On the network topology of spines

This post was originally published on Sam’s blog.

ECan recently released their draft long-term public transport plan for Christchurch and Timaru. On the other side of the globe, Jarrett Walker and co are proposing a new bus network for Dublin. The proposed Dublin network centres around the concept of a ‘spine‘ – a collection of frequent bus routes that use the same corridor in the central city, turning that corridor into a very frequent transit line.

The proposed Christchurch network consists of frequent ‘core’ routes and less frequent ‘connector’ and ‘cross-city’ routes. ECan’s definition of frequent is a bus every 10-15 minutes, with the idea being that you can just turn up to one of these bus stops without needing to consult timetabling material, knowing that you won’t have to wait long.

Here’s a sketch of what they’re roughly proposing, in diagrammatical form:

no spines

Their network has a radial topology, with 9 lines radiating out from the central city and one orbital line. New lines (with my own unofficial labelling) are Magenta, Brown, Green and Red. Under this network, a lot of Christchurch gets better service to the CBD (see Chris Morahan’s model). However, could we combine some of these frequent routes into common rapid transit corridors? Below is an attempt:


Under this altered network, we have two main lines – A (North/South) and B (East/West) – that branch into individual lines after travelling along common corridors (orange is lonely because I couldn’t figure out if it would be an A or a B). This would mean the Papanui to Moorhouse and Riccarton to CBD lines would be incredibly frequent, with a bus every 4 minute-ish. I’ve also shown heavy rail for a potential connection point at Moorhouse. It should be noted that the spines aren’t just collections of bus routes travelling along the same road – they are intelligently timetabled frequent routes, so that there’s a bus every 3ish minutes, 7am-7pm every day.

Is this better? I don’t think so. The extra frequent coverage under ECan’s plan is hard to argue with, and I doubt the benefits of a couple of very rapid corridors outweigh that. Also, the Christchurch City Council is putting a fair amount of effort into promoting the central city as an opportunity for development, and, to me, a rapid-spine based network may undermine that effort by encouraging [transit-oriented] development away from the CBD along Papanui and Riccarton roads. The coverage based network being proposed would likely do more for promoting development in the CBD, as evidenced by Chris’ analysis demonstrating that more people can reach it in less time.

Additionally, wayfinding can help promote some rapid corridors that exist under their plan anyway. For example, plinths like these above would be well suited outside Westfield Riccarton and at a Moorhouse Ave stop, where buses will likely be every ~2-4 mins at peak times under their network. This sort of signage would ensure people view the core network as a cohesive whole and help cement main corridors such as Riccarton – CDB and south Colombo St as rapid transit routes that can be travelled on without pre-planning.

5 thoughts on “On the network topology of spines

  1. ” it seems inappropriate for ECan to undermine that effort by promoting their own ‘transit-oriented development’ along these couple of corridors.”

    The RPTP is not solely Ecan’s plan. The plan is developed by the Joint Public Transport Committee, which is Ecan in partnership with CCC and other territorial councils, so it appears your conclusions are mistaken.


    1. Ah yes – just updated that bit with more detail. My point is that incredibly high frequency transit down Riccarton Rd and Papanui Rd (provided by a spine-based network) might result in lots of transit-oriented development there – normally good but when CCC is trying to get the CBD to take off potentially bad.
      Although that ‘might’ is a massive ‘might’. Would take true rapid separated transit for that to happen I would guess.


  2. The other point is do you really want to wade into the mass of politics surrounding the CBD development… the landowners in the CBD think the CBD is important, the ratepayers outside the CBD less so.


    1. If we want a transport system that works, then it’s essential to wade into the politics surrounding cbd development. The biggest mistake christchurch has made in its development so far is not involvimg transport experts enough in landuse decisions. I’d also argue that if we want a city that is successful we should all value a functional CBD, even if we don’t own land there- after all it is the primary reason the city exists.


      1. Yes, never have so few had so much power.

        Urban planners can write RMA land use rules without Regulatory Impact Statements & Benefit Cost analysis being required on each and every individual rule.

        Broad S42 reports are all that are required.

        Transport in the end is little more than the derived demand of the spatial land use pattern.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s