Journal of the Korean Cartographic Association ISSN:1598-6160(Print) 한국지도학회지

COVID-19 is a respiratory disease that is likely to be transmitted through respiratory droplets and aerosols. The indoor space of the building is very vulnerable to the broad spread of COVID-19, and urban environment that many people are densely populated at high-rise buildings due to the intensive land-use may be more vulnerable to these diseases. Moreover, since the population distribution in the urban area changes over time dynamically, understanding of the spatio-temporal change of urban population is a key for the epidemiological investigation of COVID-19. However, it is difficult to examine the de facto population resided in a specific building at a specific time period. Therefore, this study attempts to suggest an alternative way of estimating the horizontal and vertical distribution of urban population at a particular time period. More specifically, we estimate the de facto population at a building scale using a dasymetric mapping method based on geographically weighted regression (GWR) model. In general, a dasymetric mapping enables us to estimate more detailed population at a finer scale using ancillary data. In this study, building usage and total floor area are utilized as ancillary data and heterogeneous distribution of population is incorporated by the GWR model. As a result of analysis, we can suggest the detailed population estimation at the building scale finer than statistical enumeration units over the entire area of Seoul. It is expected that the estimated de facto population at the building scale would be utilized for important fundamental information for the epidemiological surveillance of pandemic diseases like COVID-19 and the establishment of the effective prevention plan of epidemics.