Level of Traffic Stress (LTS) is a four-level system that classifies the stress experienced by cyclists on road segments and at intersections. While LTS has been used in past studies to assess cycling connectivity, accessibility, and safety, very little is known concerning its influence on cycling preferences. This study investigates this topic using a dataset containing 323,163 unique GPS trajectories of Hamilton Bike Share (HBS) users collected over a 12-month period (January 1st to December 31st, 2019). A GIS-based map-matching algorithm is used to generate users’ routes from these trajectories along with attributes such as route length, number of intersections, and number of turns. Unique routes and their use frequencies are then extracted from all routes. The most popular routes between bike share hub (station) pairs are then identified as dominant routes while shortest distance routes are derived by minimizing distance traveled. Weighted level of traffic stress (WLTS), a novel measure of impedance (travel cost) developed for this study, is used to derive the least stressful routes between hub pairs. The three types of routes are compared statistically. The comparison finds that HBS users tend to choose longer routes with bicycle infrastructure in an effort to reduce their traffic stress. However, they do not choose to minimize traffic stress in its entirety by choosing the lowest WLTS routes. In other words, dominant routes are not the least stressful routes in a bike share system. Likewise, minimizing distance is not the sole consideration of HBS users. The findings suggest that other factors also influence route choice. This study not only enhances our understanding of cyclist route preferences with respect to LTS, it also presents a novel measure of impedance – WLTS – that could be used when planning new cycling infrastructure or as an alternative means to route cyclists between origins and destinations. / Thesis / Master of Science (MSc)

This thesis examines the existence of a dominant route between a hub pair and factors that influence bike share cyclists route choices. This research collects 132,396 hub to-hub global positioning system (GPS) trajectories over a 12-month period between April 1, 2015 and March 31, 2016 from 750 bicycles provided by a bike share program (BSP) called SoBi (Social Bicycles) Hamilton. Then, a GIS-based map-matching toolkit is used to convert GPS points to map-matched trips and generate a series of route attributes. In order to create choice sets, unique routes between the same hub pair are extracted from all corresponding repeated trips using a link signature tool. The results from t statistics and Path-size logit models indicate that bike share cyclists are willing to detour for some positive features, such as bicycle facilities and low traffic volumes, but they also try to avoid too circuitous routes, turns, and steep slopes over 4% though detouring may come with a slight increase in turns. This research not only helps us understand BSP cyclist route preferences but also presents a GIS-based approach to determine potential road segments for additional bike facilities on the basis of such preferences. / Thesis / Master of Science (MSc)