Transferability analysis of built environment variables for public transit ridership estimation in Wuhan, China

Zhixiang Fang · Lupan Zhang · Meng Zheng

8 citations0/0 files storedTransactions in Urban Data Science and Technology

No readable copy stored yet — this record is metadata only.

No open-access location is known for this record yet.

Literature note Cite in a project Citation map

Urban Transport and Accessibility Transportation Planning and Optimization Urban Design and Spatial Analysis

Abstract

New stations (such as metro stations) will bring remarkable changes to the local transportation and economic development. Understanding patterns of factors which importantly impact on public transit ridership in the surrounding areas of new stations is essential to their construction planning, like estimating the possible ridership. Built environment variables with high importance magnitude, which were thought applicable to estimate public transit ridership in other areas of the same category, were described as transferable variables (TVs) in this study. A transferability analysis method of the built environment for the ridership estimation was constructed by adopting partial least square regression (PLSR) based on available data. Taking Wuhan, China as an example, this study analyzed the changes and differences of the built environment variables in different categories of pedestrian catchment areas (PCAs) of metro stations on the importance and transferability magnitude for the metro and taxi ridership, based on the metro and taxi data of one week in January, April, and June. Performances of the ridership estimation based on TVs and all the built environment variables were compared. This study inferred that (1) most of the land use variables (about 85%) showed important influence on the metro and taxi ridership, while only about 18% of the other variables showed key impact. The importance magnitude of the built environment variables was mainly related to PCA categories and public transportation modes, but less related to time. (2) Highly important built environment variables also tended to be highly transferable. Transferability magnitude of the built environment variables for the ridership was related to PCA categories and types of public transport. (3) Compared to all the built environment variables, using TVs, the relative accuracy of the metro and taxi ridership estimation was around 20% and 18% higher respectively.

Paper brief

Loading…

Cite this work

Citation unavailable for this work.

BibTeX RIS CSL-JSON

No public copy is stored for this record, so only the metadata is cited.

Citations in this library

Citation map →

Stored only when both papers are in the corpus — 8 citations exist upstream.

References

2017Multiscale Geographically Weighted Regression (MGWR)2001Activity-based disaggregate travel demand model system with activity schedules 2015Geographical and Temporal Weighted Regression (GTWR)2004Factors influencing light-rail station boardings in the United States 2007The Four-Step Model 2012Application of geographically weighted regression to the direct forecasting of transit ridership at station-level 2011Transit ridership forecasting at station level: an approach based on distance-decay weighted regression 2018A geographically and temporally weighted regression model to explore the spatiotemporal influence of built environment on transit ridership

Cited by

2023A Model Assembly Approach of Planning Urban–Rural Transportation Network: A Case Study of Jiangxia District, Wuhan, China 2025A deep reinforcement learning approach for real-time bus operation control using departure timetabling, stop-skipping, and re-routing

Explore the map →

Papers that cite the same references — the relational fallback while embeddings are still backfilling.

Imported from OpenAlex (CC0) — nothing here is generated.