What Determines Mixed Land Use? Analysis of Big Data on Commercial Districts in Seoul, South Korea

Each downtown commercial district has a different degree of mixed land use. Some commercial areas are a huge cluster with monotonous retail use, while others consists of a wide variety of land uses including residential or office uses as well. Locational factors such as proximity to other amenities such as transit and public services have a great influence on land use mix (LUM). TOD around the railway station leads to mixed land use in the area due to efficient transit (Chen et al., 2016). Also, retail and food establishments demonstrate a certain location pattern in the urban area, making the neighborhood more diverse (Sevtsuk, 2014). A certain age or sex group of customers may also affect LUM in a commercial area due to their love of variety. Some young adults and the elderly prefer homes close to various urban amenities, and thus they tend to be attracted to a higher mix of land use (Blumenberg et al., 2019). This paper aims to find the causes of differences in LUM in terms of the location and socioeconomic environments of each commercial districts and the demographic characteristics of visitors and residents, especially focusing on public transit and automobile accessibility. To this end, we introduce a micro-scale approach wherein LUM is defined by the entropy index, utilizing data for more than 40,000 individual buildings, based on Rodriguez et al (2013)’s comparison of results using Monte Carlo simulation. The existing studies on the relationship between land use and human activity usually use data collected at a specific time by a survey or the aggregation of existing data. The fixidity of the time dimension of data makes it difficult to examine the dynamic changes of visitors to downtown areas. The real-time big data on visitors enable us to overcome this limitation. By processing mobile phone GPS data, visitor demographics are examined based on a real-time big data and comprise the Seoul Living Population (De Facto population counts with mobile phone GPS) from January 1, 2017 to December 31, 2018; we analyzed the distribution of visitors in the city of Seoul and their travel patterns in downtown commercial districts along the time dimension. We use the daily average of total visitors and residents for each commercial district, dividing them into young adults (20-34) and old adults (35-64) as explanatory variables, in addition to the locational and socioeconomic factors for each district. The expected results of the analysis are as follows: First, the total number of visitors increases the variety of LUM in a commercial district. Second, the LUM is relatively high in the commercial areas where the proportion of young adult (20-30s) visitors and residents is higher than those in other age groups. Lastly, higher transit accessibility increases the variety of LUM by reducing the need for parking space and prompting walkable streetscape. These results may help planners design and regulate land use for downtown commercial areas in terms of LUM in an appropriate way