But in a way, March might be the snowiest month anyway. Or not. You can be the judge.
I hear it all the time. I hear the words spoken in meatspace, I see it on Facebook. Just in the last week, as the snowiest February on record in the Twin Cities winds down, I’ve heard it four times, in the form of an ominous warning about what is to come next, tongue in cheek or maybe not tongue in cheek. You think February is bad? Well, March is the snowiest month! We’re in for it!
As a person who studies and tracks weather and climate, I have always known that March does not get the most snow in this exact part of the country. There are some places where it does. In parts of Montana, Wyoming, the Dakotas (mainly the one to the south), and here and there in the west, March has more snow, on average, than any other month. These regions are about a mile above sea level, so many storms that are rain in Minnesota, Iowa, and points east can be substantial snowstorms in the High Plains, and March is a wet month. You also get frost in late August on the High Plains.
But in Minnesota it isn’t true that March gets the most snow, though in a certain way one can be forgiven for thinking it the snowiest.
I looked at data from 1957 to 2016 to check this out. Why those years? I picked 1957 because a look at my data set from NOAA shows that that is the cleanest bunch, and I didn’t want to do any further cleaning, and my data stops at 2016 because that is when I downloaded the file. I’ve looked at the last few years directly, and at the older years going back to 1900, and everything works the same if I include them, but with less assurance that some data point isn’t going to be off by error.
Looking at the Twin Cities area (MSP airport data) the snowiest month, on average, is January (22% of the snow) with December a close second. March is third, and February is fourth, causing an interesting but persistent glitch in an otherwise normal looking distribution, a true bi-modality, with a brief and minor snow drought in February. Like this:
So, plain and simple, March does not have the most total snow, on average.
Another way to look at this is to list all the years for each month where that month was the snowiest that year. That might seem a little clumsy at first, but it provides a granular look at the data, and also accentuates the February phenomenon.
It is not apparent from this graph, or other looks at the data I won’t show you, that there has been much of a change in this pattern. (By the way, there is one tie between two outlier months, so there are two 1957s.)
There is, however, a way in which March can be seen as the “snowiest month.” If we look at just snow storms that involved one day accumulations of 6″ or more March had a lot more than any of the other months. (Note, this is not 6″ snow storms, but one day accumulations, which is a valid sampling but not the same thing.)
Note, however, that the seven snowiest months in the data set don’t include March. The snowiest month in this data set is actually a November, the year of the Halloween storm. The next several are almost all Januaries, and one December, until we get to March 1985 and March 1952. So it is not the case that the really big storms have been in March or have been recently, thus causing a change in perception. It is simply the case that March is when the wet season begins some years, which has traditionally been spring or early summer, and sometimes that overlaps with the cold, and so we get these big snowfalls (with the occasional one happening in April). And it has always been this way. Well, not always, but in everyone living’s lifetime.
But what about the ominous statements people are making, that since this February was a record month for snow, and since March is “the snowiest month,” that we are going to get even more snow over the next four weeks? A look at the data addresses that issue.
For the fear to be valid, it would be the case that many, or most, March snowfall totals are higher than February totals, regardless of how much snow fell in February. Since the mid 1950s, March had more snow than February 33 times, and the reverse was true 27 times. The average snowfall in a February that was followed by a snowier March was 5.5 inches, with a maximum of 17 inches, while the average February followed by a less snowy March was 12.7, with a maximum of 26.5, during this period. In other words, based just on what we see in past years, and ignoring actual meteorology or logic or statistics, the chances of a March having more snow than February is attenuated during years with a very snowy February. That does not mean that March gives us a break when February is very snowy. It just means that March does not actually say to February during those snowy years, as has been suggested, “Hold my beer.” Chances are this very snowy February will not be followed by an even snowier March.