Transportation Geography and Network Science/Legibility

= Legibility (Network Perception) =

Legibility is a visual quality associated with readability, clarity and ease of recognition. While the formal definition of "legibility" is used in typography to describe how easy a reader may recognise individual characters in a text, it is also a desirable trait in urban design. In the Image of the City by Kevin Lynch, legibility is defined as "the ease with which its [the cityscape] parts can be recognized and can be organized into a coherent pattern". For transport networks, legibility is an important design aspect as it determines how the network is perceived by its users. Legibility influences the arrangement of a network on a map presentation or even the physical layout of some parts during the design phase.

Legibility, Imageability and Mental Maps
In the book the Image of the City, the cityscape is divided into five elements: paths, edges, districts, nodes and landmarks. They are essential components for people to construct their "mental maps" (network perception) of the city. Paths represent accessible transport network links; edges are the boundaries which separate different zones of activities; districts are areas/regions with different characteristics, usually related to the land use; nodes are intersections or interchanges which are part of the transport system; landmarks are well-known places often used for navigation or as reference points. In an urban planning context, legibility is reflected by how easily a person can remember these elements and navigate through the city based on the mental map. For a transport network, the paths and nodes are particularly important because space is allocated for transport purposes, but the other elements are also very critical in some contexts.

The visual quality of legibility is often associated with imageability (or apparency) as the city becomes more "vivid" and memorable when the network is clear and easy to understand. An example is how the street names are numbered so that their spatial relationship becomes easy to remember. The road layout needs to be legible for the street numbers to be reasonable. A relevant public transport application is that bus stops often use the same name as the streets so that people can easily relate the bus routes to the road system. However, imageability can contradict with legibility because too many memorable elements would lead to increase network complexity, thus reducing the system legibility.

Legibility in Cartography
Cartography is the study and practice of making maps. It is particularly relevant in the discussion of legibility since maps are one of the most common network visualisation tools. Legibility is part of the generalisation concern which is a fundamental problem in cartography. A balance between accuracy and simplicity is required to deliver the information efficiently. Different types of maps with examples are discussed here to elaborate on how they influence the network legibility.

Topographic vs. Topological Maps
A map can be categorised either as a topographic or topological map based on their presentation style. In brief, a topographic map retains most of the geographical information and usually uses contour lines to present the elevation of terrains. The geographical distance can be calculated by using the map scale and the network layout is similar to reality. On the other hand, a topological map omits most of the geographical information such as distance, layout, and orientation. The map elements are better aligned and arranged such that the map usually has a higher level of legibility. Most public transport network schematics today are drawn in a topological style, following the convention established by the London Underground which shifted from topographic map representations to topological maps in its early years.

The London Underground (also known as the Tube) is the earliest underground passenger rail system, which is opened in 1863. The official network map was initially a topographic map on which the underground rail infrastructure was superimposed on the street map. The map format resulted in poor legibility as the network expanded over the years. In 1926, a map designer Fred Stingemore came up with a new network map that removed most of the surface features, reduced some curvature and used regular inter-station distances. In 1931, engineer Harry Beck used his expertise in the electrical circuit to realign the railway routes as orthogonal or diagonal lines. Beck's modification has been widely adopted by many public transport agencies and topological maps have become the standard of modern (public transport) network map representation.

Topological maps are generally more legible than topographic maps, especially when a large network with many routes and overlaps are displayed. It also allows for greater flexibility as to how the map designers choose to express the network layout and convey essential trip-planning information. Besides using straight lines to represent routes/links/edges, an expert on the London underground map, Maxwell Roberts, also experimented circular distribution of networks including the London Underground, Melbourne and Sydney.

2D vs. 3D Maps
With the advance of 3D modelling software and geographic information system (GIS), there is a trend to convert traditional maps to their 3D equivalent such as the perspective/street-level view in many navigation applications. With the proper use of 3D technology, the amount of information and its accuracy can be significantly improved while maintaining the same (or even higher) level of legibility.

A major difference between 2D and 3D maps is the presentation of landmarks. While traditional maps might use different markers to place an emphasis on the landmarks, the place is difficult to identify information about its appearance or surroundings without any images. For instance, the instruction to turn at a certain intersection or cross the road in front of a certain building would not be helpful if the intersection or building is hard to locate in the first place. Although the graphical representation might be legible, the physical network may still cause some confusion. In contrast, 3D maps are better equipped to give a direct view of the places and landmarks so that the spatial relationship between different elements becomes clear. In the Image of the City, Kevin Lynch found that as people become more familiar with the city, they would rely more on the city's landmark for navigation rather than paths or districts. This is related to the concept of imageability and the time to construct mental maps. This means 3D maps may accelerate the development of mental maps by introducing landmarks more frequently than traditional 2D maps.

Inherently, 3D maps are able to show an extra dimension (i.e. elevation). This is an important feature for modern multi-story infrastructures where users may need to navigate through different levels to access the transport service. It also becomes more convenient to integrate multi-modal transport maps since different transport modes may operate at different elevations (e.g. subway, road system, and aeroplane). 3D maps are also arguably more efficient at showing orientations. While a paper map usually references to the North pole, a 3D map may refer to surrounding buildings or other objects for the direction so that bearing becomes unnecessary.

Network vs. Route Maps


A legibility tool is used to break down the network to show only some parts of it on the map in order to improve the legibility. Singapore's Mass Rapid Transit (MRT) system can be a case study for the discussion on route maps/diagrams. The MRT trains are equipped with SMRT Active Route Map Information System(STARiS) which is a series of LED light indicators of routing information (SMRT is the service operator). While the network has 5 lines in operation, the STARiS on each train only displays station information on the current routes. The lights show the following stations and possible destinations of the current train. The route maps can reduce a complex network to a few lines of interest and improve legibility by only keeping the most essential information.

However, there is a risk of over-simplifying and providing insufficient information for decision-making. STARiS 2.0, a newer version of the discussed system, is criticised for its lack of aesthetics and poor system design. One of the problems is that the new digital screen only displays the next five stops instead of the whole route as the original STARiS did. This causes frustration and impatience because users need to pay attention to the displayed information at least once per five stops if he/she is not familiar with the route or interchange information. Another problem is the inconsistent use of topographic maps on STARiS 2.0 instead of the topological representation used everywhere else.

Octolinearility vs. Curvature
A study on the subjective and objective map usability for Paris Metro challenges the design principle of octolinearity which is widely adopted by graphic designers and network mapmakers. The research group conducted several experiments involving 120 participants to evaluate the map usability values for both the official map of Paris Metro and an unofficial version with curve routes. Different sets of survey questions are designed to reflect the (perceived) objective measures such as planning time, journey duration and invalid routes and subjective ratings and overall expectation. The results show that the all-curve map has the best performance on planning time and has fewer invalid routes than the commercial Metro map. The argument is that octolinearity leads to poorer legibility for large networks when the change of direction becomes frequent. The Tokyo Metro Map is another example where straight line segments result in an excessive amount of bends and cause confusion to new network users when compared to the new Paris Metro network map.

Legibility and Connectivity
The transport network is heavily dependent on the structure of paths and nodes. As a result, the network legibility has an impact on the organisation of these elements and directly influences the network connectivity. Not only is legibility a tool to show the level of (perceived) connectivity, it may even be an argument against it. While adding more paths to a network can increase its connectivity, it also increases its complexity and reduces legibility. Therefore, the network planner should be aware of the legibility requirement when the main design aim is to provide connectivity and accessibility for the transport network.

Legibility and Hierarchy
Since there is a hierarchy of roads in the transport system, it is logical to reflect such hierarchy on maps to improve network legibility. The RMS classified map is published by Roads & Maritime Services (RMS) and uses different colour codes to represent Auslink, regional roads and state roads in the state of New South Wales. Another example is the Minneapolis high-frequency network map published by MetroTransit. This map highlights several public transport routes (Metro and bus) with a maximum arrival interval of 15 minutes. It can be argued that these routes are more important for the service operation. The network becomes more legible by eliminating information about routes with longer wait time and simplifying the decision-making process.

Legibility and Wayfinding
One of the key benefits of good network legibility is the ease of navigation and wayfinding. This means network legibility, as a network characteristic or quality of its presentation style, affects travel behaviours significantly. The rate of learning is a direct outcome of the accuracy of network perception. Network users spend time and efforts to learn about the transport network system with the assistance of various navigation tools, e.g. maps and GPS technologies. The effectiveness of such tools can be quantitatively assessed by studying the user's decision-making process and results. To some extent, legibility should be considered if a comparison between hypothetical decisions and observed decisions because it might help explain different learning rates associated with certain locations in the network, e.g. some tunnels or highway ramps may be underutilised due to their poor legibility on a 2D map; there might be a larger group of tourists attracted to better-advertised places of interest with a path marked out.

Legibility in Public Transport Operation
Jarrett Walker, a public transport consultant, makes a connection between legibility and freedom in his book Human Transit. The legibility here refers to the clarity of representation of public transport routes and freedom refers to network users' ability to change travel plans. Walker claims that freedom can be derived as an outcome of legibility because efficient utilisation of the network and the trip navigation are based on a good understanding of the system layout. He conceptualises legibility as two characteristics:
 * 1) "simplicity in the design of the network, so that it's easy to explain and remember"; and
 * 2) "clarity of the presentation in all the various media."

In the context of public transport, the significance of legibility varies for different user groups. For commuters who mostly travel along a fixed route between certain origin and destination pairs (e.g. home-workplace or home-school), legibility is a less critical factor in trip planning and decision-making. However, to promote public transport and encourage people to travel to many different locations would require a recognisable and memorable network presentation. A legible network aims to minimise its layout complexity and enhance the users' perception of service routes. Legibility tools and techniques are often implemented to achieve an acceptable level of legibility when complexity is unavoidable given the physical layout of the network.

Zhan Guo, Nigel H.M. Wilson (2011) conducted a study on the cost of transfer for the London Underground system. In this study, legibility is a factor of the Tube user experience (classified under transfer environment and system design) and influences the perceived transfer cost which traditionally only concerns wait time. Although the study lacks a quantitative correlation coefficient for the transfer environment, it has proven its relevance in the decision-making process for transfers.

= Further Reading =
 * Reddit MapVsGeo has a series of animations illustrating the comparison between topological subway/metro maps and their actual geographical forms in several different cities.
 * Project Subway NYC is a website that publishes 3D modelling maps of subway stations in New York City.

= Discussion Questions =
 * 1) What are some of the missing information in a topological map? How does each of them affect the user's decision-making? What are the likely consequences and how would the network be impacted?
 * 2) Why are most public transport maps in the form of topological schematics but significantly less so for road networks? Consider the interaction with the surrounding environment and access points.
 * 3) When would a 3D map be preferred over the traditional 2D network maps? Does it mainly depend on the network size on the map or any other network characteristics?

= References =