Comment of the Week: Serene and Effective Transit

Peter Parker of Melbourne on Transit makes a link between my serene sprawl post and the “Be on the Way” principle.

To my mind the most serene are when homes are dotted along one road between two towns. In other words being on the way!

Even though residential density is low, bus service levels and patronage can be suprisingly high eg Melbourne Route 683 http://www.metlinkmelbourne.com.au/route/view/925

As soon as there are multiple parallel roads between towns, or one can’t draw a ‘line of best’ fit that’s direct between A and B, can be driven by road AND is walkable from homes along it, then the effectiveness of public transport drops greatly.

I wonder if this is one of the reasons for transit working in European mountainous villages where there’s only one road and villages dotted along it?

DSCF2065 I think this is right.  If topgraphy prevents development from spreading away from a single main road, you get a perfect geography for transit — everyone within walking distance of a single line.  This is why transit planners find so much beauty in one-dimensional urban forms, such as linear strips of beachfront towns or strings of villages along mountain roads.

Paul Mees’s recent book talks quite a bit about the amazing performance of rural transit in Switzerland.  I’m now curious if this geography is part of what makes the difference.

Photo: Cultivated mountainous landscape east of Bern, summer 2009.

22 Responses to Comment of the Week: Serene and Effective Transit

  1. Alon Levy October 13, 2010 at 7:13 am #

    Bear in mind that while Switzerland’s rural rail is better than that of any other country, rural Switzerland remains auto-oriented. It’s urban Switzerland that has very high transit mode shares.

  2. Danny October 13, 2010 at 8:18 am #

    I ran into the same phenomenon in Honduras. Developments happen mostly along major roads, even in rural areas.
    I actually lived for a while in a town called Sabanagrande. You can see it here: http://maps.google.com/maps?q=sabana+grande+honduras&um=1&ie=UTF-8&sa=N&hl=en&tab=wl
    It was a TINY town of about 1500 people (smaller than my high school), and yet because it was squarely on the path between Tegucigalpa and Choluteca, I never had to wait more than 5 minutes for an intercity bus. Well, I guess one time I waited 15 minutes, but that was at 4:00AM.
    There are other small towns that are quite a bit larger than Sabanagrande, but were out of the way of major routes. I remember one town with a population of 15k that only had buses come about 5 times per day. What a big difference location makes.

  3. Cap'n Transit October 13, 2010 at 8:45 am #

    It’s especially interesting considering how a lot of urbanists favor grids and multiple routes in cities.

  4. John October 13, 2010 at 8:49 am #

    It’s probably also why Honolulu outperforms all other US cities of its size.

  5. F October 13, 2010 at 9:07 am #

    And also why rural Japan is so well-served by trains and buses. Outside the Kantō and Yamato plains, Japan is a graph at least as much as Switzerland is.

  6. Leo October 13, 2010 at 9:32 am #

    @Cap’n Transit:
    1. A street has capacity limitations. Running more than 60 buses an hour down one street tends to encounter difficulties.
    2. Frequent cross-streets are very important for being able to walk to and from a bus stop.
    3. Cities tend to not be arranged in a straight line even without grids. The opposite of a grid in this discussion would be a giant spiral, not cul-de-sacs. Cul-de-sacs also contradict the single route approach.

  7. anonymouse October 13, 2010 at 10:40 am #

    I think the “grids” urbanists are talking about are at a finer scale, and opposed to the suburban cul-de-sac-hell style of development, where there is an arterial street with a residential area along it, but separated by a wall, and zero or one access points into the residential area from the arterial per half-mile. You can still have this along a single road in a mountain valley, which would mean that the people in the residential area potentially have to walk an extra 10 minutes to get to the bus stop.

  8. EngineerScotty October 13, 2010 at 11:44 am #

    For the better part of a century, contemporary highway engineering has viewed this type of road–the rural arterial passing throuh town centers and surrounded by commerce–as something to be bypassed. The needs of inter-urban travelers are placed above the needs of small-town residents (often on the legit grounds that there are more of them), and limited-access roadways are built (where adjoining property access is prohibited, either by law or by easement restrictions). And more often than not, if there’s bus-based transit, it gets routed on said bypass, rather than using the “old” route, on the grounds that most of its passengers are trying to get from big city A to big city B.

  9. J October 13, 2010 at 11:46 am #

    Unrelated:
    There was a very large ad today in the Fresno Bee about your upcoming appearance. Is it a surprise? I haven’t seen it mentioned here on this website.
    If you’d like, I could send you a picture of the ad.
    Of course, this being Fresno, the irony is that there is a large picture of a bus, but instead of saying which bus route serves the venue, there are instructions on where to park.

  10. Jarrett October 13, 2010 at 12:35 pm #

    @J. I’ve seen the poster, yes. It’s on my list to do a post on it. Yes, I had the same comments, but you start where you are.

  11. Jarrett October 13, 2010 at 12:38 pm #

    @All re grids.
    Tne grid is the perfect network form for for a continuously developed two-dimensional area, as I argued here:
    https://www.humantransit.org/2010/10/sydney-ambitious-new-car-sharing-policy.html
    … and in that case I am talking about a large scale network grid with routes spaced every 800m/0.75 mi or so.
    This doesn’t change the fact that you can get even better outcomes if your service area is “one-dimensional” so that the whole thing can be on one line.

  12. Jeffrey Jakucyk October 13, 2010 at 12:56 pm #

    This doesn’t surprise me too much. Historically, the busiest streetcar line in Cincinnati was the East End car line. That’s a long neighborhood (actually a string of sub-neighborhoods) that stretches about 5 miles east of downtown along the Ohio River, sandwiched between the river and the steep hillsides. This neighborhood is generally no more than two or three blocks wide, making transit access very easy. Unfortunately due to river flooding and general urban flight, it’s quite depopulated now.
    I also think that highly linear development like this makes transit ridership much higher, because there’s actually very little within walking distance from a particular point. Instead of having a large radius of things within walking distance, you only have a choice of walking up or down the main corridor. This means you have to use transit more often to get to what might otherwise be a lot closer in a flat gridded neighborhood.

  13. GD October 13, 2010 at 2:12 pm #

    Though it is implicitly there in the article, let’s just remember that you need a particular kind of alpine valley in which the centers of population are really just strung along, and not nestled on any plateaus or in side valleys. Or, for that matter, valleys along a river that does U-turns and other such nasty things.
    Also, even European transit suffers from service cuts, especially in those very sparsely populated side-valleys, which, ironically, can be found much more often in the lower parts of the Alps. The higher the mountains get, the more sharply defined and thus transit-friendly the population distribution will be.

  14. Paul C, Vancouver October 13, 2010 at 2:24 pm #

    While a linear pattern may a more effective transit experience. Simply because it is easier to run one bus route down one street. Then it is to run multiple bus routes down multiple streets.
    But this idea seriously breaks down once you population starts to rise. Simply because the distance someone has to travel increases with a linear pattern versus a grid pattern.
    If a city were laid in a 25KM/1KM pattern. With one road. Then a person at one end would have to travel at least 25KM just to get to the other end. You might also assume that the average distance travelled is half thus the average person has to travel 25.5KM
    If you take that same city and laid it out in a 5KM/5KM pattern. Now the furtherest anyone has to travel (corner to corner). Would be approx 7KM. Even then the average distance would be half that at about 3.5KM.
    So a linear lay out may work fine in much smaller populated areas. It would never work in more populated areas.

  15. Paul C, Vancouver October 13, 2010 at 2:25 pm #

    EDIT for my above comment
    I meant to say that the average distance in my linear example would be 12.5KM not 25.5KM as I mistakenly typed. 🙂

  16. Brisbane October 13, 2010 at 3:53 pm #

    Isn’t this the whole ‘linear cities’ thing all over again? Didn’t linear cities die out in the 1940s???

  17. J October 13, 2010 at 5:47 pm #

    @Jarrett, I dont know how long you’ll be in town, but if you’d like some recommendations on what to see/eat/do Id be glad to send some. Of course Im going to try to make the talk.

  18. voony October 14, 2010 at 12:22 am #

    yes Topography is very important,
    I have touch on it in
    http://voony.wordpress.com/2010/04/26/thezurichmodel/

  19. K October 14, 2010 at 4:04 pm #

    @Brisbane, I’d describe the Gold Coast as a linear city. It just follows the coast.

  20. Wanderer October 17, 2010 at 11:13 am #

    The Napa Valley in California is an interesting example of this, most of the towns are strung along one highway. So it’s possible, even with pinched American funding, to give them some level of transit service. Whereas in the “wine” area of neighboring Sonoma County (not the suburban area), population is much more dispersed and service not as good.

  21. Jase October 18, 2010 at 3:55 am #

    Recognizing the merits of the linear small town, I once did some analysis comparing a long grid with a square grid. I thought the long grid might win, and that would explain new York. Wrong. The square grid led to lower travel times. In any substantial linear city with central density, pt would be too busy in the center and too empty out in the burbs ( assuming consistent capacity along the route)

  22. francis October 20, 2010 at 2:35 pm #

    The linear system works best when there are two cities, thus providing anchors at each end.