Commuter Rail Speed (Hoisted from Comments)
For commuter rail, even more so than urban transit, there is a tradeoff involving speed, cost, and coverage. Higher speed is useful all else being equal, but all else is frequently not equal. American commuter rail is on average faster than European and East Asian commuter rail, but fails because relative to the distances people travel and the amount of sprawl it must compete with it is quite slow. Because of the need for higher speed, my previous commuter rail modernization proposals have featured average speeds higher than those that are normally found elsewhere.
That said, speed should be increased by means of better rolling stock, adequate maintenance, and better timetable adherence leading to less schedule padding. The American practice of running low-frequency trains from each suburban station expressing to the main city station makes service worse rather than better. Consider the following practices:
In Paris, RER trains are more or less local. There are some trains skipping stops, but in RATP territory (see schedules here), trains will just skip a few stops, rather than running truly limited-stop. In SNCF territory there are some express trains. In addition, the Transilien trains run on legacy routes even in the city, and run express outside it; often they’ll run nonstop between Paris and the terminus of the parallel RER line.
In Berlin, the S-Bahn lines run local. As in Paris, there are regional lines that are separate from the metro-like S-Bahn, and those make fewer stops in the urban core.
In Tokyo, there are local trains and rapid trains. Local trains make all stops. Rapid trains come in several flavors – they only stop at select stations, but sometimes there are several levels of rapid trains (all on the same tracks). The busiest lines have a track pair dedicated to local trains, and a track pair dedicated to express trains, and then there are usually two consistent express patterns, one more express than the other. The local trains (Yamanote, Keihin-Tohoku, Chuo-Sobu) run on dedicated tracks and are very metro-y in their lack of branching, and the rapid trains run on separate lines and look more legacy, just without the brand separation of France and Germany. In addition, there’s less separation of infrastructure.
For example, the Chuo-Sobu Line has local trains running to Mitaka, and the Chuo Rapid Line has trains making limited stops to Mitaka and local stops farther west as well as trains making limited stops all the way, with timed overtakes at the express stations; the express trains’ stopping pattern is consistent. Consult this schedule for details (click on “interval timetable”).
The speed of all of those lines is quite low – for example, the RER A averages 47 km/h between Boissy-Saint-Léger and Saint-Germain-en-Laye. However, this comes from short station spacing, coming from the fact that they serve dense urban areas, with reasonably dense suburbs. The need for speed in Paris is much less than in Boston or even New York. In Paris, 20-30 kilometers out of the city one already leaves the built-up area. Even in Tokyo, a much larger city, Takao, 53 kilometers from Tokyo Station, is near the end of the urban area. In contrast, Providence is 70 kilometers from Boston, and Ronkonkoma is 80 kilometers from Penn Station.
Pedestrian Observations 
Yes, Providence, Worcester, Lowell, Haverhill and Fitchburg are far away, but most Boston commuters are inside the I-495 belt, which is ~50km out. Of course, there’s plenty of empty space inside that, but mostly outside the Rt 128 belt (~20km).
One of my go-to “stunners” at parties etc. is that Boston is the least dense major urban area (defined as contiguous census tracts of minimum density, possibly with connecting tracts of non-residential land) in the US… and that LA is the densest.
http://en.wikipedia.org/wiki/List_of_United_States_urban_areas#2010_Urban_Areas
LA: 7,000/sq mi
San Francisco: 6,300
San Jose: 5,800
Delano, CA: 5,500
New York: 5,300
…
Boston: 2,200
US urban areas by land area:
New York: 3,500 sq. mi
Atlanta: 2,600
Chicago: 2,400
Philadelphia: 2,000
Boston: 1,900
Dallas-Ft Worth: 1,800
LA: 1,700
And then there’s California and our 150 kilometer commuter lines.
Well, development follows commuter rail, and rail follows commuter rail. So commuter rail is a device for sprawl.
Therefore slower commuter rail with more stops actually may make sense, with some hardwall distance from the city centre somewhere around ~40km – everything further out would take too long. So commuter rail can have rapid transit distances (1.5km in the city, 2-4km further out), while still having a total travel time of less than an hour from downtown to the terminus. This may help to foster regional cohesion around the city. And while possibly fostering sprawl to some extend, may also create a distinct metropolitan region, deliberately constrained by the transit network.
In a model like this, anything further out should really exist as an independently functioning city, with regional rail connections that are not expressively targeted at commuters.
Montreal for example is still fairly regionally cohesive, with almost all the population living within a 35km radius from downtown. The only town with commuter rail access that is farther out (St Jerome, 45km) is actually a more independent city (which you can see from the median commute distance, which is really low compared to all other suburbs). Commuter rail could help maintain this regional cohesion, by deliberately keeping average speeds low (~45Km/h), and extending at most 35km out.
Well, the problem is that Boston is not as cohesive. Matthew’s right that most commuters are within 50 km (though most commuter ridership comes from stations outside Route 128 – a little more than 50% on the north side, and nearly all ridership on the south side), but New England’s urban geography is such that there are a lot of small cities arranged haphazardly. For example, Plymouth is older than Boston, but has recently redeveloped as a Boston bedroom community. North and west of Boston, several cities (Waltham, Lowell, Lawrence, Haverhill) are 19th century mill towns that have been drawn into the Boston orbit since their industries went into terminal decline. The area in between those cities and Route 128 has been filled in with very low-density suburbia, too spread out for short station spacing.
Also, not all commuter lines are the same. Some lines have naturally long interstations, for example the Providence Line, or else they have enough traffic to support both local and express trains, with long interstations on the express trains, for example the LIRR Main Line and the NEC in New Jersey. Those lines are also straight enough to support high top speed, 160-200 km/h rather than 100-140. Others are the opposite: some of the MBTA lines feeding into North Station desperately need infill stations in Cambridge and Somerville, and some of the shorter LIRR branches would make fine low-speed local-only RER lines.
New England’s urban geography is such that there are a lot of small cities arranged haphazardly.
They aren’t arranged haphazardly. You have a somewhat different perspective if you walk or paddle to where you are going. Usually they were walking or paddling to where the mill was. It’s a lot easier to put the mill where the falls are instead of moving the falls.
In Paris, 20-30 kilometers out of the city one already leaves the built-up area.
Scale. Lop off the more suburban parts of Nassau, Westchester, Essex and Bergen Counties and you have a metro area that has the same population as metro Paris. Or New York City itself is roughly 75% of metro Paris.
Geography. Metro Paris, London, Berlin, are on wide flat plains. New York has a mile wide harbor on one side and half mile wide estuary on the other, then mountains, not very tall ones but mountains none the less, a few miles away from the coast. Everything south of Coney Island and east of Sandy Hook is ocean…. Raritan Valley Line, Harlem River Line….
Political. Highways were going to free the people from the evil grip of the greedy monopoly railroads in the US. Only works with cheap gas. In the rest of the world rail was a government service just like roads, Gas, along with all the other imported energy, was a highly taxed luxury item. Tokyo made it and makes it even harder to own a car with things like having to prove you have off street parking. ( I can’t find a reference, so did Manhattan until the 60s. Parking regulations island wide were “No Parking midnight to 6, for street cleaning”) … the nickel fare meant people in Corona and Rego Park didn’t have to pay LIRR fares to get into Manhattan…
Alon, is it possible you could put the imperial units in parantheses? It’s hard to grasp the distances and speeds if you’re not entirely familiar with the metric system.
Nikko, not meaning to be a dick, but if you don’t make the effort to understand what a km is, you’re restricting yourself to the APTA “special olympics” of transit, comparing Chicago to Baton Rouge while ignoring Seoul and Stockholm. That aside, 0.6 times the unfamiliar number is more than accurate enough in general. Hell, halve the km to get miles and it’s close enough for here.
(Disclaimer: I’ve lived in the US for nearly three decades and I still have no idea what a stupid degree Fahrenheit is: almost certainly I’ll be mentally converting – 32 * 5 / 9 to my dying day.)
I’m studying engineering so I’m familiar with the metric system. I just can’t intuitively grasp what 70km/h is in my head while I can intuitively grasp 70mph. I can easily convert a measurement using 0.6 but it’s a pain to be keeping track of a page full of numbers that are in metric and having to mentally keep tab on their conversions. And viewing the world in terms of my native measurement system does not limit me to to a “special olympics” mindset. I’m sorry but that is just really arrogant.
Some things I forgot to add: when I say I intuitively understand 70mph, it’s in the same vein as you intuitively understanding what temperature is like in Celsius but not so much in Fahrenheit. And the only thing that makes the Imperial/English system inferior to the metric system is the difficulty of converting between units. But for the sake of understanding speed, either one is just as good as the other.
Ugh. I try to include both numbers when talking about a post or article or study that uses English units, but sometimes this means dropping mph numbers when talking about the US independently of what other people say.
(It’s kind of weird – I’m bilingual, but use metric more than English, except for fuel economy, for which I’m much more familiar with mpg because I almost never see the corresponding numbers outside an American context.)
England is no stranger to long distance rail commuters. Our suburbs are dense by comparison to american ones but London stretches it’s tentacles far and wide. Train services around London are a rich mix of routes on a half hourly basis. Some will be just with London, and stop at all the inner stations, the next service will go to town just outside and will only stop at major interchanges in the inner suburbs before becoming all stop further out. This patterns repeats several times as you go further out with trains running further and further out of London before they make a stop.
The result is minor towns and villages may just have a half hourly service to London, while a few stops along the line a major centre will 6 to 10 trains an hour all day.
Time marks the commuter field around London. With commuter services on the main lines running at 90 to 100 mph on the slow lines with expressed running at 125 mph on the fast you can find seasoon ticker holders scattered far and wide. 100 miles out of London on the fastest lines. Though ticket prices for these long distance routes means the numbers doing so declines as prices rise. The rise of the part time commuter has flung the numbers higher though. People who only need to be in London 3 or 4 days a week have escaped even further out for the lower property prices and better schools while maintaining a small flat in the city.
Toronto copies the English pattern of express services stopping at “outer” stations, then skipping all (or almost all) the inner ones. That always seemed to a better option than having the express stopping at one station in four or five, with rest done by an all-stops. The latter forces a change [transfer] on most users; the former gives everyone a one-seat ride while increasing average speed for the “outer” stations.
Toronto doesn’t have anything like the frequent trains London has, though.
It’s very rare to actually have an express that stops at one station in four or five throughout the line. More commonly, the express runs local in the outer segment, and then stops at one station in four or five farther in, when it’s paralleled by a local. Most express subway trains in New York work like this, as do the rapid trains in Tokyo.
It’s not quite all or nothing on the Uk ones though. Yes, it’s a pattern of local stops further out then express or semi fast but important suburban stations get called at by expressess. For example the Windsor lines out towards Reading from London Waterloo runs 45 miles to Reading (and is considered the slow line to Reading). The Reading service is half hourly (more at peak) and is all stations to Staines, which is on the edge of London, it then only stops at Feltham, Twickenham, Richmond, Clapham Junction and Waterloo. The half hourly Windsor train joins the line at Staines where it continues as all stations to Twickenham and then only stops at Richmond, Putney, Clapham Junction, Vauxhall and Waterloo. An additional service from Weybridge also joins at Staines and then is all Stations to waterloo via the Hounslow Loop, additional inner all station services aslo join the main linr further in from the Hounslow and Kingston Loops. So Staines has 6 trains an hour to London, though only four run at a decent speed. Feltham has 6 an hour, Twickenham and Richmond 8, while Putney has 10 an hour.