Recent research has suggested that the discharge rate of traffic across the stop-line of signalized intersections is not constant (a conventional assumption when deriving saturation flow), but that it varies over the green time. Different studies have shown very different patterns of variation, but unfortunately the reasons for these differences have not been fully explained. This study compares previous results on the variability of the discharge rate over the green time, and then attempts to explain why different patterns are produced by different situations. Secondly, in order to give a more comprehensive picture of this phenomenon, a study is undertaken at four signals in a developing city ? Hanoi, Vietnam. It appears that the pattern of the discharge rate at signalized intersections depends on: (1) traffic composition; (2) proportion of turning vehicles; (3) length of green time; (4) type of operating signal control; (5) the number of vehicles actually stopping behind the stop-line; and (6) those simply decreasing their speed before crossing the stop-line. Furthermore, it seems that the period of time from the onset of green to the moment the discharge rate reaches a temporary stability or maximum value depends much on the proportion and manageability of the predominant type of vehicles at the head of the queue.
Abstract