Some of you may have noticed that during the early part of fall, and again in the spring, I include with the forecast the probability of precipitation falling in the form of snow. By the middle to the end of November that probability is usually over 90 per cent, so I no longer include it with the forecast. To determine this number really isn’t very difficult. I simply take the number of days with recorded snowfall and divide it by the number of days with recorded precipitation. A year or so ago I created some different snow probability statistics, so I thought I would share some of that information with you now.
The snow statistics I created only use Winnipeg’s data, but after looking at Brandon’s data, I feel that this will be fairly representative for much of southern Manitoba. A couple of areas where these values will probably be off a little bit are for the areas along the escarpment, where larger snowfalls happen a little more often due to the orographic effect of the escarpment, as well as in the extreme southwestern corner of the province.
There are a number of ways to examine snowfall data, but for this look at snowfall probability I chose to examine two areas:
1) Probability of the amount of snowfall you could expect during a single snow event; and
2) Probability of the number of snowfall days you could expect each winter.
When I examined how much snow falls on any given day in the winter I was actually a little surprised at what I found – at least at first. If you look at the probability graph, you can see there is a 50 per cent probability that the amount of snow to fall in one day is about 1.5 cm. Over half of the time it snows, we see less than this amount. If we look at the 10 per cent probability, we find that we will see around five cm of snow. This means that for 90 per cent of our one-day snowfalls, we should expect to see less than five cm of snow. After thinking about it for a while, that actually sounds about right. After all, we only remember the big snowfall events – all the little ones where we don’t really have to clear any snow are simply forgotten.
The second snow statistic we’ll look at is the probability of the amount of snowfall you might expect during a snow event – in other words, a snowstorm. This looks at more than just how much snow fell on any one day, but takes into account consecutive days that recorded snow, calling those a single snow event. Understandably, the amounts go up compared to the single-day event, but as you’ll see, we don’t get big dumps of snow very often!
The probability of receiving 10 cm of snow during a snow event is around 10 per cent, or, in other words, every 10th snow event you could expect to see 10 cm of snow. If we jump to 20 cm the probability drops down to around three per cent, at 30 cm it’s down to one per cent, and if you get a storm with around 40 cm of snow, the probability is about 0.2 per cent, or once in every 500 snow events. To help put this into perspective, the average number of snow events per winter is around 27. So that’s roughly 18 years worth of snowstorms to get one 40-cm event. Overall, a good chunk of our snowfalls are less than two cm! So if you thought right from the start that getting big snowfall was unusual, you were right!