Post by BCcampus Research Fellow, Rob-Roy Douglas, Northern Lights College
For some reason weather has been on my mind lately — possibly due to rainfall and mudslides in the south and rapidly oscillating weather in the north (from +7 degrees Celsius to -23 and back again). Weather, of course, means meteorology, which is the science of the atmosphere. Meteorology is very much a statistical science, based on a combination of traditional weather recording and observations that now stretch over centuries and new predictive forecasting based on probability algorithms that attempt (with mixed success) to predict the weather. These algorithms are in turn based on models that draw on all those weather records. So meteorology is a statistical science in two ways: it is based on traditional statistics — the accumulation of large quantities of historical data, but these in turn serve as a means of generating probabilities about future events.
In Canada there are weather records as far back as the first European visitors. Environment Canada has data from some of its weather stations from 1840 onward, and sources as diverse as the Jesuit Order, Royal Navy, British Army, and Hudson’s Bay Company added to the weather records in the country. This is a large database now supplemented by satellite observations of regional and global weather patterns. But still, somehow it is either colder or warmer (or wetter or drier) than forecast when we go outside in the morning, which is why most weather forecasts deal in probabilities rather than absolutes. Weather is complex and involves many probabilities, so it is challenging to model statistically.
Most of us notice meteorology only when it fails, when something happens that wasn’t predicted. This may be changing, as extreme weather leads us all to become more familiar with terms like weather bomb and atmospheric river that have more in common with meteorologist slang than the pop-culture references that usually dominate our everyday language. Even so, we remain passive consumers of weather information for the most part rather than actively interrogating our environment to determine how weather is changing and what it might be like tomorrow, next week, or next year.
Traditional Indigenous approaches to weather and weather forecasting are interesting, because they are fundamentally statistical but in ways that are hard for Westernized, modern people to understand. The problem is partly one of scale and partly epistemological. We are now accustomed to think of weather as large-scale, symbolized by weather maps showing formations like jet streams, high-pressure cells, low pressure cells, and so on over all of Canada and its environs. Modern Canadians are also conditioned to regard weather forecasting and observation as expert science (which is why we call it meteorology) and expect someone with special knowledge to both generate these maps and explain them to us.
Beyond this, Canadians, like other Western peoples, regard weather as something that is epistemologically separate from them. It is part of the natural world, separate from the human world. This separation is why Western science regards weather as something to be classified, catalogued, and perhaps even controlled. For traditional Indigenous cultures, weather and nature itself are part of what it is to be human. Weather needs to be understood and accommodated in much the same way as we understand and deal with any other aspect of the world, human or otherwise.
Understanding starts from the same base as Western meteorology. Traditional Indigenous approaches to weather forecasting are also based on accumulated data, often over generations. Fundamentally, this is statistical, and I encourage my students to approach it from that perspective. However, these statistics have always been accumulated on a personal scale, based on individual observations and passed on from generation to generation. Their interpretation is based on algorithms that are individual and social and evolve with the input of new information. This is not really all that different from Western meteorology, but it is obviously not formalized or written in textbooks. It can also be remarkably more sophisticated. In a study of weather changes at Clyde River, Nunavut, published in 2009, the authors noted that local Inuit had identified changes in wind patterns and seasonal shifts that went unrecorded by the local meteorological station. In effect, the local hunters were able to identify the changes in weather caused by climate change before instrumentation detected it. They had to: both their lives and their livelihoods depend on understanding weather patterns, and climate change was already making this much more difficult (Gearheard et al. 2009).
In the context of a statistics class, it is useful to assign students to engage with weather through making some observations, generally simple ones, such as changes in the weather and what they notice precedes or follows it. This can take the form of a simple climate diary over about two weeks, which gives students a base of data to think about statistical record-keeping and how it can inform predictive modelling (the subject of a later lecture). More important, they engage with weather personally and, by sharing their observations with the class at large, get a sense of how we as humans can do our own internal and social predictive modelling — using that much more sophisticated computer, the one located handily within our cranium, to predict the weather instead of one made of silicon.
This research is supported by the BCcampus Research Fellows Program, which provides B.C. post-secondary educators and students with funding to conduct small-scale research on teaching and learning as well as explore evidence-based teaching practices that focus on student success and learning.