Connectivity for Cyclists? A Network Analysis of Copenhagen's Bike Lanes

英语. Cycling has been identified as a central element of the solutions to some of the most pressing challenges for today’s cities, such as poor air quality, rising CO2-levels due to emissions from motorised traffic, traffic congestions, and sedentary lifestyles. A high-quality, safe, and widespread network of cycling infrastructure is a key part in the quest to encourage more people to choose the bicycle over the car. There is however a lack of methodologically sound, quantifiable, and consistent methods for evaluating networks of cycling infrastructure, partially due to issues with data availability and quality, inconsistent categorisations, and the fact that cycling as a means of transportation historically has received much less attention and funding, compared to other modes of transport. Meanwhile, there is a rich tradition of using network analysis within e.g. geography and transportation studies to describe, analyse, and model spatial networks. The purpose of this thesis is to examine how a spatial network analysis can be applied to describe and evaluate a network of cycling infrastructure, specifically the network in Copenhagen, Denmark. The analysis is centred on a range of traditional network metrics used for evaluating connectivity and accessibility, primarily based on the computation of shortest paths in a weighted graph, and applying mostly open source technologies such as PostgreSQL/PostGIS, Python, and NetworkX. From this analysis, it is evident that, despite its status as one of the world’s best cities for cycling, there are substantial variations in connectivity and access within the Copenhagen network of cycling infrastructure. The analysis moreover shows that using different weighted graphs to model the network, with weights based on both geographical distance and other factors influencing the cycling experience, can give a fuller and more nuanced picture of how accessibility and connectivity vary throughout the city. Combined with data about e.g. demographics and the socio-economic status in the city’s neighbourhoods, a network analysis of cycling infrastructure can thus highlight issues with areas with a disproportionately high or low network connectivity, identify overloaded network segments, and suggest where it might be beneficial to extend and strengthen the network.

英语. Cycling has been identified as a central element of the solutions to some of the most pressing challenges for today’s cities, such as poor air quality, CO2- emissions from motorised traffic, traffic congestions, and sedentary lifestyles. For cycling to become a viable alternative to motorized transportation, a high-quality, safe, and widespread network of cycling infrastructure is necessary. When constructing and expanding said networks of cycling infrastructure, a method for analysing, evaluating, and comparing them is needed, in order to define minimum standards for network quality, to quantify, measure and track progress, and to detect areas in the network in particular need of improvements. The study is focused around the network of cycle paths and routes in Copenhagen, Denmark. The city is known for its well-developed network of cycling infrastructure, which nevertheless still has room for improvement. The goal of the study is firstly, to examine the network in Copenhagen, with a focus on how well-connected the network is and how access to and from other places by bike varies throughout the city. Secondly, the purpose is to explore how a network analysis of spatial data can be applied as a method within transportation and cycling planning. The analysis is founded on an abstraction of bike lanes and routes as a network consisting of edges (lanes/routes) and vertices (intersections and endpoints). With the use of the concept of shortest paths between vertices in the network, different metrics for connectivity and centrality (how central a network element is for the overall connectivity) are computed. Interpreted in the context of cycling, these metrics can both describe the network of cycling infrastructure at a general level and highlight areas and network elements with a noteworthy low or high connectivity or centrality. Importantly, the shortest paths used throughout the analysis are not just based on geographical distances but are instead based on the idea of the ‘costs’ from moving through the network. The costs are in this case based on variables which influence the experience of cycling, as for example the traffic speed, number of cars, and presence of separated bike lanes. The analysis shows that the network of cycling infrastructure in Copenhagen overall is well-established and has a level of connectivity comparable to regular street networks. Meanwhile, there are substantial variations in connectivity and accessibility across the city. The access to cycling infrastructure is thus unevenly distributed, in a pattern that to some extent matches the unequal distribution of other types of transportation infrastructure as well as variations in socio-economic characteristics. The analysis moreover demonstrates that a spatial network analysis is a feasible and useful method for analysing cycling networks, and that the method can be adapted to incorporate some of the factors influencing the experience of cycling in a city. A network analysis of cycling infrastructure can thus be used to, for example, identify weak or missing links in the network, model the effects from new bike lanes on overall connectivity, and provide metrics with which the networks in different areas can be compared.