SRTC currently uses a traditional trip-based four step model on the PTV VISUM (version 12) software platform. Previously, SRTC used TMODEL and then transitioned to EMME/2 specifically to support Federal Transit Administration (FTA) New/Small Starts analysis. The most recent regional travel survey was conducted in 2005 with a sample size of more than 1,200 households. The model was most recently calibrated and validated in 2012 for the 2010 base model. When the 2010 base model was built, it was enhanced to include more robust transit assignment capabilities, trip generation rates were recomputed, model run procedures were streamlined, a combined bike/walk mode was added to the nested mode choice structure, and the volume delay function was updated among other refinements. An assessment of the performance of the 2040 forecast model was also conducted during the development of Horizon 2040, the region’s long range regional and metropolitan transportation plan, which was approved by the SRTC Policy Board in 2013. SRTC also has forecast year models for Years 2020 and 2030.
There are 565 transportation analysis zones (TAZs) including 34 external zones and 12 park & ride zones in the SRTC model. The model has the following land-use and employment categories:
SRTC does not have a land-use model. The land-use and population forecasts are developed via a manual process of allocating population growth based on discussions with local agencies. There is not one agency that does coordinated socio-economic forecasts. Employment is assumed to grow at the same rate as population.
The model includes the following trip purposes:
The periods included in the model procedures are AM (6am-9am), Mid-day (9am-3pm), PM (3pm-6pm) and Night (6pm-6am) with two peak hours AM (7am-8am) and PM (5pm-6pm). The trip distribution model uses a gravity model formulation with no k-factors.
The mode choice model is a nested logit model with auto, transit, and non-motorized travel at the upper level nest. The lower level nest has drive alone and shared ride grouped under auto, and walk access and drive access grouped under transit. Currently, the only transit available in the region is local bus. There is a placeholder for a light-rail mode also included in the model. This placeholder creates a third-level nesting structure by splitting walk access into walk to bus and walk to light-rail and splitting drive access into drive to bus and drive to rail.
The model’s vehicular network contains I-90, U.S. highways, state routes, principal arterials and a few minor arterials, collectors and local roads as necessary. In the 2010 model there are about 2 million person trips and 1.4 million vehicular trips on an average weekday with 8.4 million vehicle miles traveled (VMT) and 227,000 vehicle hours of travel (VHT). The model’s transit network is comprised of 38 fixed bus routes with about 1,700 bus stops and 41,000 unlinked passenger trips per day. There are five express routes, one of which is primarily utilized by university students. A combined bike/walk mode is incorporated in the nested model choice model, but biking and pedestrian trips are not assigned to a network.
SRTC is currently well positioned for an update to the regional travel demand model. A major focus of the agency recently has been developing a performance management system to implement Horizon 2040, called the Horizon 2040 Implementation Toolkit, while meeting MAP-21 requirements. The model will play an integral role in informing this toolkit. The next iteration of Horizon 2040 will be updated by December of 2017 and the model will be used for scenario evaluation, project prioritization and other analysis as necessary. An updated travel survey and enhanced model are important to support the development of the plan and other SRTC activities. SRTC wants to enhance the travel demand model while continuing to address on-going needs. SRTC is considering the following model updates, but understands that not all of these enhancements are necessary or even desirable:
The primary reasons that SRTC is pursuing a model peer review are to:
