mapping

I have been MOOC’ing this summer and learning how to do maps. Geography as an adult is much more fun than my 10th grade geography class.

Chad Skelton over at the Vancouver Sun intrigued me with his data retrieval and mapping of British Columbia’s Smart Meter uptake. if you’re not from BC, here’s a short intro (#BCpoli-aware feel free to skip the next two paragraphs).

BC Hydro is the government owned (Crown Corporation) utility that produces and distributes electricity for the province of British Colombia in Canada. In 2011, BC Hydro announced its intention to spend $$$ upgrading all its electricity meters to “smart meters”. These meters are capable of being read via wifi by meter readers, and potentially also give BC residents the ability to monitor their electricity usage in near-real time.

Many concerns were raised about the smart meters. One was about the costs of the program vs. perceived benefits. The others, which gained traction were around an emerging movement in BC connecting wifi, cell signals and wifi-enabled smart meters with a whole variety of health effects. While few, if any of these health concerns have been actually causally linked to smart meters, or even to the amorphous descriptor “wifi radiation”, these health concerns have gained traction even among official bodies such as the Union of BC Municipalities, municipal councils and school boards. The BC provincial election in 2013 was a chance for people to voice their concerns. The opposition parties all brought the issue up during canvassing.

For my peer assessment mapping project, I wanted to see if areas of relatively high smart meter refusal were correlated or co-located in any way with voting against the ruling BC Liberals.

This is the map I made, my first ever map not scrawled on paper.

View Larger Map

Reading the Map

The electoral districts are colour-coded by BC Liberal Party percentage, darker means higher vote for the BC Liberals. I chose this rather than “who won” because I was looking more for an anti-BC Liberal effect. I will, at some point in time, try to overlay “who won” as well. The smart meter refusal data is in three different coloured and sized circles. Large and red means higher refusal, and small and green means low refusal. This is a hybrid of a graduated circle symbol scheme and a diverging colour scheme. Clearly, using points to represent areas is a big limitation, but it is sufficient for a quick peek.

Anything to See?

• An overwhelming majority of people had smart meters installed, > 90% in most places. So, BC Hydro’s brute force, no options, default installation plan was mostly successful
• Places of higher than normal refusal tended to vote against the BC Liberals. I believe this had more to do with existing anti-BC-Lib tendencies influencing smart meter refusal rather than the other way around.
• Urban centres like Victoria and Vancouver had relatively low rates of refusal. Is this because of higher apartment proportions, or because smart meter refusal was restricted to a small number of high information, highly motivated individuals whose number varied by location and whose numbers in places like Victoria were muted by larger populations?. Note that my home area of Victoria had the most (7300) rejected smart meters, even though the percentage is small. The ageing white (l)iberal enclave of Saltspring Island (Ganges), aka hippieville, Canada had by far the highest refusal percentage. So, is this smart meter refusal map mostly a hippie population distribution map?

The take home message for me was that the anti-smartmeter movement had little influence on the election, which was most likely won on the usual and mundane issues of the economy, trust and corruption.

Methods

1. I downloaded data on smart meter refusal from the Chad Skelton’s post and Tableau public
2. The data from BC Hydro is categorized using their division of BC into distinct geographical billing areas. I used billing area names to geotag the information. The site http://www.findlatitudeandlongitude.com/ has a feature where addresses can be uploaded in bulk via a text interface, and the site returns the place, and latitude longitude. I added province and country to the place names, and edited ambiguous names to make the search more effective.
3. I uploaded this table to arcgis to form one layer. Arcgis is a big and expensive GIS software, with a limited free online playpen where this map is displayed. I used graduated circles and natural breaks to represent the different levels of smart meter refusal. A big limitation to this approach is that the BC Hydro billing areas are just that, areas, not points on a map. However, the area boundaries are not available as a shape file, and geographical areas vary widely. So, the points correspond to the centre of the nearest big population area mentioned in the BC Hydro billing area description
4. I downloaded BC electoral district shape files from Paul Ramsey of Open Geo. These shape files are an improved version of those available from Elections BC, again, thanks to Chad Skelton for pointing me in this direction
5. Elections BC lists 2013 provincial election results information by party by district. However, there is no publicly downloadable mapped source for the election data results. I used the open source GIS desktop software QGIS to open the shape file and add the attribute of BC Liberal percentage to the shape file. I uploaded this shape file to arcgis and layered it with the smart meter refusal rate graduated circles to look for patterns.

Maps are fun to play with, and I know very very little about them, which is a great combination. Every minute I spent making this map was a learning experience. Comments and feedback, please. I think I will slowly incorporate mapping into my skill set. But I think I will use open source/free solutions in the future.