With household travel surveys (HTSs) to produce a regional-scale policy model, this research develops a methodology for predicting context-based vehicle-trip reductions applied to ITE’s Trip Generation Handbook at a site-level development. This methodology may be used as a supplement to ITE trip generation rates, providing justification to vehicle-trip reductions based on known contextual vehicle mode splits. With the 2006 HTS of the Puget Sound Regional Council, Washington State, non-home-based trip ends were selected, and common built environment measures were extracted. A clustering analysis was then applied to all trip ends to determine clustered groups or contexts. With contexts, sociodemographic characteristics, and trip characteristics as model variables, a binary logistic model was developed to predict the mode split. Mode splits were then calculated for all context types (Type A through Type H) with the use of average built environment, sociodemographic characteristics, and trip characteristics variables. External establishment survey rates and mode splits, published from reports in California, were then used to verify the prediction power of the model on the basis of the type of establishment’s context. For each establishment survey, the same built environment measures were extracted, and the location was classified into contexts with a linear discriminate analysis. In general, mode splits predicted across contexts showed variation expected for areas with greater residential or employment density, land use mix, and connectivity. Establishment data showed that predicted values fell within the observed range or fell on the conservative side of estimation. Future applications of this research are discussed.
ASJC Scopus subject areas
- Civil and Structural Engineering
- Mechanical Engineering