information request: peaking patterns

The book is coming along.  Expect a number of information requests in the next weeks, as I start pinning down details.

I need some nice charts showing ridership per hour of the day for a typical inner-city transit line, and for an outer-suburban one.  Very simple bar or line graphs, covering a weekday, that I can use to illustrate the fact that outer-suburban service tends to be much more peaked and inner-city service busier all across the day.  If you're inside a transit agency and have access to such things, I'd appreciate it. 

6 Responses to information request: peaking patterns

  1. Joseph January 23, 2011 at 9:34 pm #

    I don’t have any data for you. But I can tell you that this is absolutely true!

  2. Alon Levy January 23, 2011 at 10:23 pm #

    I think there’s a chart for the Moscow Metro in the link I sent you about it a few months ago.
    In addition, if you look for New York’s Hub Bound Report, you’ll also see fairly detailed numbers for ridership entering the Manhattan core by hour. In addition, if you go here you’ll be able to see LIRR ridership per station broken down based on peak vs. off-peak. Peak presumably means trains defined as peak in the schedule – I forget what hours they enter Manhattan, but you can check.

  3. voony January 24, 2011 at 1:13 am #

    That is certainly interesting,
    But one should be made aware of too fast conclusion.
    Inner city system ridership could see smoother ridership across the day, because it works already at capacity in off peak hour (so has little room to accept extra capacity)… that could be the case of numerous metroplis like Paris, or even Vancouver for some route like 99B (or skytrain in its downtown part).
    Speaking of Vancouver, at you will find hourly traffic on its bridge.
    Not surprisingly on the most congested bridges, like Port Mann bridge, you will not find “peak hour” traffic: bridge is at capacity most for the day…
    Offer drives the demand.
    On other crossing, not at peak, you will see some peak traffic…and on other one using reversal line, like George Massey tunnel, you will clearly identify spike in traffic…but that becomes more Offer driven by demand
    the former case could be more in line with inner city ridership, while the later is more typical of the suburban service…
    hopefully, this point (of close to capacity or not of service) will be integrated in your study.
    PS: for Vancouver, some people had put the revenue bus service per hour for the whole translink, and that provide a graph not dissimilar to the the average hourly traffic observed on a typical (like 2nd narrow) Vancouver bridge. (you can see the graph at )

  4. In Brisbane January 24, 2011 at 1:25 am #

    I would be asking Melbourne’s department of Transport and also TransPerth for data on this.
    It would be very interesting to compare the peakiness of demand on their high frequency, long scope of hours Orbital SmartBus routes and also TransPerth’s 98/99 CircleRoute which runs through these outer areas.
    It would also be interesting to compare the peakiness of demand in cities which ran trains frequently all day (Perth) with those that focus on peak hour only (Brisbane).

  5. Robert Wightman January 24, 2011 at 2:23 pm #

    The TTC posts service summaries, schedules of all routes on line. If you look at this and their route map you can get an idea about the level of services run in the a.m. mid day and p.m. rush hours as well as weekend and evening service. the most recent one that I can find is at
    They are not very consistent about making all of these available on line, or at least easy to find.
    Steve Munro on his site has done a detailed analysis of a number of street car lines showing delays at different points on the line.
    Take a look at the 510 Spadina line which has an a.m. rush hour headway of 2:30 and a Sunday headway of 2:13.
    I hope this is of some help.

  6. jack horner January 24, 2011 at 7:10 pm #

    @ In Brisbane ‘It would also be interesting to compare the peakiness of demand in cities which ran trains frequently all day (Perth) with those that focus on peak hour only (Brisbane).’
    Remembering that other matters also affect whether the system is favourable to peak period travel or all day travel, and the effects may be hard to disentangle. For example (with my guesses on the likely effects):
    – design of the network: if more focussed on central city, will be more peaky.
    – legibility: ‘bowel of spaghetti’ networks, and networks with poor information services and poor transfers, will be more peaky as they are hostile to the less peaky occasional users and cross suburban trips.
    – fare structure: greater discount for distance will encourage peaky commuting. Offpeak discount will encourage all day ridership.
    – design of vehicles: for example double deck rail cars, compared with single deck cars, are better for longer distance commuters who give high value to getting a seat, but their unattractive environment (stairs, low ceilings) is worse for short distance and off peak riders.**
    ** For example, Sydney, Australia: an entirely double deck operation with 400 tonne, 160 metre trains serves the inner city underground, preWW2 inner ring suburbs with 1km station spacings and 50km lines to the urban fringe with 2-3km station spacings.
    The result is an unhappy hybrid of metro and commuter style operation. Peak hour seating is maximised at the expense of high operating costs and low load factors at most other times. A more metro style operation would be cheaper to operate and would have higher total capacity, though with a smaller proportion of seats. Being more suited to shorter trips, I bet it would be less peaky.
    NB Sydney Cityrail has station entries/exits broken down by periods of the day, but I don’t think they are on the internet. Comparing the time of day brackets you could see the relative peakiness of different stations/lines.