Published on Sep 13, 2010
After posting about the efforts of climate scientists to build a global climate databank for predicting extreme weather events, I got a note from Dave Jarvis who recently built a cool open source tool for exploring weather trends across Canada since 1900. Just choose a location in Canada and within seconds you can pull up historical trends ranging from minimum, mean and maximum temperatures to average daily rain and snowfalls.
Apart from the fact that it’s cool, open source and freely available to anyone using the Web, there are three things worth noting about the tool.
1. It’s fantastic that Jarvis is doing this because this is exactly the kind of tool that can make the science of climate change more accessible to the public. Jarvis concedes that the results are not 100% scientific, but perhaps that could change with a little additional investment. For example, a brief tutorial on how to interpret the data would help. It would also be cool if users could upload their own custom reports and make them available for comment and debate. Environment Canada could host something like this on their website. I could also see a tool like this being used in classrooms across the country.
2. Using good design sense, Jarvis built multiple levels of complexity into the interface. The most basic users can generate simple scatterplots, while advanced users get to control a variety of variables (e.g., elevation, more precise date ranges) and even the type of data modelling used to generate their reports. This fits with my general proposition that the best way to maximize participation in a collaborative project is to allow for differing levels of engagement among users and contributors.
3. Contrary to what I reported last week, high-quality climate data is not as readily available in North America and Europe as I had first thought. I suggested that it was mostly countries outside these regions that opt to keep their weather station data proprietary, thus creating a rather large obstacle for scientists hoping to create a broadly-accessible global climate database. Not so, says Jarvis, who has encountered problems getting data from countries across Europe and here in NA too. Jarvis reported having to pay Environment Canada $100 for the weather station database he is currently using, which is not too steep. But a dataset of similar quality in the US costs $100,000. According to Jarvis, there are other repositories of US weather data that are free, but the datasets do not go back as far and do not have the same geographic coverage. Nevertheless, he intends to begin integrating the free US data soon.
Jarvis claims that he has not uncovered any comparable tools in his research and neither have I. If you happen to know of any, please let me know!