THE CREATION OF POINT-BASED HOUSEHOLD AND JOB GEOGRAPHIC DATASETS FOR TRAVEL MODELING: DISCUSSION OF THE DEVELOPMENT PROCESS AND OF MODELING BENEFITS Erik Sabina, P.E., Regional Modeling Manager

The Denver Regional Council of Governments is conducting a project to develop a disaggregate, activity-based travel modeling system for use in regional planning. Many of the advantages of such models are derived from their disaggregate approach to decision-modeling, and the disaggregate demographic data required to support this approach. Population synthesis permits disaggregate specification of household and person characteristics. However, household and job location in the model typically are known only at the TAZ level. This limitation leads to aggregate specification of numerous key variables, such as transit walk access distance and walk mode distance, or requires the use of complex, non-intuitive approaches to reducing aggregation error through zonal subdivision. The project team concluded that it was practical to assign an x-y location to each household and job in the region, and that this approach would produce a variety of benefits both in model estimation, and in model application. Example benefits include simpler, more accurate calculation of short-distance trip lengths and transit access distances/times, which in turn significantly improve mode choice model outcomes for transit and non-motorized modes. Model scenario analysis also is greatly improved, as it is possible t