Several issues related to the upper branch of the speed-flow curve are addressed using data gathered at a freeway bottleneck in toronto. Some important findings evolve, some of which challenge conventional beliefs. At low to moderate flows, speed on the upper branch is found to be insensitive to flow, a notion that is becoming increasingly accepted; at higher flows, the data suggest that speed decreases with increasing flow, but the fall-off is not nearly so precipitous as is commonly thought. The data further suggest that the presence or absence of an upstream queue is a more important variable than flow for predicting freeway speeds. Perhaps the most important finding is that belief in a precipitous speed drop may very well have resulted from a misinterpretation of data that arises because the speed of vehicles discharged from a queue varies with location in the bottleneck.
Abstract