Paris vs Tokyo: the two different models for express commuter rail stopping patterns

A commuter at Tokyo's Shinjuku station. Image: Getty.

Many cities have large commuter rail networks, which function as urban rapid transit and extend into the suburbs. They use mainline rail rather than separate subway tracks, but are identical in other respects to conventional metro systems: urban stop spacing, frequency, and fares are all within the range of metro systems.

The biggest systems are in Japan, where Tokyo and Osaka carry the vast majority of their public transport passengers on commuter trains and not metro trains. In Europe, the biggest system is the Paris RER, while in the German-speaking world all major cities have S-Bahn networks.

But while the concepts are all broadly similar – mainline trains serving the suburbs share tracks in the core, so as to provide metro-like frequency – the stopping patterns vary.

First, some regional rail systems run express trains, whereas others are all local. The Munich and Berlin S-Bahns only have local trains. In contrast, the Paris RER and the Tokyo commuter rail network combine local and express trains – sometimes on four tracks, and sometimes on two, using the schedule to avoid conflicts.

Usually, systems that run express trains are bigger than systems that do not, but there are exceptions: Copenhagen's S-Train has express trains on most branches, and the Zurich S-Bahn has express trains on some lines as well.

On systems that are not modernised, express trains are especially common. That’s because the traditional function of commuter rail is to connect the suburbs with city center at rush hour; local services, connecting the suburbs with each other and not just with the city, are less important.

As a result, American commuter lines, even ones with very little ridership by European or Asian standards, generally have express trains: each stopping pattern might only get 1 or 2 trains per hour at rush hour. So do a few European branches, for example some of the outer commuter lines in Paris, not connected to the RER. Since these lines carry few riders, the important distinction is between different local and express patterns on busy lines that run frequently all day.

There isn't much to say about local trains, which (mostly) stop at every station. Express trains are more complex, and there are two ways to run them: one common in Paris, the other common in Tokyo. The Parisian model is to have long central segments with only two tracks, on which every train makes every stop. (In London, Crossrail is planned to follow the same pattern.) Tokyo’s railways have four-track segments, and express trains skip some stops even in the core.

Tokyo-style express trains may skip fewer stops in the centre than in dormitory communities, but they still skip even some central areas. On the eight-track main between Tokyo and Shinagawa Stations on the Tokaido Line, for example, the local Yamanote and Keihin-Tohoku Lines make all four intermediate stops; but the express Tokaido Main Line and Yokosuka Lines only make one intermediate stop, at Shimbashi. Central Tokyo stretches roughly between Tokyo Station and Shimbashi, and there is one station between them, Yurakucho, with transfers to four Tokyo subway lines. But in the judgement of rail planners, it made sense to skip this station, and for express trains to serve just one in every so many stops on the inner part of the line.

In Paris, no such thing service exists: the central tunnels only have two tracks, so it is hard to arrange local and express trains on them. Even on the few segments of the central network that have four tracks, such as part of RER C, there is no stop skipping. The transport authorities judge it best to have every commuter train make every stop within the city proper, which extends about 5 km out of the center.

Conversely, in the suburbs, Paris does mix some local and express trains on two tracks: the RER B runs 12 trains per hour off-peak – just enough room for trains which run non-stop between Gare du Nord and Charles-de-Gaulle Airport, and some express trains in the southern suburbs.

The Parisian approach ensures that the RER can function as high-frequency trunk lines within the city proper. The RER A averages a stop per 2.5 km on the central trunk, and the RER B and C a stop every 1.2-1.3 km (the other two RER lines, the D and E, only make three city stops each). The Metro averages a stop every 500 meters, of course – but nonetheless, 1.2-1.3 km is well within the range for international metro systems, comparable to the spacing of stations on the London Underground. The central Crossrail trunk will average a stop every 1.6 km – wider than the Underground but not much more so.

In Tokyo, of course, the commuter rail frequency in the core is even higher, since the inner lines are all at least four-tracked. But farther out, there are express and local trains mixed on two tracks, with timed overtakes, using the legendary punctuality of Japanese railways to schedule trains to avoid conflicts. The result is that the express routes have quite wide stop spacing, which permits higher speeds, approaching 60 km/h on the Tokaido and Tohoku Main Lines.


A smaller city, with trunk lines not as full to capacity as in Tokyo, Paris, or London, could mix local and express trains even at rush hour. In Tokyo, local and express trains are mixed on some lines on the shoulders of rush hour (but not at rush hour, when trains arrive every 2 minutes); it is unclear what the absolute upper limit of this system would be, but it appears to be in the range of 15-20 trains per hour. In cities without Japanese punctuality, the limit is about 12 trains per hour: a local train and an express train each coming every 10 minutes, with an overtake every 6-8 stations.

Such cities have a choice. The Paris approach works very well for Paris, and the Tokyo approach works very well for Tokyo. There is always a tradeoff in mass transit between narrow stop spacing for service to more places, and wide stop spacing for higher average speed.

The two different approaches for commuter rail express stopping patterns display a related tradeoff, between higher frequency to all stations and higher average speed at express stations. Which of the two approaches is better depends on local factors. These include city size and density (more sprawl encourages the faster Tokyo approach, more density encourages the more frequent Paris approach); punctuality (better punctuality makes mixing local and express trains on two tracks easier); and how important it is that suburban commuters be able to reach every urban station, rather than just a few major stations.

There is no inherent better choice. The tradeoff is not that one option is more beneficial but more expensive, but rather that the two options have different benefit levels, depending on local conditions.

Those conditions can vary widely between cities, even in the same country. A smaller French or British city might find that its home and job distribution makes the Tokyo approach better, and a smaller Japanese city might find that the Paris approach works better for it.

Cities anywhere might even find that the German approach of not having any express trains works best. This means that planners should consider all stopping patterns, and not just default to what is familiar from nearby cities.

Alon Levy blogs at Pedestrian Observations and tweets as @alon_levy.

Want more of this stuff? Follow CityMetric on Twitter or Facebook.

 
 
 
 

Which pairs of capital cities are the closest together?

Vienna, which is quite close to Bratislava, but not quite close enough. Image: Thomas Ledl

It doesn't take long to get from Paris to Brussels. An hour and a half on a comfortable Thalys train will get you there. 

Which raises an intriguing question, if you like that sort of thing: wich capital cities of neighbouring countries are the closest together? And which are the furthest away? 

There are some that one might think would be quite close, which are actually much further part. 

Buenos Aires, Argentina's capital, sits on one side of the estuary of the Río de la Plata, while Montevideo, Uruguay's capital lies on the other side. 

Click to expand: Image: Google Maps

But at 207km apart, they're not really that close at all. 

Similarly, Singapore – capital of, er, Singapore – always sticks in the mind as 'that bit on the end of the Malaysian sticky-out bit'. But it's actually pretty far away from Kuala Lumpur, Malaysia's capital. A whole 319km away, in fact:

Click to expand: Image: Google Maps

Thinking of 'countries that cause problems by being close together', you inevitably think of South Korea and North Korea. 

Click to expand: Image: Google Maps

And while Pyongyang in the North and Seoul in the South are pretty close together, 181km just isn't going to cut it. 

Time to do some Seoul-searching to find the real answer here.

(Sorry.)

(Okay, not that sorry.)

Another place where countries being close together tends to cause problems is the Middle East. Damascus, the capital of Syria, really isn't that far from Beirut, in Lebanon. Just 76km:

Click to expand: Image: Google Maps

Seeing as Lebanon is currently host to millions of refugees fleeing the horrors of Syria's never-ending civil war and the atrocities of Daesh, or Isis, this is presumably something that authorities in Beirut have given a certain amount of thought to.

Most of the time, finding nearby capitals is a game of searching out which bits of the world have lots of small countries, and then rooting around. So you'd think Central America would be ripe for close-together capital fun. 

And yet the best option is Guatemala and El Salvador – where the imaginatively named Guatemala City is a whole 179km away from the also imaginatively named San Salvador.  

Click to expand: Image: Google Maps

Another obvious place with lots of small-ish countries is Europe – the site of the pair of capitals that drove me to write this nonsense in the first place. 

Click to expand: Image: Google Maps

And in fairness, Vienna and Bratislava do make a pretty good showing of it. Austria's capital sits on the Danube; drift downstream, and you swiftly get to Slovakia's capital. As the crow flies, it's 56km – though as the man swims, it's a little longer. 

There are more surprising entries – particularly if you're willing to bend the rules a little bit. Bahrain and Qatar aren't really adjacent in the traditional sense, as they have no land border, but let's just go with it. 

Click to expand: Image: Google Maps

Manama, Bahrain's capital, is 140km away from Doha, the centre of the world's thriving local connecting-flight-industry which moonlights as Qatar's capital. 

Sticking with the maritime theme, Port of Spain in Trinidad and Tobago is 152km from St George's, Grenada. 

Click to expand: Image: Google Maps

Good, but not good enough. 

Castries, the capital of the Carribbean country of St Lucia, is 102km north of Kingstown, the capital of St Vincent and the Grenadines. 

Click to expand: Image: Google Maps

Better, but still not good enough. 

Basseterre, the capital of St Kitts and Nevis, inches ahead at 100km away from St John's, the capital of Antigua and Barbuda.

Click to expand: Image: Google Maps

But, enough teasing: it's time to get down to the big beasts.

If you ask Google Maps to tell you the distance between the capital of Congo and the Democratic Republic of the Congo, it comes up with a rather suspect 20km. 

 

Click to expand: Image: Google Maps

A short distance, but considering the only thing separating the two is the River Congo, something's up: Google places the centre of Brazzaville a little north of where it should be, and the centre of Kinshasa many many miles south of where it should be, in some sort of suburb.


So, in true CityMetric style, we turn to train stations. 

Though such transport hubs may not always perfectly mark the centre of a city – just ask London Oxford Airport or London Paddington – in this case it seems about right. 

Kinshasa's main train station is helpfully called 'Gare Centrale', and is almost slap-bang in the middle of the area Google marks as 'Centre Ville'. On the other side of the river, 'Gare de Brazzaville' is in the middle of lots of densely-packed buildings, and is right next to a Basilica, which is always a good sign. 

 

Click to expand: Image: Google Maps

And when marking that distance, you get a more realistic 4.8km. If you want to be really keen, the ferry between them travels 3.99km, and the closest point I could find between actual buildings was 1.74km, though admittedly that's in a more suburban area. 

Pretty close, though. 

But! I can hear the inevitable cries clamouring for an end to this. So, time to give the people what they want. 

Click to expand: Image: Google Maps

If you ask Google Maps to tell you how far away the Holy See, capital of the Vatican, is from Rome, capital of Rome, it says 3.5km. 

Click to expand: Image: Google Maps

If you set the centre of Rome to be the Palatine Hill, the ancient marking point for roads leading out of Rome, that narrows to 2.6km.

 

Click to expand: Image: Google Maps

Fiddle a bit and put the centre of the Vatican as, well, the middle bit of the roughly-circular Vatican, that opens up a smidge to 2.75km.

Click to expand: Image: Google Maps

Mark the centre of point of the Vatican as the approximate location of St Peter's Tomb within St Peter's Basilica, which is after all the main reason the Vatican is a thing and not just a quirky suburb of Rome, and 2.67km is your answer. 

Though obviously in practice Rome and the Vatican are as far away as one single step over the railings at the entrance of St Peter's Square, which fairly blatantly makes them the closest capital cities in the world. 

But that would have been a very boring thing to come out and say at the start. 

Oh, and if you hadn't worked it out already, the longest distance between a capital city and the capital of a country it shares a land border with is 6,395km. 

Click to expand: Image: Google Maps

I know it's tough for you, Vladimir and Kim. Long-distance relationships are a real struggle sometimes.

I can't make a pun work on either Moscow or Pyongyang here, but readers' submissions more than welcome. 

Jack May is a regular contributor to CityMetric and tweets as @JackO_May.

Want more of this stuff? Follow CityMetric on Twitter or Facebook.