Impacts of the Urban Spatial Landscape in Beijing on Surface and Canopy Urban Heat Islands

How does the urban spatial landscape (USL) pattern affect the land surface urban heat islands (SUHIs) and canopy urban heat islands (CUHIs)? Based on satellite and meteorological observations, this case study compares the impacts of the USL pattern on SUHI and CUHI in the central urban area (CUA) of Beijing using the satellite land-surface-temperature product and hourly temperature data from automatic meteorological stations from 2009 to 2018. Eleven USL metrics—building height (BH), building density (BD), standard deviation of building height (BSD), floor area ratio (FAR), frontal area index (FAI), roughness length (RL), sky view factor (SVF), urban fractal dimension (FD), vegetation coverage (VC), impervious coverage (IC), and albedo (AB)—with a 500-m spatial resolution in the CUA are extracted for comparative analysis. The results show that SUHI is higher than CUHI at night, and SUHI is only consistent with CUHI at spatial–temporal scales at night, particularly in winter. Spatially, all 11 metrics are strongly correlated with both the SUHI and CUHI at night, with stronger correlation between most metrics and SUHI. VC, AB, and SVF have the greatest impact on both the SUHI and CUHI. High SUHI and CUHI values tend to appear in areas with BD ≥ 0.26, VC ≤ 0.09, AB ≤ 0.09, and SVF ≤ 0.67. In summer, most metrics have a greater impact on the SUHI than CUHI; the opposite is observed in winter. SUHI variation is affected primarily by VC in summer and by VC and AB in winter, which is different for the CUHI variation. The collective contribution of all 11 metrics to SUHI spatial variation in summer (61.8%) is higher than that to CUHI; however, the opposite holds in winter and for the entire year, where the cumulative contribution of the factors accounts for 66.6% and 49.6%, respectively, of the SUHI variation.