollution removal potential of New York City's urban forest.
• Land Use Regression reveals that tree cover represents absence of emissions sources.
• Total NO2 and PM10 emissions exceed canopy deposition rates at the county level.
• PM10 canopy deposition is slightly higher than traffic-based emissions.
• Estimated total emissions of both pollutants are spatially disconnected from deposition.

Urban forest pollution removal potential has not been well explored at the neighborhood resolution and in relation to neighborhood-level emissions. In NYC's five counties, modeled NO2 removed by the primarily-deciduous urban forest ranges from <1% (New York) to 13% (Richmond) of total emissions; modeled PM10 removal ranges from <4% (New York) to 20% (Richmond). Across a 900 m2 grid, average traffic NO2 emissions are over an order of magnitude greater than canopy removal; PM10 canopy removal slightly exceeds average traffic emissions. NO2 and PM10 removal are weakly but significantly inversely correlated in space with traffic emissions at the grid level (r = −0.126, p < 0.0001). Land Use Regression modeling of monitored levels of NO2 and PM2.5 reveals an inverse correlation with tree cover in winter (leaf-off) and summer (leaf-on) suggesting that canopy indicators represent lack of pollution sources rather than active pollution removal. Tree canopy deposition likely has at most a small impact on neighborhood air quality relative to emissions. Planners should emphasize a holistic view of the benefits of urban trees when prioritizing urban neighborhoods for tree planting.