Map of the Week: Smoking, redux

STEVE RUSSELL/TORONTO STAR

Maps:

Small (1000x540)

Medium (1200x700)

Large (1600x700)

---

One of the limitations on our online mapping projects has been that maps with large numbers of polygons (or two-dimensional shapes, as seen in a map of city neighbourhoods) are slow to display, especially in IE. Another was that the polygons themselves had to be created by hand. This meant that the 200-odd postal codes in the GTA were about the outer limit of what we could display. More ambitious maps, like one depicting all the electoral districts in the province, were impractical both for us to create and for a reader to load.

So I was delighted when Sanjay Singh of Pitney Bowes Business Insight approached us with alternative way of creating this kind of map which elegantly solves both problems. Without going into the details, it involves two techniques which I had tried at different times but not together, which turns out to be the key.

For the first week of Map of the Week 2.0, we have a map showing smoking patterns in the GTA (unfortunately not including Durham for now) by census tract. Just over a thousand areas are represented. This map loads much more quickly than our .xml format in Firefox and IE, but if there are any unexpected issues please let us know in the comments.

The map shows average annual expenditure on smoking.

Back in August, we created a smoking map covering the whole province by health region (identical to city/county/regional municipality in most cases). That map showed a concentration of smokers in rural and northern areas, and the GTA map shows something of the same pattern within our own region.

Rural/deep suburban regions spend much more on tobacco than the rest of the GTA, with the highest tobacco spending being around Kettleby. There is a pocket of very low tobacco spending in northern Brampton, interestingly.

Within Toronto, Lawrence Park, Hogg's Hollow, chunks of Forest Hill and the Bridle Path all have higher-than average tobacco spending, while the traditional low-income V seems to have lower rates. Looking more closely at neighbourhoods, there does seem to be a rough positive correlation between income and tobacco spending.

Comments

[Byron Moldofsky]

Hi: I teach a course in cartographic design and I am interested in the new method for displaying polygons faster you refer to as Map of the Week 2.0. Can you give me some more information about the method used?