TBEST Model Enhancements : Parcel Level Demographic Data Capabilities and Exploration of Enhanced Trip Attraction Capabilities

2011PDFTech ReportFDOT BDK85 977-05DTRT07-G-0059Public transitTrip generationSimulationDemographicsPublic transit, transit modeling,ridership, trip generationResearch HubFloridaUnited StatesFlorida. Dept. of TransportationPolzin, Steven EdwardChu, XuehaoBunner, RodneyPinjari, AbdulCatala, MartinNational Center for Transit Research (U.S.)Institute Center for Urban Transportation ResearchFlorida. Dept. of Transportation. Public Transit OfficeUnited States. Dept. of Transportation. Research and Innovative Technology AdministrationUS Transportation CollectionFDOT, in pursuit of its role to assist in providing public transportation services in Florida, has made a substantial research investment in a travel demand forecasting tool for public transportation known as Transit Boardings Estimation and Simulation Tool (TBEST). TBEST incorporates supporting databases that allow users to model transit services for purposes of determining future needs and optimizing current resource deployments by targeting the best markets and route configurations. This research effort is designed to explore enhancements to TBEST to increase its predictive capability and further enhance its value to transit planners. Two key and related areas are targeted. First, the project explores model calibration with parcel-level data. This involves increasing the geographic precision of transit ridership modeling by using parcel-specific data on land use to understand the activity at the parcel level, and hence, the potential for transit ridership. Second, the project explores strategies to more robustly address the issue of special generators. The project determined that transitioning to a parcel-based model is a promising improvement for TBEST. It enables a more precise capturing of the accessibility of transit stops, which has been shown to be critical to transit use. In addition, it accommodates a shift to a trip production/attraction-based data framework that enhances the information on which one can base a transit forecast. In summary, increased computing power, improved databases such as the parcel property inventory, and a strong understanding of factors that influence transit use have enabled the development of more powerful tools to support transit planning. While transit ridership remains highly variable at the stop level and hence difficult to model, great strides are being made, and the full deployment of parcel-level transit models seems inevitable as a logical advancement in the state of the practice