Evaluation of Bicycle Network Connectivity Using Graph Theory and Level of Traffic Stress

Abstract

The quality of the bicycle network determines ridership, safety, connectivity, equity, and livability. Very few former research studies investigated network connectivity for individual user types and identify network needs and barriers based on these rider types. This study measures the network connectivity for different rider types using level of traffic stress (LTS) and graph theory concepts. As a symbolic representation of a road network and its connectivity, a graph represents the structural properties of networks and compares one measure over another by taking into account spatial features. In addition, this study defines a bicycle network for different types of riders using LTS metrics based on traffic speed, road geometry, and traffic volume. This study evaluates the OpenStreetMap (OSM) bicycle network for Portland, Oregon, as a case study. Three transit stations in the downtown, riverside, and residential area were considered to assess the connectivity and barriers with a home at block level for last and first-mile coverage. The analysis shows that 29% of links in Portland need to be improved with more bicycle facilities to provide access to basic adult riders, and 33% of links require improvement for children. The networks are well connected for “strong and fearless” and “confident and enthused” users but not well connected for basic adults and children in many neighborhoods with low alpha and grid tree pattern (GTP) indices. The results indicate that planners and designers need to improve their network connectivity for all types of users to ensure equal active transportation opportunities beyond a particular portion of the network.Practical ApplicationsIn general, a well-connected network is important to provide the shortest route from origin to destination and safe traveling paths for all ages of people. It is critical for cities or government agencies to understand how their network is connected to different users because this knowledge will provide a fundamental basis for resource prioritizations on bicycle network improvement. This study developed a strategy using traffic stress and geometric properties of the network to assess their network connectivity. Practitioners can apply these techniques on a small scale (e.g., around transit stations) as well as large scale (e.g., entire city network) to identify the network connectivity. This study extends the applications to evaluate transportation equity in bicycle networks using served/ unserved populations where disparities in network connectivity exist to favor higher-income people.