Document Type

Journal Article

Publication Date


Subject Area

operations - traffic, infrastructure - vehicle, ridership - mode choice, ridership - demand


Trucks, Trip tables, Trip generation, Trip distribution, Travel time, Travel models (Travel demand), Travel demand, Traffic models, Traffic generation, Traffic assignment, Traffic analysis zones, SOVs, Single occupant vehicles, Road freight vehicles, Multiple occupancy vehicles, Mode choice, Modal choice, Mathematical models, Lorries, Logits, Logit models, Land use models, Journey time, High occupancy vehicles, Heavy goods vehicles, Costs, Choice of transportation, Census tracts, Case studies, Cargo vehicles, Atlanta Regional Commission


The Atlanta Regional Commission (ARC) regional travel demand model is described as it relates to its link-based emissions postprocessor. In addition to conformity determination, an overview of other elements is given. The transit networks include the walk and highway access links. Trip generation addresses trip production, trip attraction, reconciliation of productions and attractions, and special adjustments made for Hartsfield Atlanta International Airport. Trip distribution includes the application of the composite impedance variable. In the mode choice model, home-based work uses a logit function, whereas nonwork uses information from the home-based work to estimate modal shares. Traffic assignment includes preparation of time-of-day assignments. The model assigns single-occupancy vehicles, high-occupancy vehicles, and trucks by using separate trip tables. The procedures can accept or prohibit each of the three types of vehicles from each highway lane. Feedback between the land use model and the traffic model is accounted for via composite impedances generated by the traffic model and is a primary input to the land use model DRAM/EMPAL. The land use model is based on census tract geography, whereas the travel demand model is based on traffic analysis zones that are subareas within census tracts. The ARC model has extended the state of the practice by using the log sum variable from mode choice as the impedance measure rather than the standard highway time. This change means that the model is sensitive not only to highway travel time but also to highway and transit costs.