Concordance between Regional Functions and Mobility Features Using Bike-sharing and Land-use Data near Metro Stations

Research has revealed various relationships between regional functions, land use, and mobility features. However, how or to what extent the planned mobility features of designed regional functions differ from observed mobility features has not been sufficiently investigated. Here, the concordance between regional functions and mobility features was evaluated with datasets of dockless bike-sharing (DBS) usage and land use near metro stations. The K-means cluster method was applied to land use for capturing regional functions and to DBS features for identifying DBS usage patterns. The Kappa statistic for testing the concordance was proposed to show the differences between regional functions and mobility features. The results show poor concordance between regional functions and observed DBS features under the transit-oriented development context; the Kappa value was 0.09 on average for both DBS attraction and production. The concordance was found to be vulnerable to moderation by regional functions, on which commercial and business centers showed fair (Kappa = 0.20) and moderate (Kappa = 0.40) concordance with DBS attraction and production, respectively. The findings revalidate known relationships between land-use and mobility features based on mobility performance, and can assist in policy interventions to achieve the coordinated development of regional land-use and transportation systems.