Diurnal and Seasonal Variations in the Effect of Urban Environmental Factors on Air Temperature: A Consecutive Regression Analysis Approach.

This study investigates the diurnal and seasonal variations in the effect of environmental features on air temperature in Seoul, Korea. We expect that this study will lead to the identification of factors that can be applied for urban heat island mitigation strategies in summer without leading to an unintended result in winter. As our dependent variable, we employed the smoothed 31-day moving average of air temperatures, where we controlled the seasonal variation by normalizing the values observed from 247 automatic weather stations (AWS) from 2015 to 2016. Subsequently, we conducted consecutive log-log regression analyses of each day to examine patterns of change in regression coefficients and the significance of each independent variable. For independent variables, we applied built environment features including albedo, land-use, average building floors, the sky view factor, and green and water areas. This study provides analytical results regarding the relationship between environmental factors and air temperature. This study also addresses imperative issues for planners, especially regarding albedo, wind path, building geometry, and land use types. Finally, this study gives useful insights for managing the diurnal and seasonal variations of urban thermal environment in the mega-city.