Friday, September 27, 2013

Mapping Baltimore's Frequent Transit Routes

Let's take a break from the discussion on border vacuums to revisit the topic of frequent transit service by examining how this service can be effectively promoted via frequency mapping.

Many of us are accustomed to adhering to relatively rigid schedules when commuting to/from work, and transit has long accommodated such predictable linear commuting. But imagine having to adhere to a rigid travel schedule for every other mundane task – grocery shopping, doctor's appointments, a spur-of-the-moment trip to the library, and so on. Consulting a spartan bus schedule and waiting 45 minutes just to pick up a prescription and a tube of toothpaste would be intolerable for most people, so it's not surprising that driving is often considered the only practical option (at least until the roads inevitably become congested). The good news is that transit can be restructured to accommodate the same spur-of-the-moment freedom and flexibility that the car offers, and Baltimore already has several transit routes that, if leveraged and promoted properly, could become the backbone of a lucid frequent transit system.

Frequent transit in Minneapolis: You can always see the next bus coming.
Simply put, frequent transit service is transit that – as its name implies – runs often enough that you don't have to consult a schedule whenever you want to go somewhere. 15 minutes seems to be the threshold at which people become comfortable using a service without consulting its schedule. At this threshold transfers between routes no longer have to be timed, and there's a compelling argument that transit systems with frequent transfer-dependent service are far more useful than systems with barebones point-to-point service. (For example, many of us are perfectly comfortable transferring from one subway to another when visiting New York, but would never consider using a spartan-but-direct bus line when visiting, say, Atlanta.)

The 15 minute frequency threshold is merely a starting point. The better the frequency – say 10, 5, or even 2 minutes between buses – the more freedom and flexibility you get, and the less frustrating phenomena like bus bunching are. Of course, economic malaise has reduced the ability of many transit systems to introduce (or even maintain) 15-minute-or-better service, but I think planning for it is still a worthwhile goal.

Since the benefits of frequent transit are self-evident and have been thoroughly discussed here and elsewhere (the Human Transit blog does a great job discussing the concept in all its particulars), we don't really need to dwell on the subject and can instead move on to the issue of how to present frequent transit service effectively.

At the Human Transit blog, Jarrett Walker introduces the concept of frequency mapping by asking us to imagine using a road map in which all the roads were drawn to look the same regardless of their being freeways, arterials, or local streets. Most of us would find such a map useless, even if it was geographically accurate, because we rely on hierarchical visual cues to read maps and other images: we want to know first which highway to use to reach the city, then which arterial road to use to reach a particular district in the city, then finally which local street to use to get to a particular house or business. Walker points out that a confusing one-size-fits-all method of depiction is precisely what's used for most transit maps:

“Some of the lines [transit routes on the map] have frequent all-day service. Some of the lines have infrequent all-day service. Some of the lines are peak-only. So a transit map that makes all these lines look equally important is like a road map that doesn't show the difference between a freeway and a gravel road. What do we achieve by making [all the lines on the map] look the same? The most powerful message seems to be: These are all buses and all bus services are alike. A key secondary impression is: This map is complicated because bus service is intrinsically complicated.”

Philadelphia's bus network: Some of these routes run frequently and some of them don't, but you can't tell the difference without perusing stacks of schedules.
Walker argues that transit maps should be hierarchically organized like road maps to be truly legible, suggesting a classification of routes into frequent all-day service (15 minutes or less), infrequent all-day service, and peak-only service:

“These three categories are useful in such completely different ways that I would argue they are at least as fundamental as the three basic categories of urban road – freeway, arterial, and local – that all street maps clearly distinguish. When these three different services are drawn with the same line, as in most of today's system maps, you get chaos. The three categories of usefulness should be represented by three kinds of line: a strong line for the frequent network, a lesser line for the infrequent network, and some kind of dashed line for the peak-only network. The test of such a map should be that anyone can see the frequent network standing out from the rest of the map, and that anyone can see the all-day network – both frequent and infrequent – standing out from specialized peak commuter services.”

WMATA's frequency map for DC: frequent service is denoted by thick red lines, infrequent service by thin blue lines, and peak-only service by thin gray lines.
Walker also counters the common “we don't need transit maps anymore because we have smartphones, online trip planners, or <insert gee-whiz techno-gadgets here>” argument by pointing out that many people (myself included) are spatial navigators:

“Some of us need to understand our trip on a map rather than as a list of steps. Only maps can show us where we are and what our options are at every moment of the journey. If you want to feel free in your city, capable of moving about at will to do things you want to do, that's the kind of understanding you need. That's why people love the regular Manhattan grid. We may not need this understanding to follow a trip planner's directions, but it's essential if we want to use the transit system spontaneously [so we don't even have to pull out our phones and use a trip planner in the first place!].”

I'm reminded of the anecdote in which a fellow Millennial wandered around Brooklyn looking for the Brooklyn Bridge with her eyes glued to the directions on her “smart” phone, unable to find the bridge even when it was right in front of her. I therefore strongly agree with Walker that auto-generated online directions and narrative lists offer rather limited means for acquainting oneself with a transit system, even if these tools offer dynamic service updates (detours, stop closures, etc.). Furthermore, transit organizations need to serve riders who may not have regular access to smartphones or the internet.

Walker concludes by arguing that frequency mapping can even affect urban design: “If those who want good transit had access to [a frequency map], they could make decisions about where to locate that would gradually reorganize the city so that people who value transit were closest to good transit, thus making better use of the system's finite resources. The frequent transit network is potentially useful to anyone deciding where to live, where to shop, or where to start a business, and also to planners and developers making the same location decisions. If you just show people a map of undifferentiated services, they won't be able to identify which services would be most useful to them, and they'll thus be less likely to make good location choices.”

Several transit organizations have begun offering frequency maps of their systems, using various tactics to distinguish between the three kinds of service (frequent all-day, infrequent all-day, and peak-only). I particularly like WMATA's Washington DC map, First Group's Leeds map (which manages to assign each route its own color and clearly highlight frequent services!), and Spokane Transit's Spokane map.

As far as I know, Baltimore's MTA hasn't yet created such a map. The regional map is quite comprehensive – it shows MARC, Metro Subway, Light Rail, Local Bus, Quick Bus, Express Bus, Commuter Bus, and Charm City Circulator routes along with connections to outside services (it even includes water taxi landings!), and a viewer can effortlessly follow any color-coded bus route across the Baltimore region.

The MTA's map of metropolitan Baltimore is incredibly comprehensive and easy to follow, but it doesn't really differentiate service frequency.
Unfortunately the map offers only a rudimentary classification of service frequency: Bus routes are sorted into limited and regular service by using solid lines or hollow (white) lines. But “regular service” can refer to anything from a bus coming every ten minutes to a bus coming only once an hour!

I decided to create a frequency map of Baltimore's transit system to see what the system would look like if it was organized into Walker's three service frequency categories. To keep the project manageable (I squeezed it into spare time over several months), I decided to map only a portion of central Baltimore, where most of the city's important destinations lie. Hopefully the map can be expanded to cover the entire region in the future.

I grouped Baltimore's transit services into Walker's three frequency categories by using thick colored lines, thin colored lines, and thin gray lines.
Inspired by WMATA's Washington DC map, I decided to employ a similar tactic to clarify Baltimore's transit services: I used heavy colored lines to denote frequent all-day service,* thin colored lines to denote infrequent all-day service, and thin gray lines to denote peak-only service and other limited-trip service. I decided to minimize the use of color by grouping transit services by type: I used black for all rail services, red for all circulator/shuttle services, light blue for all local buses, purple for all limited-stop quick buses, and dark blue for all water taxis. I used light gray for all peak-only and limited-trip services (regardless of type) so their complicated routing would recede into the background, yet remain available for anyone requiring information on specialized commuting. Finally, I used yellow to highlight those corridors where two or more infrequent all-day routes overlapped to offer frequent service.

*Defining the threshold for frequent service was tricky because it varies by time of day. For example, while the Charm City Circulators and the Metro Subway offer 15-minute-or-better service seven days a week during their (varying!) service hours, most “frequent” local bus routes offer consistent 15-minute-or-better service only on weekdays from early morning to early evening, and anywhere from 15 to 30 minute service on evenings and weekends. Then there are miscellaneous services like the water taxis, which actually offer daily 15 minute service late into the evening for a sizable portion of the year, but then cut back to 45 minutes during the winter months. I therefore decided to use a reasonably loose definition of frequent service, akin to what Spokane Transit did, by including all routes that offer 15-minute-or-better service on weekdays from early morning to early evening. To clarify the differing evening/weekend service frequency on these routes, I added direct links to each route's schedule in the map legend.

Of course, service frequency will vary over time as finances and demand dictate: Some of the currently-frequent routes on this map might slip below the 15 minute threshold, and some currently-infrequent routes might rise above it. For example, not too long ago the 22 offered a 15 minute frequency midday, but now only offers a 20 minute frequency. Someday it may once again offer better midday, evening, and weekend service, at which point this map will need to be updated. Updating the original CAD map only takes seconds, but updating the pdf map with its designated lineweights, offsets, and route badges can take longer. Since this isn't a full-time project, I can't guarantee consistent updates, but I will do my best to incorporate major changes in frequency and/or routing whenever possible.

I'd like to thank the many people who've already studied this issue and published their findings online, making my job so much easier: Stu Sirota and the TND Planning Group studied Baltimore's 20-minute neighborhoods and proposed strategies for tying these neighborhoods together with a primary transit network. Josh Fruhlinger assembled a Google frequency map for the Baltimore region several years ago. The folks at Welcome to Baltimore, Hon! have long maintained a thorough guide to the city's maddeningly-disparate transit services. Elliott Plack managed to map all of the MTA's transit stops (I included them in the CAD map but left them out of the final pdf map due their cluttering effect). Finally, the MTA offers its collected transit data to the public, and it publishes a series of nifty neighborhood maps that allow visitors to see important transit stops within walking distance of major institutions like the Convention Center, Charles Center, and State Center.

I hope this frequency map is useful! Next time we'll return to the discussion on border vacuums by examining those around parks and "green spaces," so stay tuned!

- Marc Szarkowski


  1. Hi Marc

    I enjoyed your article and your work on the MTA map is very nice. You know what they say: "imitation is the sincerest form of flattery". I say that because we designed the WMATA maps, the Spokane maps and the MTA map now in use. We adopted the frequency approach after we created the current MTA map. Perhaps we'll go back and redesign at some point in the near future. In the meantime, what is the best way to reach you?

    Rick Wood
    President and CEO
    CHK America

  2. Absolutely perfect! A lot more riders - if not all, including me - needless to say, only care for the frequency and expectancy of a bus arriving rather than a placed time. Really, I want the MTA to dump timetables and include frequency (timetable will remain to place at one time the frequency standard begins) on schedules. This mind-math of scheduling that a bus may arrive at location midway through the route 2-3 minutes early because parking is cleared BDOT patrol... anyway, that inclusion of something like that ever occurring will factor in the return trip in the opposite direction and create in discombobulation intervals.


Related Posts Plugin for WordPress, Blogger...