Neil Freeman recently posted a great collection of rail rapid transit maps, all drawn to scale, and all at the same scale. The image at right, of course, is New York City,
He calls them subway maps, but of course that term suggests that the service is all underground, which few “subway” systems are. What matters is that they’re rapid transit. In this case, they’re specifically rail rapid transit, which is why Staten Island’s rail line in the lower left appears disconnected from the rest. In reality, it’s just connected by rapid transit of a different mode: the Staten Island Ferry.
(By “rapid transit” this blog always means transit services that run frequently all day in an exclusive right of way with widely spaced stations — linking centers to each other, for example, rather than providing coverage to every point on the line as local-stop services do.)
system extends over distances vaster than New York City, but with an
extremely sparse network.
The opposite extreme is the Paris Métro, where a thick network of interconnected subway lines really does cover the dense part of the city, to the point that most of the population can easily walk to a station.
Neil’s Paris map is a little misleading, because it doesn’t show the outer-suburban RER network — much of which does achieve rapid-transit levels of frequency. Yet it’s helpful to look at the Paris Métro in isolation, compared to these other maps, just to see how small Paris proper is, and how densely it’s covered with service.
It’s worth noticing that no rapid transit system with anything like Paris Métro’s function– complete coverage of the dense core city of a region — has been built in North America since World War II. What has been built since then are networks designed primarily to connect city and suburb. Why? These systems could be built only as a matter of regionwide consensus, and in most American urban regions, the core city is in the minority. So these new lines have rushed to get out of the core city and reach across many suburbs, resulting in networks like that of the San Francisco Bay Area above: enormous but thin, and not really covering the core city where transit demand is highest. Atlanta, for example:
As a result, the internal needs of dense cities — the places where transit needs per capita are highest — are often not met very well in late 20th century American rapid transit planning, except to the extent that they can be met on the way to a suburban market. Here’s Washington DC:
Every line in this system has a long suburban tail, but they’ve been overlaid at the center in a way that creates a bit of an inner-city network.
Of course, it helps that among the American rail transit systems created entirely in the late 20th century, Washington’s is the most extensive in terms of total miles of lines. Could this have something to do with the fact that Federal funding is administered by people who live there? No, couldn’t be …
Jarrett- you might find interest in recent artwork of my good friend, Matthew Picton. Matthew has explored many aspects of geography and urban form through his art, but his most recent set of City Sculptures capture scale features of urban form, often focusing on street and rail networks. These three dimensional pieces are made by tracing various map features, so do represent a scale model of a limited subset of urban features. Most represent a city at point in time, but in others he’s captured temportal changes in urban form. They may not be as practical in professional application, but they are a lot of fun to view.
This is great, Jarrett. Feeds right into my mapophilia. 😉
One of the buzzwords (or terms) of 21th rail transit projects is TOD (transit oriented development), the idea that residential and commercial construction will center around rail stations. Back over 100 years ago, Henry Huntington and his business associates built the Pacific Electric system here in Southern California, not because there were big bucks to be made in hauling people around the area, but because fast, clean electric transit was a prime selling point for the real estate that they subdivided. Likewise, there’s more money to be made in increasing the density and redeveloping areas previously devoted to industry than in running trains to already-dense neighborhoods.
OOPS! that first sentence should read “21st Century”
Terrific post, Jarrett.
I wonder what the maps would look like if ridership volumes per segment were mapped using a user volume-correlated color scheme or line width to passenger volume and to watch as changing volumes are simulated in 1-hour increments over a 7-day period. Does anyone know if the turnstile to turnstile ridership data is collected and made public by any of these cities? Great blog Jarrett!
Kevin. You can do this with GIS technology, and most rapid transit systems (unlike bus systems) will have turnstile counts. It would look a lot like a map of traffic volumes on a road network. Jarrett
“These systems could be built only as a matter of regionwide consensus, and in most American urban regions, the core city is in the minority.”
This iss often compounded by governmental or intergovernmental agency structures that proportionally over-represent low density areas (the way the US Senate does). BART, for instance has only three out of nine directors representing the Oak/Berk/SF urban core – leaving them with a minority of votes on the BART Board despite having a majority (or strong plurality) of residents and BART riders.