Stephen Rees's blog

Thoughts about the relationships between transport and the urban area it serves

What the Census Dotmap tells us about Metro Vancouver

with 6 comments

Screen Shot 2013-01-06 at 4.00.42 PM

I copied this map from Brandon Martin Anderson‘s website. “This is a map of every person counted by the  2011 Canadian census” in Metro Vancouver. He states “I wanted an image of human settlement patterns unmediated by proxies like city boundaries, arterial roads, state lines, &c.” Actually I found it pretty easy to pick out the border between Canada and the US (his map covers both) and the very obvious proxy that gets left is the census tract – since the placement of the dot that represent each person is necessarily random within that space.

It is a truism that transit is all about density, and this map for me tells us a lot about our region. I am sure someone else with the right electronic chops can merge in a transit map – preferably one that has a Human Transit feel to it – where thickness of the line represents service frequency. For now I think I will just state the glaringly obvious. The neglect of Surrey by our transit system cannot be justified by claiming that city’s density is inadequate to support frequent transit service. Secondly while you cannot see Highway #1 on the map south of the Fraser it is pretty obvious where it is through Burnaby.

It is also worthwhile to take a look at the zoom function, pick out your neighbourhood and look at its density. I was quite surprised by the (relatively) high density of Richmond at No 4 Road and Steveston Highway north east corner. (The Shell Road corridor and McNair high school both stand out clearly) And certainly stand in stark contrast to the low density of Shaughnessy and Arbutus Village in Vancouver. Note too how Main Street divides a low density West Side from a high density East Side. Considering the city does not have wards, this is quite a remarkable political achievement. Notice also how the midtown Broadway corridor and Kits are much darker than the surrounding areas.

I only saw this map for the first time this afternoon – but I think it does indeed spark some ideas worth discussing

Written by Stephen Rees

January 6, 2013 at 4:35 pm

6 Responses

Subscribe to comments with RSS.

  1. Unfortunately this map does not work for me, even zoomed I can’t tell the difference between the dense and moderate density. Voony posted a map that is way clearer to me and matches my perception of the region better (though I don’t know the year of the data). He also added an employment density map, a very important component of transit use.


    January 7, 2013 at 8:47 am

  2. It is indeed an interesting map with some humorous anomalies like the rather large living population in Mountain View Cemetery and a higher population density in a section of Pacific Spirit Park than is reported for some nearby neighbourhoods.


    January 7, 2013 at 4:56 pm

  3. This is a rather nonsensical graphic excercise. Political boundaries and arterials indeed strongly mediate population and job density and therein mediate transport access. What purpose does it serve to remove streets where people actally live, work, shop, get educated and travel on and substitute population locus with randomness?

    This map on its own does not justify transit south of the river over Broadway and other truly more dense corridors lined with neighourhood after neighbourhood of high density development that have existed for decades longer than Surrey has only recently produced in a tiny pocket.

    Moreover, the Trans Canada is clearly visible (as is the Fraser hwy) through Surrey and Langley, but disappears into the edge of the geographical boundaries of Burnaby Lake and Still Creek watershed.


    January 7, 2013 at 5:37 pm

  4. Rico, I think the data on my map are of 2006, the point was to show it was one component but not the only one.

    “humorously” but it is not an anomaly, the DT CBD Vancouver is empty, and still it is where all the transit lines go 😉 … That eventually makes also the point that employment density map is very important (probably more important that residential one : you could be willing to bike from home to the railway station, much less on the other end, for a host of reason…one is to let your bike out at the station overnight…)

    What shows the Bradon Martin Anderson’s map (at correct resolution) is that you can quickly identify the Richmond downtown as well as the North Van one or ambleside in WestVan, you can’t say the same for Surrey (may be Guilford emerges a bit)…in otherword it is a LA type density type rather a New York time one (That is the ” density fallacy” argument used by Paul Mees) : Surrey could be relatively high density on average but, the average density is the Surrey density pretty much everywhere

    * Cluster of high density in overall low density average is eventually better fro transit that higher than equally spreaded density (that is the point of another post ( )

    *In the Zurich post, I mention the importance of clustered job (strong CBD lead to good transit usage,…even in Calgary thought this later argument has been recently challenged by Jeffrey Brown in an Journal of Public Transportation article…)

    *A last point, the map is not showing, is that the high density need to be layout in a fashion favourable to Transit service. here is what is done in the Clayton area:,+Surrey,+BC&hl=en&ll=49.122998,-122.70366&spn=0.007422,0.01207&sll=54.112352,-126.555646&sspn=27.4365,49.438477&oq=clayto&t=h&hnear=Clayton,+Surrey,+Greater+Vancouver+Regional+District,+British+Columbia&z=17

    Have you ever wondered what means “car oriented high density”?


    January 7, 2013 at 10:17 pm

  5. Hi. I’m Brandon. I made the map.

    It’s important to understand that the census dotmap is _not_ an effective illustration of population density. The dots have to be large enough to see even while zoomed out. In dense areas, the dots overlap – essentially saturating an area black above a particular density. At high zoom levels areas with very different population densities are equally saturated. You can see differences if you zoom in, but in that case you can only see a single neighborhood; not a comparison of two. A log-scale choropleth math would be more appropriate for illustrating population density, especially when it varies by several orders of magnitude.

    Moreover, the census dotmap isn’t meant to answer any questions or support any hypotheses. It’s meant to _raise_ questions, which more precise quantitative methods can answer.

    Brandon Martin-Anderson

    January 12, 2013 at 12:37 pm

  6. Thank you for that very helpful clarification. In this case the question has been asked for many years. The answers have been generally unsatisfactory.

    Stephen Rees

    January 12, 2013 at 2:33 pm

Leave a Reply

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

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

Twitter picture

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

Facebook photo

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

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: