Urban Areas of Canada

[YannickTO]

Hello!

I've been wanting to discuss this topic for a while and putting it into the StatsCan thread might not be a good idea, it will get deluded with everything else being discussed.

That being said, what are, for you, the real numbers to determine a city population?

For me, it always has been the «urban population data«. Comparing cities with their distinct municipal boundaries or with the CMA boundaries is like comparing apples to oranges. So many cities have different types of boundaries or governments, it's really hard to take these numbers seriously.

However, the urban data numbers are reflecting the continuous buildup of an area, and it makes way more sense when comparing cities.

I was looking at the data provided by StatsCan for the 2011 census, and I was very pleased with the Urban Areas numbers. I then looked at all these areas on Google Earth, and I completely fell in love with the Urban data.

To me, a city, is not defined by political divisions, but defined by continuous buildup area.

It should be a priority for StatsCan to release yearly estimates of all urban areas over 1,000 population. Urban data never get the love that it should, and it's quite a shame.

What are your thoughts? Are you excited with the upcoming 2016 Census data coming soon? As you may know already, the urban data is the one I'm looking forward the most.

Thanks for reading and sorry for my poor English.

PS: Trying to find Ontario cities and locations through the subdivisions list always has been a pain with all the nonsense municipal names and amalgamations.

[Acajack]

I'm not on board for this as it seems a bit of a flawed measure.

For example, Ottawa's includes one of the suburbs within city limits that is separated by a strip of Greenbelt, Orleans, but excludes the other two - Kanata and Barrhaven - also separated from the city by Greenbelt land.

All there are definitely part of Ottawa's "urban area" and even "core urban area", no less than Etobicoke is part of Toronto even though it's separated from the downtown by Humber River parklands.

[YannickTO]

Originally Posted by Acajack

I'm not on board for this as it seems a bit of a flawed measure.

For example, Ottawa's includes one of the suburbs within city limits that is separated by a strip of Greenbelt, Orleans, but excludes the other two - Kanata and Barrhaven - also separated from the city by Greenbelt land.

All there are definitely part of Ottawa's "urban area" and even "core urban area", no less than Etobicoke is part of Toronto even though it's separated from the downtown by Humber River parklands.

I agree and I hope that the new figures from 2016 will rectify these examples. Kanata is a great example that needs to be added to Ottawa. But all in all, when you look at the complete list (down to the 1,000 urban areas), it's the ultimate list to really have an accurate view on our cities and urban landscape.

[1overcosc]

Originally Posted by Acajack

I'm not on board for this as it seems a bit of a flawed measure.

For example, Ottawa's includes one of the suburbs within city limits that is separated by a strip of Greenbelt, Orleans, but excludes the other two - Kanata and Barrhaven - also separated from the city by Greenbelt land.

Barrhaven is part of the urban area as well. Only Kanata is separated out from it.

Kanata will always be that way unless the Greenbelt is removed or narrowed, because the Greenbelt in that part of the city is just wide enough to meet the Statscan definition of separate.

[WhipperSnapper]

If it's not easily measured and compared than perhaps it shouldn't be measured and compared.

[YannickTO]

Originally Posted by WhipperSnapper

If it's not easily measured and compared than perhaps it shouldn't be measured and compared.

Not easy to measure maybe you're right... (although I doubt it's that difficult to do), but for comparing cities and urban development, the ultimate real list is the urban figures list.

The municipalities list (subdisivisions) and CMA list (the metro areas) are the worst lists when it comes to comparing cities between each other. The urban list definitely gives a more accurate look to our urban environment. Don't you think?

You really think we can fairly compare a municipality of 700 square kilometres to a municipality of 20 sq km? CMAs also have many different geographic sizes and rules (Oshawa and Hamilton not included with Toronto is one), and the whole list gets blurred and non logical. On the other hand, the urban area lists provides an excellent look and fair comparison between our urban centres.

[Acajack]

Stats Can are fairly conservative in the way they count things. I'm quite comfortable with CMAs personally. CMAs may be imperfect too but to me they are less imperfect than urban areas.

[YannickTO]

Originally Posted by Acajack

Stats Can are fairly conservative in the way they count things. I'm quite comfortable with CMAs personally. CMAs may be imperfect too but to me they are less imperfect than urban areas.

I would love a diagram where the main category is the CMA/MA, the 2nd category is the urban area and the 3rd category is the subdivision.

That way, we could have a definite list, with all the layers underneath.

Example (those are obviously not the right numbers, just as an example);

1- Toronto CMA - 6,000,000 - 2000 sq km
a)- Toronto Urban - 5,000,000 - 1200 sq km
b.1)- Toronto city - 2,800,000 - 400 sq km
b.2)- Mississauga city - 775,000 - 300 sq km

2- Montréal CMA - 4,000,000 - 1800 sq km
a)- Montréal Urban - 3,400,000 - 1100 sq km
b.1)- Montréal city - 1,800,000 - 400 sq km
b.2)- Laval city - 450,000 - 250 sq km

Something along those lines. CMAs, urban areas and cities, all into 1 giant statistic and breakdowns for all of us to enjoy and analyze.

[GreaterMontréal]

Originally Posted by YannickTO

I would love a diagram where the main category is the CMA/MA, the 2nd category is the urban area and the 3rd category is the subdivision.

That way, we could have a definite list, with all the layers underneath.

Example (those are obviously not the right numbers, just as an example);

1- Toronto CMA - 6,000,000 - 2000 sq km
a)- Toronto Urban - 5,000,000 - 1200 sq km
b.1)- Toronto city - 2,800,000 - 400 sq km
b.2)- Mississauga city - 775,000 - 300 sq km

2- Montréal CMA - 4,000,000 - 1800 sq km
a)- Montréal Urban - 3,400,000 - 1100 sq km
b.1)- Montréal city - 1,800,000 - 400 sq km
b.2)- Laval city - 450,000 - 250 sq km

Something along those lines. CMAs, urban areas and cities, all into 1 giant statistic and breakdowns for all of us to enjoy and analyze.

your numbers are not up to date for Toronto and montréal . you need to use the Island of Montréal.
https://en.wikipedia.org/wiki/Urban_...on_of_Montreal
2,000,000 - 498km²

[eemy]

Originally Posted by YannickTO

Not easy to measure maybe you're right... (although I doubt it's that difficult to do), but for comparing cities and urban development, the ultimate real list is the urban figures list.

The municipalities list (subdisivisions) and CMA list (the metro areas) are the worst lists when it comes to comparing cities between each other. The urban list definitely gives a more accurate look to our urban environment. Don't you think?

You really think we can fairly compare a municipality of 700 square kilometres to a municipality of 20 sq km? CMAs also have many different geographic sizes and rules (Oshawa and Hamilton not included with Toronto is one), and the whole list gets blurred and non logical. On the other hand, the urban area lists provides an excellent look and fair comparison between our urban centres.

The thing with CMAs is that despite the fact that you get major variations in CMA size (say between a fairly tight boundary around KW versus the massive boundaries of Halifax), the population is still fairly representative and comparable for both due to the fact that the rural fringes tend to have very little population. Urban Areas tend to be more representative of the built environment, but the consequence of their delineation means that areas can be missed that represent hundreds of thousands of people as is the case in Ottawa.

Both have their place, but you have to understand how they can be utilized. The rules for CMAs are designed to reveal economic connection to a city. This is why they are built on commuting patterns. The CSD (municipality) is the building block because the data is more useful to governments that way. This is largely why Urban Areas aren't really that useful when it comes to policy discussions.

Urban Areas are meant to reveal built form, delineating urban areas and rural areas. They can be useful for understanding how population is spread out in a given jurisdiction (lots of small settlements, a few large ones, or concentrated in one, for example). They are built on population density within census tracts and the determination of whether to consolidate them into a single urban area is based entirely on how close together they are to one another (4km spacing, I think). This means that economic ties that aren't evident in the built form aren't reflected in the result.

A basic rule of thumb: if your analysis is concerned with area (density for example), urban areas are a useful tool of comparison. If not, CMAs will likely be better.

[Acajack]

Originally Posted by jeremy_haak

The thing with CMAs is that despite the fact that you get major variations in CMA size (say between a fairly tight boundary around KW versus the massive boundaries of Halifax), the population is still fairly representative and comparable for both due to the fact that the rural fringes tend to have very little population. Urban Areas tend to be more representative of the built environment, but the consequence of their delineation means that areas can be missed that represent hundreds of thousands of people as is the case in Ottawa.

Both have their place, but you have to understand how they can be utilized. The rules for CMAs are designed to reveal economic connection to a city. This is why they are built on commuting patterns. The CSD (municipality) is the building block because the data is more useful to governments that way. This is largely why Urban Areas aren't really that useful when it comes to policy discussions.

Urban Areas are meant to reveal built form, delineating urban areas and rural areas. They can be useful for understanding how population is spread out in a given jurisdiction (lots of small settlements, a few large ones, or concentrated in one, for example). They are built on population density within census tracts and the determination of whether to consolidate them into a single urban area is based entirely on how close together they are to one another (4km spacing, I think). This means that economic ties that aren't evident in the built form aren't reflected in the result.

A basic rule of thumb: if your analysis is concerned with area (density for example), urban areas are a useful tool of comparison. If not, CMAs will likely be better.

Taking the urban area theory to the extreme, Burley Griffin or Oscar Niemeyer designed cities like Canberra and Brasilia might also not be considered one city or a single urban area, given that they have a lot of green buffers between built-up areas, similar to what Ottawa has.

[someone123]

Originally Posted by jeremy_haak

The thing with CMAs is that despite the fact that you get major variations in CMA size (say between a fairly tight boundary around KW versus the massive boundaries of Halifax), the population is still fairly representative and comparable for both due to the fact that the rural fringes tend to have very little population.

One funny bit of trivia about the Halifax CMA is that in the old regional council for the amalgamated municipality, the easternmost 2/3 of the total land area got one councillor out of 23. Most of the people in that district live in the western part (i.e. in the bit of suburbia that had to be included in the huge eastern district to give it a population roughly equal to the others). There are around 10,000 people who live in the eastern half or so of Halifax's CMA boundaries.

Guysborough County NS is east of the Halifax CMA and has only 8,000 residents on 4,000 square kilometres. Nova Scotia's eastern shore is by far the most sparsely populated part of the province.

Far more people live west and north of the Halifax CMA and that is where the borders are tighter, because there are adjacent CSDs, independent municipality-level entities supported by the larger population. There are a lot of commuters from these places too but the commuter-dominated parts make up only a fraction of the adjacent districts so they are not included in the CMA.

If the borders of the CSDs were a little different you could create a Halifax CMA that would be half the size of the current one, more compact than many other CMAs, and would have maybe 450,000 people in it. The Enfield area is part of Halifax suburbia, for example.

Some people think that the large total land area of the Halifax CMA means that the population number is inflated with respect to other metropolitan areas but that's not really true.

[eternallyme]

Also the definition of urban needs to change. I think it needs to be limited to built-up areas of 20,000 of more with densities at least par to today's suburbs (single family home based).

If that definition was used, I believe Canada would be about 65-70% urban instead of 80%.

[YannickTO]

Originally Posted by eternallyme

Also the definition of urban needs to change. I think it needs to be limited to built-up areas of 20,000 of more with densities at least par to today's suburbs (single family home based).

If that definition was used, I believe Canada would be about 65-70% urban instead of 80%.

Well, I do love small urban centres as well. There's nothing I like more than my little urban areas of 1,000 people, in remote northern B.C. or Nfld.

When I'm on Google Earth, I almost prefer looking at small urban centres over the major ones. Having a 20,000+ list would limit the list to about a handful of cities. I want the WHOLE list with more than 1,200 localities.

[waterloowarrior]

The definition of 'population centre' (new name of 'urban areas' since 2011 Census) hasn't been released yet. Curious if they are doing some more tweaks.
https://www12.statcan.gc.ca/census-r...eo049a-eng.cfm

Another related concept is 'core' 'fringe' and 'rural area'
http://www12.statcan.ca/census-recen...geo052-eng.cfm



re: size

Population counts for population centres are published according to the class of population centre, regardless of whether they are inside or outside of a CMA or CA. Population centres are classified into one of three groups, depending on the size of their population:

-small population centres, with a population between 1,000 and 29,999
-medium population centres, with a population between 30,000 and 99,999
-large urban population centres, with a population of 100,000 or more

[GlassCity]

I agree with the definition of "metropolitan areas" used in CMAs, but I don't agree with the fact that they can't be merged or altered as these conditions change. This leaves strange situations where Abbotsford is separate from Vancouver, Oshawa is separate from Toronto, and worst of all, Okotoks is separated from Calgary. Doesn't give accurate impressions of true city sizes.

My own personal definition for metropolitan boundaries is quite broad, as I think municipalities within even large regions are much more interconnected than we pretend. I hate the needless divisions that arbitrary boundaries create. So I use local transit service. If you can get from one place to another via local transit, that's a tangible way to demonstrate connectiveness within an urban region. In this way, Abbotsford and Chilliwack are part of Greater Vancouver, and Oshawa, Hamilton, Niagara and Waterloo are all part of Toronto. Seems to me a more accurate description of reality.

[isaidso]

Originally Posted by GlassCity

I agree with the definition of "metropolitan areas" used in CMAs, but I don't agree with the fact that they can't be merged or altered as these conditions change.

It's done so that data from year to year is comparable. If one combines metros one can't calculate changes to that metro over time periods as separate numbers would no longer be counted. If one really wants to figure out the true population in the metro it's very easy to just add them together. For Toronto: Toronto CMA + Oshawa CMA.

From a statistical pov the current system makes more sense as one can garner more information out of the data.

[speedog]

Originally Posted by GlassCity

I agree with the definition of "metropolitan areas" used in CMAs, but I don't agree with the fact that they can't be merged or altered as these conditions change. This leaves strange situations where Abbotsford is separate from Vancouver, Oshawa is separate from Toronto, and worst of all, Okotoks is separated from Calgary. Doesn't give accurate impressions of true city sizes.

My own personal definition for metropolitan boundaries is quite broad, as I think municipalities within even large regions are much more interconnected than we pretend. I hate the needless divisions that arbitrary boundaries create. So I use local transit service. If you can get from one place to another via local transit, that's a tangible way to demonstrate connectiveness within an urban region. In this way, Abbotsford and Chilliwack are part of Greater Vancouver, and Oshawa, Hamilton, Niagara and Waterloo are all part of Toronto. Seems to me a more accurate description of reality.

Can one use local transit as a measuring stick? In Calgary's case, Calgary Transit does not provide service outside of the city's boundaries. There are specialized commuter transit buses running out of Airdrie, Cochrane and Okotoks into Calgary but I don't think one could consider them a part of Calgary's local transit system. Also, communities such as Chestermere and Langdon, which are currently in Calgary's CMA do not have any type of transit service into Calgary and Chesteremere is a city of 16,000+ sitting pretty much a hop, skip and jump to the east of Calgary.

CMA's, urban areas, whatever one wants to use as a defining measuring stick - there just isn't one way that'll work well everywhere. Put it this way, the community of Entwistle is some 95km west of Edmonton is a part of Edmonton's CMA while Okotoks which is around 15km south of Calgary isn't a part of Calfary's CMA. Better yet, Heritage Pointe which abuts Calgary's southern border is not a part of Calgary's CMA and I would suspect most people would have more difficulty trying to understand that Heritage Pointe isn't considered part of Calgary's CMA while Entwistle is a part of Edmonton's CMA.

[Rollerstud98]

Does Okotoks have transit to Calgary? Unlike Airdrie, Cochrane and Chestermere there are not even direct highways leading from Okotoks to Calgary

[YannickTO]

Mmmmm, reading all those replies now, I doubt there will be much change in the upcoming data. Population centres will be the same rule, same with Metro Areas.

Hamilton CMA being separate is 1 thing (Hamilton is historically more significant), but Oshawa is a mystery. It's clearly way way more a direct suburb than Newmarket is.

And why wouldn't Guelph be added to Kitchener CMA while St. Catharines has the whole Niagara region? Things don't add up here.