Calculating station to station passenger flows

[deltic]

The rail industry is very secretive about passenger flows on the network on the spurious grounds of commercial confidentiality. In the much more competitive world of aviation, the CAA requires all airlines to provide data on every one of their flows serving the UK and publishes it in aggregate. So you can find out how many people fly each month from Edinburgh to each London airport for example.

Given the information that is in the public domain are there any data wizards on this forum who could produce a rough estimate of station to station flows, ignoring any limitations with regard to that data?

The data we have on an annual basis is:

Total passenger numbers using every station
The passenger timetable so the number of trains serving each station and the destinations served
Total number of passengers using each TOC
The average number of passengers per train for each TOC
Region to region passenger flows
Number of passengers travelling solely within each region
Number of passengers traveling solely with each sub region (NUTS or now ITL2 area)

There are over 6m possible station to station flows on the national rail network but probably 99.9% of those are negligible.

Any one up for a challenge?

 

[yorkie]

That is not enough information by itself to estimate this.

However some people will have additional information which could be used to roughly estimate; indeed some people on this forum will no doubt have access to the actual figures.

 

[deltic]

That is not enough information by itself to estimate this.

However some people will have additional information which could be used to roughly estimate; indeed some people on this forum will no doubt have access to the actual figures.

With each of the known data being a constraining factor it should be possible to come up with a reasonable guestimate but would need considerable computing power to do so - effectively trying to solve a few million simultaneous equations.

 

[Dr Day]

Most of this is readily available to the DfT from analysis of ticket sales data and timetables through the ORCATS algorithms. These date back to BR times so would imagine are not particularly constrained by computing power. Maybe this kind of data will be more openly available, at least for passenger journeys, once GBR is more up and running. Good luck to anyone trying to reverse engineer it though!