Validation and Sensitivity Considerations for Statewide Models

The Transportation Research Board, at the request of the American Association of State Highway and Transportation Officials (AASHTO) Standing Committee on Planning, initiated a study on model validation and sensitivity analysis of statewide travel demand forecasting models in 2009. Cambridge Systematics, Incorporated was contracted to perform this study under National Cooperative Highway Research Project (NCHRP) 836-B Task Order 91. This paper describes research findings on the validation and conducting of sensitivity analyses on statewide travel demand forecasting models, based in part on the review of 30 statewide models. This paper is divided into five statewide model validation and sensitivity topics. The first section describes the market segments and trip purposes used in statewide models. The next section discusses approaches to validation, reasonableness comparisons, and ranges of statistics found in statewide models. This is followed by a section that examines modal issues in statewide models. Statewide freight models are discussed in the next section and a summary of research study findings is then provided in the last section. The resulting final report will act as a supplement to Federal guidance on model validation for urban models and highlight ways in which statewide models should be treated differently from urban models.