Consideration should be given to the scope, location, and the nature of the project when selecting a model to be used for Project Traffic Forecasting. If a project and its influencing area lie completely within an urbanized MPO/TPO/TPA area, the adopted MPO/ TPO/TPA model should be used unless there is a good reason to use a different model and all involved parties reach an agreement before the project starts. If a project and its AOI lie outside or cross MPO/TPO/TPA planning boundaries, a regional model that covers the entire project lies in a rural area or an urban area not covered by an MPO/TPO/TPA or regional model, the Florida Statewide Model or Florida Turnpike Model can be used as a starting point to develop a subarea model. In addition to the system-level regional LRTP models, some Districts also develop and maintain project specific models that can be customized for the project at hand. These could be models validated to a different base year, having features that allow for evaluating different travel options, such as toll roads or transit services that have direct impact on the project, or including a different horizon year with updated model input data and transportation network. Using these models could significantly reduce the time and costs for modeling work. District Planning Offices should always be consulted regarding the availability, capability, and applicability of these models. The Florida’s modeling web portal has a list of available models for all MPO/TPO/TPAs and districts in the State and can be used as a source of information for this purpose.

Review of Model Applicability

Users should verify that the latest version of the model is obtained and conduct a review of the base year validation and forecast year projections within the project study area. This is to determine if the model reasonably reflects the current travel conditions and whether the projections are consistent with the expected growth in population and employment. If the level of accuracy in the base year model is deemed to be unacceptable for the purposes of forecasting traffic for a project, then the model should not be used until the District Planning Office and/or the agency having jurisdiction over the model has addressed the situation.

Models are generally calibrated on a systemwide level and not on a corridor or project specific level. The Project Traffic Report stage is NOT the appropriate place to perform a recalibration of the base year model. Should the calibration of the model remain an issue, it is suggested that the procedure in Chapter 5, Forecasting Without a Traffic Model, be utilized instead.

Model Applicability Revision

All models used for Project Traffic Forecasting must be approved by the District Planning Manager or his/her designee and determined to be suitable for forecasting traffic for the project. The suitability check should include Percent-Root-Mean-Square-Error (%RMSE) and screenline volumes in base year evaluations. If the model is acceptable, perform project refinement; if not, perform historical trend analysis comparison.

Model Refinement

After the initial review of the model to verify its usability for the project, further refinement of the model is usually needed. Model refinement involves correcting any errors in the socioeconomic data and model network, adding more details that are not included in the regional model by splitting traffic analysis zones (TAZs), coding more local facilities into the network, and creating new centroid connectors. In some cases, further adjustments to the model parameters are needed to produce a better match between the model outputs and observed conditions within the study area. However, adjustments can only be made with supporting evidence that demonstrates the implied travel behavior. Adjustments made to the model should also comply with the established FSUTMS standards and should be fully documented. This document should then be reviewed with the District Planning Office and the agency responsible for the model to obtain consensus on the results.