Activity-based TOD typology for seoul transit station areas using smart-card data. (December 2022)

• Record Type: Journal Article
• Authors: Shin, Yonggeun; Kim, Dong-Kyu; Kim, Eui-Jin
• Is Part Of: Journal of transport geography. Volume 105(2022)
• Publication Date: 2022-12
• Subjects: Transit-oriented development -- Smart-card -- Activity identification -- Trip-chain -- Hidden Markov Model (HMM) -- Spatio-temporal modeling; Transportation -- Periodicals; Telecommunication -- Periodicals; Transport -- Périodiques; Télécommunications -- Périodiques; Telecommunication; Transportation; Periodicals; 388
• DOI: 10.1016/j.jtrangeo.2022.103459
• Languages: English
• ISSNs: 0966-6923

Abstract

Abstract: Transit-oriented development (TOD) is a planning strategy to encourage the use of public transportation by clustering urban development centering on transit stations with dense and diverse land uses. TOD typology is important information for the urban planner to monitor and improve the TOD. However, many studies have focused only on static features such as land use and population density around the transit station. Such analysis does not reflect dynamic features based on human mobility patterns. This study focuses on a series of processes that evaluate the activity-based TOD typology of the transit station based on smart-card data. To consider the context of multimodal transit system, transit station areas consisting of one subway station and multiple bus stops are created and clustered according to the proportion of the identified activities. Passengers' trip-chain with missing activities are clustered by applying a hidden Markov model (HMM), and the activities are identified through land use and temporal patterns. Interpretation of each cluster is made through spatio-temporal patterns. The results provide six types of identified activities that are well aligned with general expectations. The transit station areas in Seoul, which were clustered with six identified types, describe the urban spatial structure, which is spatially distributed around a business district in order of mixed and residential zones. Also, the same category of clusters shows different functions depending on temporal patterns, indicating the need for consideration of dynamic features. Our approach provides insights into transit station changes throughout the day that has been overlooked in previous studies by considering the passenger's sequence of activity, start, and duration.