Arterial model networks typically capture both one signalized intersection upstream and one signalized intersection downstream of the study intersections. If signalized intersections are not present, such as in rural areas, it is recommended that the network capture approximately half a mile of the arterial in both the upstream and downstream directions. Visual inspection of field data is used to determine if more of the roadway network will be included in the analysis. Like freeways, all intersections with significant influence on the arrival patterns of vehicles entering the network are included in the model. Additionally, it is recommended that all resulting congestion during the peak period be captured within the study limits. If congestion is identified on a network boundary, extend the study limits to capture the entirety of the queue. If extending the study limits is no longer practical, additional information on capturing unmet demand is provided in

Section 13.6

.