Atlas of NSW a Pleasant Surprise

Last month, the NSW government launched its online Atlas (in Beta). Developed under the radar, it would have taken many by surprise. And surprise me it did. Not in the least because government mapping sites, especially in NSW, need to be viewed with a healthy dose of suspicion. With a few exceptions, we have over the years been underwhelmed with clunky, slow, unusable and unmaintained mapping initiatives that – if they are still around – are gathering dust, rather than servicing taxpayers.

So I had a look at the NSW Atlas with some trepidation and maybe a little sense of impending doom. However, I was pleasantly surprised with that I saw, though there are a few concerns.

The Atlas of NSW does many things right. First and foremost: it’s an actual Atlas in the proper sense of the word; that is it presents information in a number of topics, where appropriate in map form. Thus the information provision is the starting point of the Atlas, and the mapping functionality (part static maps, part dynamic through the ‘Atlas Explorer’) is quite rightly subservient to this.

Secondly, the site is slick, well designed and performs fast, including the Atlas Explorer, which zooms, pans and redraws layers at a speed that we’ve become used to from Google Maps.

All in all, it’s a good-looking and fun resource to explore, with topics ranging from Economy, Elections, Environment, and History, to People. The content is illustrated with plenty of well-designed maps and infographics such as this trade-routes map.

Having heaped all this praise on the NSW Land and Property Management Authority (LPMA) who published the Atlas, now is a good time to voice thee key concerns I have with the site:

1. Conceptually, the Atlas is very similar to the Australian Natural Resources Atlas, which was launched about a decade ago. At the time, ANRA was ground breaking, but as a static snapshot, it was not maintainable and the data quickly became outdated and 'stale'. In that context, it must be hoped that LPMA have a good maintenance plan (and budget!) in place, or risk a very short Atlas lifespan;
   
2. As the latest Apps4NSW contest (and many similar initiatives) has shown, making government data available under a gov 2.0 agenda, works best when the raw data is published in addition to sophisticated viewing applications such as the Atlas, preferably as web services. I’m giving LPMA the benefit of the doubt here (despite previous track record), and am looking forward to seeing the dataservices in the next release, so that the (mobile) app developer community can go knock themselves out;
   
3. And finally one of my pet soapbox issues: “thou shalt not visualise absolute values in choropleth maps”. Choropleth (shaded area) maps are not meant to be used for the presentation of graded absolute values such as total population. This presentation provokes a wrong data interpretation.
   
   Look at this example from the Atlas – a map of Total Population. The darker blue area is the Shire of Cobar. To the East is Bogan Council. Clearly lots of people more in the Cobar area, right? Well, yes and no. You will notice that Bogan covers a much smaller area, it is actually only 1/3 of the size of Cobar. So though Cobar has more people (5200 over 3000), its population density is only half that of Bogan (0.11 people per km2, with 0.21 people per km2 for Bogan). So, on average, Bogan has almost twice as many people on its land surface than Cobar.
   That is why every cartographic textbook will tell you that Choropleth maps should only be used for relative data such as population density, labour participation rates, or average family income.
   
   But hey, in this era of spin, maps can lie just as easily as statistics.