A Transit Ridership Model Based on Geographically Weighted Regression and Service Quality Variables

This report presents the development of Geographically Weighted Regression (GWR) models for predicting public transit use for home-based work trip purpose. A large array of potential transit use predictors were considered, including demographic, socioeconomic, land use, transit supply quality, and pedestrian environment variables. The best predictors identified through model estimation include two global variables, regional accessibility to employment and percentage of households with no car, and three local variables, employment density, average number of cars in households with children, and percentage of Black population. The models were estimated based on 2000 CTPP data for Broward County. Model testing showed significant improvement in model accuracy. The GWR models indicate that for local variables, their significance and influence vary by location. It is also observed that for this particular problem, an unexpected local sign of a variable is an indication of the insignificance or irrelevance of the variable at that location. This understanding of where a variable is significant and how significant it is at different locations is useful because it provides a basis for the development of transit service improvements by determining where improvements are needed and how much benefit the improvements will bring. The findings indicate that the local model estimation technique of GWR can help improve transit demand analysis and the identification of priority areas where transit service may need improvements. Reproduction of completed page authorized ii ACKNOWLEDGEMENTS This study was sponsored by a research grant from the Public Transportation Office of the Florida Department of Transportation (FDOT). The author would like to thank the Project Manager, Mr. Ike Ubaka, for his guidance and support throughout the research. We also thank Ms. Lina Kulikowski of Broward County Office of Urban Planning and Development for providing employment data and the Broward County Planning Council for providing land use data. The contents of this report reflect the views of the authors and do not necessarily reflect the official views or policies of the Florida Department of Transportation. This report does not constitute a standard, specification, or regulation. INTRODUCTION As congestion in urban areas continues to worsen and highway solutions become less effective, many local governments and communities are turning their attentions to public transit systems. To maximize the return on investments in public transit systems, effective analysis tools are needed to determine transit ridership potential and the benefits of investments. Such tools, including those for long-rang planning or short-term ridership prediction, …