Land surface temperature (LST) and outdoor air temperature (OAT) have been widely used to investigate urban heat island (UHI) phenomena. Although LST and OAT were reported to show a substantial correlation, whether urban factors play a consistent role in both LST and OAT has not been well addressed in the existing literature. To fill this gap, this study examined the effects of urban factors on both LST derived from satellite image data and OAT measured at the pedestrian-level through a case study in Seoul, Korea. Two different temperature datasets were used for the same days in summer and winter, and corresponding urban spatial characteristics were calculated on the basis of geographic information system (GIS) data. The effects of urban variables were analyzed using the Genizi and partial correlation methods. The results from the Genizi method revealed that land use characteristics (i.e., building coverage and pervious ratios) were important factors influencing LST, whereas urban morphological characteristics (i.e., building height) had a greater influence on OAT. The partial correlation results confirmed these separate sets of urban factors above are key factors for LST and OAT. In addition, the partial correlation analysis demonstrated that the pervious ratio showed a negative correlation with LST and the building height consistently showed a significant influence on OAT in both seasons. In contrast, the canyon ratio showed a contrasting effect on OAT depending on the season, underscoring that urban factors can affect temperatures differently depending on the meteorological conditions. This study highlights the different effects of urban factors depending on temperatures.
Read full abstract