Evaluation of Macro- and Mesoscopic Models for Congestion Impact Analysis

Congestion in urban arterial road networks increases steadily. For long-range road investment planning it is essential to be able to evaluate the impacts of this congestion on road user costs and on the environment. The current method is to use transport planning models with static traffic assignment. This method, however, may significantly underestimate the impacts of congestion because of the inability to model the extent of queuing caused by bottlenecks. An alternative method to analyze congestion impacts is to use dynamic mesoor microscopic models that are capable of representing the effect of queuing and the resulting delays. A study to look into these matters was undertaken in Stockholm in 1999. The study area included the arterial road network in the Stockholm region with some 2300 links. Traffic data was collected for morning peak flow conditions using aerial photography, floating cars and stationary flow and speed surveys. The traffic flow counts were used to update the trip matrix, after which traffic assignment for comparative analysis was made using the CONTRAM7 model as well as the macro model DSD-IRS. Finally the results from these runs were compared with the actual situation as observed in the field studies. The conclusions from the study was that the mesoscopic model was well suited to describe traffic performance on congested road segments, which was not the case with the macro models tested. A considerable effort is however required coding the network in sufficient detail to describe the capacity characteristics of the bottlenecks in the system.