Sustainable Cities and Society, cilt.101, 2024 (SCI-Expanded)
The composition and configuration of the built environment affect intra-urban heat variability and human thermal exposure. We investigated how 2-D and 3-D building and vegetation characteristics, represented by various landscape metrics, affect the mean radiant temperature (MRT) distribution in Phoenix and Tempe, Arizona, USA, to determine which urban form characteristics are most important for minimizing thermal exposure. For a hot-dry summer day with low wind speed and a maximum air temperature of 43.3 °C, the relationship between 1-m resolution MRT data and seventeen 2-D and 3-D landscape metrics were analyzed at the Census block (micro) and urban (macro) scale. The landscape metrics were calculated with FragStat from a Digital Surface Model, a 3-D point cloud obtained from high-resolution (0.5-m) USGS LiDAR data, and a 1-m resolution land use/land cover map classified from the 2015 National Agriculture Imagery Program (NAIP) data. Hourly MRT for June 27, 2012, was simulated using the SOlar LongWave Environmental Irradiance Geometry (SOLWEIG) model. After testing for autocorrelation in hourly MRT using Moran's I, the relationship between landscape metrics and MRT outcomes was analyzed using correlation coefficients and multiple linear regression. The best predictive power was achieved using 2-D and 3-D metrics together, with an explanation MRT of 86 % (8:00 h) to 96 % (12:00 h). The five most important factors were 3-D vegetation height, 2-D percent tree surface cover, 3-D building height, 2-D building edge density, and the 3-D vegetation height coefficient of variation. Results show that MRT is driven by the composition and configuration of 2-D and 3-D urban features. The horizontal arrangement impacts MRT through varying land cover, and the vertical extent influences shade patterns, with more complex urban forms providing more shade and lowering MRT. Findings advance our understanding of how urban design can reduce thermal exposure on hot days using passive cooling strategies that rely on changes in the configuration and composition of landscapes.