Greater Greater Washington

Visualize Metro ridership

A recent Data Visualization Hack Day not only drew coders from around the region, but also some WMATA planning staff. They used a Javascript library to build a calendar of Metrorail ridership, showing each day's ridership since 2004 as a square of a different color.


Image by WMATA. Click to enlarge.

If you go to the large and interactive version on PlanItMetro, you can mouse over individual squares to see the date as a tooltip.

The darkest red days have the lowest ridership, the darkest green the highest. You can see high ridership events like President Obama's January 2005 2009 inauguration, the Stewart/Colbert rally in October 2010, Snowmageddon/Snowpocalypse in February 2010, and more.

Stepping back, it's clear how ridership is highest in April, June, and July, and the number of very high ridership days jumped significantly in 2008 but then has stayed flat or a bit down since. Weekend ridership has gotten lower in recent years, probably because of all the trackwork.

What do you notice?

David Alpert is the Founder and Editor-in-Chief of Greater Greater Washington and Greater Greater Education. He worked as a Product Manager for Google for six years and has lived in the Boston, San Francisco, and New York metro areas in addition to Washington, DC. He loves the area which is, in many ways, greater than those others, and wants to see it become even greater. 

Comments

Add a comment »

What do you notice?

An enormous potential for growth on the red and light green days.

This graph is a reminder that outside rush-hour, there is enormous untapped capacity in metro.

by Jasper on Jun 5, 2013 10:25 am • linkreport

President Obama's January 2005 inauguration

IF ONLY!

This is a great visualization. The July 2008 boom (when gas prices hit their peak) is clearly evident as is the yearly drop-off in August during traditional summer recess time. The April bump every year is cherry blossoms, I assume?

Monday holidays appear to have about as much ridership as your average Saturday, which makes sense.

by MLD on Jun 5, 2013 10:32 am • linkreport

October 29th, 2012 shows a ridership of ZERO. That can't be right... or did I miss something that day?

by Dirk on Jun 5, 2013 10:43 am • linkreport

Dirk: That was Hurricane Sandy. Metro was closed: http://www.wmata.com/about_metro/news/PressReleaseDetail.cfm?ReleaseID=5362

by David Alpert on Jun 5, 2013 10:44 am • linkreport

Haha of course. I was out of town that week. Totally missed it. Thanks David!

by Dirk on Jun 5, 2013 10:46 am • linkreport

@MLD, probably a combination of Cherry Blossoms and baseball season starting

by Birdie on Jun 5, 2013 10:47 am • linkreport

This is an excellent visualization of the reason metro was built: get about 750,000 people to work Monday - Friday.

Saturday and Sunday = it's a nice thing to have around.

by Nick on Jun 5, 2013 10:50 am • linkreport

Nick: I don't know if that's fair. People move around less on weekends, period. I agree Metro isn't as useful on weekends as it should be, but just because the ridership is a lot lower doesn't really prove anything.

by David Alpert on Jun 5, 2013 10:55 am • linkreport

While Mondays have lower ridership than the rest of the work week (ignoring holidays), it appears that since last August and more pronounced in 2013 to date, there has been a trend of most Mondays falling below 700K. Less people going in to work on Mondays?

Thanksgiving and Christmas day riderships are really low. That is when Metro should do some single tracking work!

by AlanF on Jun 5, 2013 11:03 am • linkreport

David I don't think it's at ALL true that people move around less on weekends! Do you have data on that?

Most folks I know pretty much go to/from work during the week (with maybe a detour for happy hour, ferrying kids, etc.). During the weekends it's errands, social engagements, and other things. This results in many more discrete trips for most of us on a given Saturday or Sunday (sample: regular grocery store, costco, home depot, brunch with friends, walk in the park, dinner/drinks).

I would LOVE to take Metro/Metrobus for these trips, but it's just not reliable enough. I more often find myself driving or maybe cycling.

by Kristen on Jun 5, 2013 11:04 am • linkreport

I used to take the Metro every weekend but trackwork killed it. I almost only take the bus now because why wait 20 minutes for a train?

by Alan B. on Jun 5, 2013 11:06 am • linkreport

kristen

leaving aside the weekend days that some of us spend gardening, doing household chores, home centered recreation, etc - the kinds of trips you mention are intrinsically harder for transit to serve. each leg of a multidestination trip is short, and yet transit, even IF it had weekday frequencies, has a wait time (and access times) independent of trip length. Plus those locations typically have lower parking costs than workplaces do, and often lower parking costs than the same places do on weekdays. If one has a car, and is inclined to use it, its difficult to make the economics for transit work - at least for a large part of the area that metro serves. And for the areas in the core where driving is least convenient even on weekends, biking and walking are sometimes strong alternatives.

by AWalkerInTheCity on Jun 5, 2013 11:11 am • linkreport

RE: Weekend auto travel
There are fewer vehicle trips on weekends (it's about 77% of weekdays). Those trips do have higher occupancy so person miles on the weekends is about the same.

There are many reasons why transit travel is lower on the weekends (it has basically always been this way.) The biggest reason is that the kinds of trips people take on the weekends are not the kinds of trips that are best suited for or served by transit for a wide variety of reasons.

It is not mostly about service - even systems that have excellent service at non-peak hours (e.g. SkyTrain or NYC subway lines) have much lower ridership on the weekends.

by MLD on Jun 5, 2013 11:30 am • linkreport

I don't own a car and I've notice how using the system on weekends has increased my frustration level. It has gone hi enough that I am willing to pay uber for a pretty-much-reliable ride to my home. A lot of people I know tend to agree, we assume that metro does not exist on the weekends. Unless, using it during the day, going more that 2 metro stations (if they live in a walkable area), and not transferring lines. Come on, how is it possible that the combined waiting time is longer than the actual trip (e.g. 20 min waiting time at origin + 20 min waiting time transfer). It is faster to bike from George Town to Bethesda....

by WM on Jun 5, 2013 11:31 am • linkreport

It's a really cool graphic. Although I hate to be a nitpicker here, I find their color scheme to be completely counter-intuititve. Perhaps it's because I spend so much time staring at Doppler radar images, where the usual coloration has green for the lightest storms and red for the most intense. Same for the air quality index. Or the standard temperature coloration where blue is the cold, small temperature limit and red is the hot, high temperature limit.

Perhaps I just need to see intense red as high anti-transitness in this graph.

by thm on Jun 5, 2013 11:33 am • linkreport

red is bad, green is good

by Alan B. on Jun 5, 2013 11:35 am • linkreport

Until the heavy weekend track work started, weekend ridership was increasing at a much higher rate than weekdays. From 1995 to 2010, weekday morning ridership increased 43% and Saturday night ridership increased 142%.

by Ben Ross on Jun 5, 2013 11:48 am • linkreport

@David Alpert

Ehh, I think it does.

Unless I'm dead wrong (don't think I am) the bulk of the metro ridership is suburbs---> core. (We talked about core-core and suburb suburb in recent posts, but those were not big chunks of ridership). These people stay in their neighborhoods on the weekends, and don't ride the train if going into the city b/c driving is cheaper and faster.

by Nick on Jun 5, 2013 12:16 pm • linkreport

It's indisputable that Metro was originally built to move people to work Monday - Friday, with weekend service being an afterthought.

Of course, people like me, who don't own a car and rely on metro as our primary transit option, curse this short-sighted decision every Saturday night, but it seems there is little we can do given current constraints and the general ineptitude of WMATA.

by Hadur on Jun 5, 2013 12:16 pm • linkreport

Dumb question: Why is it sideways?

by Chris from Arlington on Jun 5, 2013 12:29 pm • linkreport

It's indisputable that Metro was originally built to move people to work Monday - Friday, with weekend service being an afterthought.

I don't understand what this statement means. The infrastructure is there. How can you build it so that it's only usable 5 days a week and not the other days?

by MLD on Jun 5, 2013 12:57 pm • linkreport

Is there any way to access the raw data at this level? I'd like to make some other visualizations of it that would make a little clearer how the level and seasonality have changed over time.

by Gray on Jun 5, 2013 1:05 pm • linkreport

AWalkerInTheCity + MTD

1) I'm not talking about a multi-destination trip here. In DC, it's possible to take the Metro to a neighborhood for brunch, go grocery shopping nearby, and Metro home. Later in the evening, one might go out for dinner and drinks. My suspicion is that people in DC WANT to take Metro in these circumstances on weekends, but don't - because the service is beyond terrible. In many cases (given 20-30 min headways and malfunctions), taking Metro could add 30-45 min over driving. No sane person would choose Metro in that situation.

2) I'm not sure we know what the "biggest reason" is that people don't much ride Metro on the weekends. Your arguments to that effect are conclusory - not that you are necessarily incorrect; we just don't know. Having lived in other cities with extensive public transit - and having visited many others - I can tell you that I used those systems on the weekends for all purposes. The reason I don't do the same with Metro has to do solely with reliability. Your mileage may vary.

by Kristen on Jun 5, 2013 1:15 pm • linkreport

@Chris from Arlington - it's so all the days are aligned. Sundays at the top, Saturdays at the bottom.

by 7r3y3r on Jun 5, 2013 1:23 pm • linkreport

...and Americans read left to right so you can see how the ridership increases as the year progresses in an easier fashion than if it were arranged top to bottom like a standard calendar

by 7r3y3r on Jun 5, 2013 1:26 pm • linkreport

These people stay in their neighborhoods on the weekends, and don't ride the train if going into the city b/c driving is cheaper and faster.I don't stay in my neighborhood on the weekend. But it's definitely not faster and cheaper to drive after spending 30 minutes circling around to find a spot, only to give up and pay a private lot something outrageous to park.

by dcmike on Jun 5, 2013 2:12 pm • linkreport

I just realized you were talking about suburbanites, but my point remains. I live in the city and I'd rather wait for a bus/train or pay Uber than drive even a short distance.

by dcmike on Jun 5, 2013 2:16 pm • linkreport

kristen

I specifically said "its difficult to make the economics for transit work - at least for a large part of the area that metro serves" Id suggest that includes almost all the suburban areas except possible the RB corridor in Arlington, and several parts of DC as well - almost everywhere EOTR and also parts at least of upper NW.

by AWalkerInTheCity on Jun 5, 2013 2:26 pm • linkreport

and while the core is different, again I would note that some of the worst service problems are outside the core. For example the EFC-WFC work related to the silver line.

suppose you had just closed the orange line for a week instead. Would that have really meant that many more people using metro to/from WFC, Dunn Loring, and Vienna on weekends? I dont think so. The preponderance of riders there that have autos, do NOT take metro, unless they are going into DC on day (cherry blossom peak, inauguration, etc) when parking looks to be impossible. Closing down for 7 days OTOH would have been a catastrophe for weekday riders, and for the entire northern virginia region.

by AWalkerInTheCity on Jun 5, 2013 2:36 pm • linkreport

@MLD, the NYC subway system does have lower ridership on weekends, but the percentage dropoff is less, especially on Sundays. The MTA stats (http://www.mta.info/nyct/facts/ridership/) for 2012 shows average ridership of 5.38 million weekdays, 3.17 million on Saturdays, 2.49 million on Sundays. The NYC subway is seen and operated as a complete transit service.

As more TOD projects open around the DC Metro stations, we should be seeing increasing ridership in the evenings and weekends. The current level of weekend service disruptions which are cutting into weekend use are due to a 6 year capital program. One would hope that the WMATA planners and board will be looking to follow the big catch up on maintenance program with making the system more responsive to local 7 day transit use. The Silver Line, Purple Line, DC and Northern VA streetcars are going to expand the use of the Metro as a 7 day a week transit system.

Charts of daily ridership numbers presented in the same format (but adjusted scales) for the NYC Subway and Boston T over the same period would be interesting. I expect there would be less seasonal and monthly variation for the NYC subway.

by AlanF on Jun 5, 2013 2:45 pm • linkreport

by the way, in this context its interesting to look at ridership fall off on mondays and fridays, which I think I see in the above visualization - days when there are more leave days, RDOs, and teleworking. That again suggests the dominance of commuting, and the challenges facing metro in responding to more concentrated peak riding. Non commute use may well become more important to balance out that peaking problem.

by AWalkerInTheCity on Jun 5, 2013 2:56 pm • linkreport

Weekend ridership actually took less of a hit than I thought.

I scraped the data from the visualization, and threw this chart together:

by andrew on Jun 5, 2013 3:12 pm • linkreport

Of course NYC has better weekend numbers than other systems. I expect you would see the same thing in any other world city. And I expect that if we had access to daily station entries/exits for WMATA, something some transit agencies already do, you would see a huge difference in the weekday/weekend drop-off between suburban and inner stations, and a difference depending on bus route.

And I would expect in DC that as we put more TOD projects online we will see higher weekend ridership.

But there seems to be an underlying and unsaid criticism in some of these posts, that saying it's somehow all WMATA's fault; they provide crappy weekend service ("untapped capacity") and so people won't ride. But there are many reasons that people do not ride on the weekends that have nothing to do with frequency of service:
- types of places people go on weekends are not well-served by transit/network design
- weekend trips involve more people so cost comparison between transit/driving is lower
- no monthly passes so every transit trip has marginal cost

The reality is that for trips into the core people on the edges are probably still using Metro. But those trips aren't the majority of trips - the majority of trips are people going to visit friends, or grocery shopping, etc. Those are not well-served like trips to the core.

by MLD on Jun 5, 2013 3:20 pm • linkreport

I scraped the data off of the site, and threw together a few charts.

The most striking thing I noted was that weekend ridership actually declined (a lot) less than I thought.

(Note that 1 = Sunday; 7 = Saturday in the legend)

by andrew on Jun 5, 2013 3:21 pm • linkreport

@andrew: Is it possible to provide a link to that? I can't see the entire graph on this page.

by Gray on Jun 5, 2013 3:22 pm • linkreport

If you want to play around with the data yourself, here it is in a Google Spreadsheet.

(And, apologies that GGW seems to cut off my images)

by andrew on Jun 5, 2013 3:27 pm • linkreport

Proof that they need to lower prices on the weekend.

by David C on Jun 5, 2013 3:37 pm • linkreport

@andrew: Thanks! Interesting that median ridership is consistently lower on Mondays than other weekdays. I would have expected Friday to look like that, since it was my understanding that many Feds work from home then. But maybe they're actually doing so on Monday, then leaving early Friday?

Or maybe it's due to the holidays. But I wouldn't expect that to affect the medians...

by Gray on Jun 5, 2013 3:48 pm • linkreport

Monday is consistently the lowest weekday because of holidays. 10% of Mondays are holidays.

by MLD on Jun 5, 2013 4:04 pm • linkreport

@MLD I'm looking at the median rather than the average, so I'm a bit surprised if a handful of holidays are causing such a huge drop.

That being said, Monday's ridership levels are indeed the least consistent.

A few weird trends I'm seeing here:
* Eyeballing this, Sundays ridership levels appear to be the most consistent, followed by Saturday and Wednesday.
* Some time around 2006, our Tues/Wed/Fri/Sat commuting patterns started to become a lot less predictable, but Thursdays became much more predictable for some reason.

Don't read too much into the 2013 data points. We've only got half a year's worth of data.

by andrew on Jun 5, 2013 4:33 pm • linkreport

Any ideas on what the weekend ridership is for DC only, VA to MD, MD to DC and DC to VA trips and trips for Metrobus with the same locations.

by kk on Jun 5, 2013 10:23 pm • linkreport

Thank you for sharing this great visualization!
We were inspired to re-create it using Excel, here you can find our model:

http://goo.gl/5PU45

The concept is simple, using conditional formatting and gradient color.
From the link, you can download the file.

Ps: I really like your anti-spammer map!

by Kris on Jun 6, 2013 12:36 pm • linkreport

Add a Comment

Name: (will be displayed on the comments page)

Email: (must be your real address, but will be kept private)

URL: (optional, will be displayed)

Your comment:

By submitting a comment, you agree to abide by our comment policy.
Notify me of followup comments via email. (You can also subscribe without commenting.)
Save my name and email address on this computer so I don't have to enter it next time, and so I don't have to answer the anti-spam map challenge question in the future.

or