Spatial and Temporal Effects of Built Environment on Urban Air Temperature in Seoul City, Korea: An Application of Spatial Regression Models

Abstract

In this study we examined the relationships between the built environment and urban air temperature in Seoul city, Korea. We developed multivariate regression models that address the relationship between built environment characteristics and the ambient air temperature with spatial statistics techniques. In addition, we analyzed the difference in daytime and nighttime air temperature to identify the built environment characteristics that affect the intensity of the nocturnal urban heat island effect (UHI). The large sample size of AWS locations in Seoul makes it possible to analyze the factors that influence ambient air temperature and UHI effect. The analysis results indicate that the sky view factor (SVF) and gross floor area significantly influence the daytime air temperature, while the building coverage and albedo showed strong relationships with the nocturnal air temperature. This study also demonstrated the importance of advanced spatial statistics techniques that control spatial autocorrelation and spatial heteroscedasticity in urban air temperature research. Our models confirmed the need to capture the effects of spatial autocorrelations within our spatial data. The findings of this study are valuable for understanding the complicated associations between the built environment and urban air temperature and to develop public policies to mitigate UHI effects.

Keywords

• air temperature
• built environment
• urban heat island effect
• automatic weather stations
• spatial statistics