Fantasy Maps: Columbus

At the Urbanophile today, a fantasy transit map of Columbus, Ohio by Michael Tyznik.  This fuzzy image will show you the general shape:  For the very detailed map, click here.


It imagines a network of rapid transit subways many of which flow into surface rail lines, so presumably a model similar to San Francisco or Brussels or the Boston Green Line.

(He calls it a “tram-train” system, but it’s actually the opposite of the pattern used by the celebrated tram-trains in Karlsruhe, because here “train” means a subway flowing through the center of the city, while in Karlsruhe the tram portion is the city and surface train lines are used to reach outer suburbs.)

Some problems to note even in a proposed utopia:  Subways that flow into streetcar lines are often a poor fit, for three reasons:

  1. Subways are very high capacity and streetcars usually lower, if only
    because you physically can’t run streetcars as frequently as you can
    run subways.  In both San Francisco and Brussels, the subway service
    branches onto two or three streetcar branches, and that manages the
    problem somewhat.  But alas:
  2. Streetcars in mixed traffic are exposed to many causes of delay, subways are not.  Since a line is only as reliable as its least reliable point, the full capacity potential of the subway can never be used because cars are flowing in from the streetcar portion at unpredictable times.
  3. Subways are very fast and streetcars very slow.  It’s rare for a demand pattern intense enough to require a subway to end so abruptly that the only thing you need are streetcars.  In both San Francisco and Brussels, the feeling is that if a subway has to transition into a surface light rail form, it should ideally go into exclusive right-of-way so that it’s still somewhat fast, and then perhaps transition to streetcar further out as loads are lower.  (The M-line in San Francisco tries to do this.)  Of course, this is not precisely what happens in either San Francisco or Brussels, because like all systems they are prisoners of the long-ago design of their infrastructure.  The city has grown in so tightly around the infrastructure as it is that it would be unimaginably expensive (in political pain as much as money) to change it.

All these issues will be familiar to transit riders in San Francisco and Brussels.

Note also that ending routes in one-way loops, as this fantasist does, is not the best practice unless the loops are very small (e.g. one block wide).  That’s because you need a driver break and recovery point at the end of the line, and you want the vehicle to be empty at that point.  If you’re ending in a large loop (planners sometimes call them “balloon loops”) you have the same problem as the London Circle Line: there’s never a point where everyone is off so the driver can have a rest and the vehicle can get back on schedule if it’s a bit late.  Even driverless systems need some spare time at end of line to recover from delays.

18 Responses to Fantasy Maps: Columbus

  1. Aaron M. Renn December 2, 2009 at 12:20 pm #

    I’m curious to know your take on the Muni metro system.

  2. Jarrett at December 2, 2009 at 12:47 pm #

    Post expanded a bit in response to Aaron’s comment.

  3. JamesL December 2, 2009 at 3:04 pm #

    Does the map creator in fact envision a downtown subway? All I can find on his site is reference to light rail with an exclusive right-of-way.

  4. Jonathan December 2, 2009 at 3:08 pm #

    Jarrett, is there some system

  5. Jonathan December 2, 2009 at 3:11 pm #

    Jarrett, is there some standard for how to name and number transit lines? Should “A” be the most used line, or the one with the most westerly northern terminus, or the longest one? I think they should go in some kind of alphabetical order, west-to-east, but what do you think?

  6. Michael Tyznik December 2, 2009 at 3:20 pm #

    I created the map. I’d just like to say that I am an amateur with an interest in transit and transit design, but these maps are mostly an excuse for me to do information design and to maybe spur conversation about transit. I’m sure there are major problems with my design.
    For clarification about the “subway”, they’re actually existing surface-level tracks, but the street grid around the Union Station stop is actually raised above ground level, which means that main trunk line does function like a subway. This photo shows where those tracks run under the current Convention Center, on the site of the former Union Station:
    Let me know if you have any more questions, although I am probably not qualified to answer them. 🙂

  7. Michael Tyznik December 2, 2009 at 3:23 pm #

    Oh, and as for the numbering/lettering of the system, letters are used for the light rail lines, and numbers for streetcars, arbitrarily. They are lettered/numbered in the order I’d propose building them, starting with the A line because it runs along the planned 3C tracks, and the B line because it is on the Ohio Hub tracks.

  8. Alon Levy December 2, 2009 at 5:11 pm #

    Michael, are those Ohio Hub trains supposed to function like German-style tram-trains, which function as mainline commuter trains outside urban areas and as streetcars within city centers?

  9. rhywun December 2, 2009 at 6:46 pm #

    I can’t speak to Columbus or the usefulness of the proposed network, but I like the design of the map. Clean and crisp, and very Vignelli. I especially like the depiction of the neighborhoods–good job! My only criticism is I never liked Vignelli’s depiction of transfer stations–the white bubble that is commonly used in such situations makes a better symbol I think.

  10. Alon Levy December 2, 2009 at 7:00 pm #

    For fantasy maps, the Vignelli design makes it harder to figure out whether the proposed transit is any good. Vignelli-style maps can make a snaking line look straight and vice versa, and can distort scale, making a suburban system look local.

  11. rhywun December 2, 2009 at 8:00 pm #

    Alon, I’m a big fan of when agencies provide both a schematic and a geographic diagram. As a transit geek, I want to see both. I have strong opinions on which one is more useful to the general public, but I do see a need for both types of diagrams, each in their proper place.

  12. anonymouse December 2, 2009 at 11:14 pm #

    The point of the subway sections on subway-surface lines is as much capacity as speed. Muni Metro or the Green Line wouldn’t be able operate as many trains as they do if the lines ran on the surface. Also, the subway portions tend to be under downtown areas, with narrow, congested streets which mean low traffic speeds and little room for dedicated surface ROW. Once you get further out, you can usually get a boulevard median to run the line in, and speed limits around 35. The bang for buck of a subway there is rather lower.

  13. Pedestrianist December 3, 2009 at 12:27 am #

    Your points about the disadvantages of subway/streetcar systems is absolutely spot on!
    Our legacy rail lines (JKLMN) are surface/subway for the reasons you and anonymous describe – they were built a long time ago and when time came to upgrade the city only mustered the money and will to upgrade the downtown segment under Market Street.
    But maybe because all we have is surface/subway we seem hell-bent on only building that kind of line in the future.
    The new T-Third will become the Central Subway – a surface/subway line that doesn’t benefit from the collector phenomenon you describe.
    And the only subway option studied for rail on Geary was also a surface/subway line.

  14. James Dowden December 3, 2009 at 8:57 am #

    The other thing that concerns me is that the central “subway” has too many branches feeding into it. Assuming 2-minute minimum headways on the core section (which may cause problems with reliability), the average branch cannot get more than 6 trains per hour. At 2½ minutes, that becomes an ugly 4.8tph.
    I would also question the wisdom of running an express service at such frequencies. Assuming 30 seconds saved per stop (which approximates New York’s express runs well), the western express run saves 2 minutes, which is less than the 5 minutes average waiting time for the express (assuming 6tph per branch). The eastern one looks more reasonable, but it sure would annoy the local riders.

  15. Michael Tyznik December 3, 2009 at 10:53 am #

    Good points, James. I do think at a certain point tunneling another trunk line directly under downtown would be necessary.

  16. Alon Levy December 3, 2009 at 7:07 pm #

    Rhywun: I’m not even getting to that debate… all I’m saying is that for fantasy maps, geographic accuracy makes the map clearer. Real maps produced by transit agencies are another issue.

  17. Ted King December 3, 2009 at 7:44 pm #

    More on San Francisco –
    For those who are curious the links below have details on S.F.’s Central Subway and one of the water problems they’ll be dealing with. Personally, I think the project is a ruptured duck due to cost (~ $1B), deep tunneling, and short routing (why not try for Fort Mason, just west of Van Ness Ave.?). Plus the ETA is 2018 (MMXVIII in Roman).
    I think my beloved city is already a poster child for the dangers of trying to do transit on the cheap. The Central Subway and the CAHSR projects bid fair to make it a flaming mural.
    Central Subway Project page etc. :
    Hallidie Plaza redesign w/ cistern :

  18. Nathanael December 22, 2009 at 8:17 pm #

    The problem with the Columbus proposal turns out to be that those train tracks have a lot of freight trains running on them; there’s no bypass line for the freight; and the construction of the Convention Center eliminated the space for extra tracks which used to be there back when it was Union Station.
    Columbus really screwed up with that Convention Center.