Influence of the proportion, height and proximity of vegetation and buildings on urban land surface temperature

Abstract Urban areas are characterised by the dominance of impervious surfaces and decreased presence of vegetation compared to their rural surroundings. The resultant increase in temperature is known to amplify global warming, with negative impacts on health and increased energy requirements for cooling. Intra-urban variations in temperature have received less attention than urban–rural variations, although the former can be even larger than the latter. Land cover composition is known to influence surface temperature, while the influence of heights, of buildings and vegetation, is less explored. There are also fewer studies in high-latitude cities although extreme heat events are increasing in frequency and severity in these cities, and high-resolution geospatial datasets are often available for detailed analysis. The aim of this study is therefore to assess the influence of selected land cover variables on the estimated surface temperature in the four largest cities in Denmark—Copenhagen, Aarhus, Odense and Aalborg. Land surface temperatures (LST) of the four cities were estimated using Band 10 (10.60–11.19 μm) from Landsat 8 imagery. Vegetation cover, building cover, vegetation height and building height were estimated using 4-band aerial imagery, building footprints and LiDAR-based elevation models, and their correlations with LST were estimated. Moving average filters, with window sizes from 3 × 3 (90 m × 90 m) to 11 × 11 (330 m × 330 m), were used to understand the area of influence of surrounding land cover on the LST within 30-m cells. When vegetation cover and building cover increased from 0–5% to 95–100%, median values of LST decreased by 4.16 ± 0.76 °C and increased by 4.31 ± 0.69 °C, respectively. Land cover variables within 7 × 7 windows (210 m × 210 m) are shown to have strong correlations with the LST of 30-m cells. The area of influence of building heights on the LST of 30-m cells was the largest in Copenhagen, which also has the tallest buildings among the cities. LST reduced by 4.10 °C when the mean vegetation height within a 30-m cell increased from 0–2 m to 20–22 m, and by 5.75 °C for 210 m × 210 m patches with the same height range. A combination of increased vegetation cover and height could therefore be used to regulate temperature in or close to hot spots in cities depending on the availability of space.