Evaluation of the safety impact of access management in urban areas

The access provided by streets and highways to adjacent lands are managed by controlling the spacings between the access points including signals, driveways, and media openings on mid-block segments, and setting the limit on the corner clearances around intersections. There have been studies on evaluating the impact of access management techniques on safety and mobility in urban areas. Samples of mid-block segments and intersections can be collected from selected arterials. Because the midblock segments or intersections in the same arterials share the same missing information, safety and mobility on them show unique features that should be taken into account when modeling.In this study panel data models were proposed for safety analysis on mid-block segments and intersections. A virtual "mid-block" segment was assumed to exist for each arterial. The observations of the mid-block segments on this arterial were viewed as repeated observations for the virtual "mid-block" segment. This perspective of the midblock segments or intersections over space made it feasible for the panel data model to evaluate the impact of access management techniques on safety. In addition, this study also recognized that interdependency existed between safety and mobility for a mid-block segment or an intersection. Therefore, for mid-block segments, simultaneous equation models were adopted by integrating with the panel data modeling structure. For intersections, the interdependence between safety and mobility wasn't considered due to the lack of data, and only count data models combining with the panel data structure was developed.Data were collected from different sources for the urban areas of Southern Nevada. The results from the models for mid-block segments indicate that there is a strong interdependency between safety and mobility. The length of mid-block segments, driveway density, and median opening density are very significant factors that influence crash rate on mid-block segments. From the results of the models for intersections, it was found that corner clearance significantly influenced the number of crashes occurred at intersections. Other factors also influence the occurrence of crashes at intersections that include land use, traffic flow, number of lanes, and posted speed limit.