This dataset maps forest patches across the Chicago region using a 2010 land cover layer (University of Vermont 2016). First, the land cover layer was converted to a binary raster of tree and no tree and resampled to a 3-m resolution. Then, roads from a land use layer (CMAP 2015) were masked from the canopy raster to ensure that no forest patches extended over paved surfaces. Forest cores and edges were identified by applying a morphological spatial pattern analysis (MSPA) (Soille and Vogt 2009) from Guido's Toolbox (Vogt and Riitters 2017). Methods were adapted from Alonzo et al. (2021). Briefly, cores were specified as pixels that were surrounded by 15 m of forest edge on all adjacent and diagonal pixels. The 15-m edge threshold was set as compositional and structural changes have been observed at this distance (Baker In Review). Additionally, this threshold allowed for the detection of relatively small forest patches, as the smallest patch can be a single core pixel surrounded by 15 m of edge--a 16.5 m radius circle (0.021 ha). The output from the MSPA were then converted to edge and core polygons, and other output types (e.g., connectors) were omitted. Citations: Alonzo, M., Baker, M. E., Gao, Y., & Shandas, V. (2021). Spatial configuration and time of day impact the magnitude of urban tree canopy cooling. Environmental Research Letters, 16(8), 084028. Baker, M. E. (In Review) Distributed urban forest patch sampling detects edge effects and woodland condition for monitoring and management. CMAP. 2015. Chicago region land use inventory. Soille P. and Vogt P. 2009. Morphological segmentation of binary patterns. Pattern Recognit Lett 30: 456–9. University of Vermont. 2016. Chicago regional land cover dataset. Vogt P. and Riitters K. 2017. GuidosToolbox: universal digital image object analysis. Eur J Remote Sens 50: 352–61.