Preparing transit forecasts for New Starts and Small Starts applications can require a rigorous review of a travel model to ensure that all model steps are well validated and functioning correctly. For smaller transit projects, a comprehensive travel model update may not be feasible, especially for preliminary forecasts. This presentation reports on an alternate approach that can be used in such cases.
VIA, the transit agency in San Antonio, TX is considering a high capacity transit corridor that may be submitted as a New Starts or Small Starts project. The agency has a recent on-board transit survey available, but the regional travel model would need several revisions to produce transit forecasts appropriate for use in an FTA submittal.
The agency decided on an approach that is based on an observed transit trip table generated from survey data. The observed data are used to support an incremental logit model that estimates change in transit share as a function of changes in transit service. All other variables such as auto travel time and parking costs are held constant. With guidance from the FTA, the method can accommodate alternative-specific effects including station amenities such as lighting, shelter, and real-time information.
Use of this alternative approach allowed model validation to focus on transit model components, resulting in an accelerated schedule. Transit networks were validated using both aggregate and disaggregate validation statistics such as boarding totals by route and stop, transit path prediction success, and transit path distance. The need to review and validate other model components was minimized.
While this approach has several benefits, some challenges were also encountered. The area surrounding the proposed corridor experiences heavy pedestrian activity, some of which could be served by transit. Methods of including these non-motorized trips in the analysis were explored. The observed trip table proved to be very sparse, requiring some aggregation of data from TAZs to districts. Another concern was introduction of transit to zone-pairs where transit is not currently available. This was not an issue in this particular implementation, as most corridors under consideration would serve areas with existing transit access.