Doctor of Philosophy (PhD)
Civil and Environmental Engineering
First Committee Member
Hualiang (Harry) Teng
Number of Pages
The goal of this study was to address one of the major weaknesses of the ubiquitous four-step procedure for travel demand modeling: omission of spatial interactions between the variables. While contiguity of the analysis zones is commonly used to define spatial interaction of the variables in spatial analysis, it might not capture the interactions of travel demand variables. In this study, the efficacies of four alternative methods for defining spatial relationships: contiguity, separation, a combination of contiguity and separation, and economic linkages (accessibility), were evaluated. The home-based-work (HBW) spatial models and non-spatial models for trip attraction, and trip production were developed. For the destination choice, the spatial models were developed by using separation and accessibility alternatives for defining spatial relationship. Comparison of the trip attraction models indicated that the model estimated using the separation spatial relationship had the best fit. Furthermore, comparison of the best spatial model and the non spatial model indicated that the spatial model outperforms the non spatial model by increasing the prediction accuracy by 14%. For the trip production model, the results indicated that the spatial variable is unnecessary. For destination choice, the spatial model developed using separation spatial relationship was found to he the best based on statistical tests. To compare the spatial model and the non-spatial model, the forecasted alternative destination shares were used. The results indicated that the difference between the forecasted alternative shares by using spatial and non-spatial models is small when there is a small percentage increase in casino/hotel and retail jobs. In order to use the developed destination choice models for long-term forecasting, additional variables such as housing location should be included. Also, since the design of the analysis zones used in this study may not be optimal, an attempt to design new analysis zones through a careful aggregation process in which homogeneity is carefully controlled, is recommended.
Characteristics; Demand; Destination Choice; Incorporating; Models; Spatial; Travel; Travel Demand; Trip Generation
University of Nevada, Las Vegas
If you are the rightful copyright holder of this dissertation or thesis and wish to have the full text removed from Digital Scholarship@UNLV, please submit a request to firstname.lastname@example.org and include clear identification of the work, preferably with URL.
Kwigizile, Valerian, "Incorporating spatial characteristics in travel demand models" (2007). UNLV Retrospective Theses & Dissertations. 2770.