New York counted every one of its street trees in 2015. Satellites measure how hot the city's surfaces get on summer mornings. Put the two together, neighbourhood by neighbourhood, and the question is how much the trees cool the streets around them.
FindingAcross 188 New York neighbourhoods, each ten points of tree canopy goes with surfaces about 0.7 °C cooler in summer; the count of street trees alone barely predicts heat.
How to citeAl, Stefan. 2026. “Shade and Heat.” Formlume. https://research.stefanal.com/street-trees/
DataNYC Parks: 2015 Street Tree Census; USGS Landsat 8/9; © ESA WorldCover project 2021, contains modified Copernicus Sentinel data (2021) processed by the ESA WorldCover consortium. Please also cite the original data when you use these figures.
UseFigures on white may be reused for non-commercial purposes with credit (CC BY-NC 4.0). The page design is © Stefan Al, all rights reserved; for commercial use, please get in touch.
ThemeHeat and climate
Trees. NYC Parks, 2015 Street Tree Census, living trees only. Street trees are those along streets, not in parks or yards. Canopy is the share of each neighbourhood classed as tree cover in ESA WorldCover 2021 (10 m), which includes park and backyard trees.
Heat. Landsat's thermal band is measured at 100 m and resampled to 30 m, so each value blurs over a block or two; neighbourhood averages are little affected. Landsat 8 and 9 surface temperature (USGS) on eight clear summer days, 2022–2025, via Microsoft's Planetary Computer; each scene expressed relative to the city's average land temperature, then averaged. These are roof and pavement temperatures around 10:30 a.m., not air temperatures.
Neighbourhoods are the 2010 Neighborhood Tabulation Areas used by the tree census, assembled from census tracts; parks, cemeteries and airports (codes ending 98 or 99) are drawn but left out of the charts.
Reading it. Canopy and heat are associated across neighbourhoods; this does not isolate the cooling effect of trees from that of everything that comes with them, such as lower density, more lawns and fewer flat roofs.
Tools. Python (Rasterio, NumPy, pystac on Microsoft's Planetary Computer) for the rasters; D3.js for the charts.