The introduction of a new form of at-grade intersection control in the United States, termed a roundabout, has left many researchers and practitioners puzzled about their performance level. Many researchers and practitioners have looked to foreign design and operational manuals for guidance. Although the methods contained in these manuals have been implemented in their respective countries, no one is certain how they will transfer to conditions in the United States. Considering that driver characteristics are a major contributor to operational performance, these methods may not accurately depict the performance level of roundabouts with American drivers. Driver characteristics, in relation to operational performance, of four single-lane roundabouts located in the United States are described. The same driver characteristics are compared with findings in Australia under similar conditions; in addition, the probability density function for gap acceptance is derived. American drivers do not always react the same as Australian drivers under similar conditions. However, use of the Australian methods is, in most cases, more conservative and therefore should not overpredict the capacity and performance of roundabouts in the United States. The probability density function for gap acceptance at roundabouts was similar in shape and slope to that of two-way stop control. However, comparison of the gap-acceptance values of right-turning vehicles at two-way stop control with those at a roundabout indicates that drivers at roundabouts accept smaller gaps in the traffic stream on entry. This leads to the conclusion that roundabouts should perform better than two-way or all-way stop-controlled intersections under most conditions. The question remains: When do roundabouts function better than traffic signals?
PROBABILITY DENSITY FUNCTION OF WAVE HEIGHTS OFF THE WESTERN COAST OF TAIWAN
