Snowiest March in decades! Is winter over yet?

An image from Front Royal, VA
during the March 17th storm.

An image from Spotsylvania, VA
during the March 3rd storm. 

The St. Patrick's Day snowstorm we just endured will go down in the record books as the biggest storm on that date in DC history. The previous St. Patrick's Day record was only 1.9" of snow. While this storm easily broke the record with 3.9" on the holiday itself, the real wallop was with the storm total. We got 7.2" of snow at Reagan National from Sunday night into Monday morning! This storm, plus the early-March snow and ice, bring us to a grand total of 11" for March 2014. This is the snowiest March on record since... can you believe it... 1960?!?! It's been half a century since DC has had this much snow, this late in the season!

"SnowPatricksDay", as it was dubbed, seems to be the last straw for many Washingtonians, even those who normally love the snow. Our seasonal snowfall total is now almost exactly double the 30-year average, at 30.3" (The average is 15.4"). However, this number is still not high enough to crack the top 10 snowiest winters in DC. So now the question is, will we get one or two more snowstorms that push us over the edge, and into the top 10?

While it is possible, of course, that we could get another big snow storm or two, it's highly unlikely. As of today, we are exactly nine inches short of creating a tie for the 10th-snowiest winter in DC (39.3" in 1910-1911). A snowfall of nine inches or more in DC has only happened twice this late in the season in recorded history (in 1942 and in 1891). And these storms were extreme outliers; on average, after March 17th, we only get 0.2" of additional snowfall for the entire season.

April is notoriously sparse for snow-lovers. The most recent April snow at Reagan National happened in 2007, but it was a paltry 0.4" of accumulation. Since 1888, there have only been nine April snowfalls of 1" or more, with the most recent one happening in 1924! Yep, that was quite awhile ago. We can only expect accumulating snowfall (more than a trace) at Reagan National about once every ten years. Add in trace snowfall events, and the number only increases to two out of every ten years. That means, most of the time, we do not see any snowflakes in the month of April in Washington. By stark contrast, DC gets accumulating snow nine out of every ten Marches, and we have never had a March in recorded history without at least a trace of snowfall in the month. So, the frequency of snow events rapidly declines from March to April, and so does the amount of snow in each event!

One last thing... a snowy March doesn't increase the odds of getting snow in April. Since 1888, there have been 24 Marches with at least 6" of snow. Only three of those Marches were followed by snow in April of more than a trace. Statistically, it means that there is no correlation. This winter has been surprising at every turn, though, so maybe we'll get one last surprise to finish out this sneakily snowy season!

Long-range forecasts are "Snow Joke"

There has been a LOT of buzz on social media and around the watercooler over the past couple days about the potential for a major storm next weekend. No, not Super Bowl weekend like the Farmer's Almanac predicted, but the following weekend. I hope that this talk raised your eyebrows and your level of suspicion about the accuracy of such a forecast. 

The biggest problem with this forecast is that it is way too far in advance. The rumor mill began on January 29th, which is 12 days prior to the supposed arrival of this storm. On that date, the European computer model's long-range output showed a massive storm for the Mid-Atlantic region. But that is just one run from one computer model; it's not nearly enough information to make any sort of prediction! It's like seeing a photograph from a city block away, through a frosted pane of glass, and saying that your view of the photo is crystal-clear. I'm going to show you a few maps so you can really visualize how much the forecast can change, based on small changes in the model output!

First of all, the forecast will change from one model run to the next. A model run is each instance that the forecast model receives new data ("initial conditions" from satellite and radar data, weather balloons, and other sources). Most forecast models are run 4 times per day. 

The GFS, or Global Forecast System, is the primary model used and maintained by NOAA. Last year, the GFS received the first of two upgrades, which will boost its computing power tremendously. (You can learn more about the upgrades here.) Even with the upgraded computing power, the model still has wide variations from one model run to the next! Take a look at the different forecasts from the GFS model, valid for 12Z Sunday. The map shows the GFS's prediction for precipitation during the 6-hour period leading into 7am on Sunday, February 9th. 

These two model runs happened 6 hours apart from each other, and check out the huge difference in the placement of the precipitation! The same computer forecast model came up with two very different ideas for Sunday morning's precipitation potential. 

Now, let's compare two forecast models against each other: the GFS and the ECMWF (the European model). This time, I'll use the same model initialization time, but I'll still show you the same forecast time of 7am on Sunday, February 9th. 

Here, the difference is even bigger! The European model keeps us dry during the early morning hours... a huge contrast to the approximately 1.00" of liquid precipitation over the same 6-hour period that the GFS model is predicting. This model variability is extremely common in the long-range, which is why we professional meteorologists don't put any stock in any extreme weather predictions that they might produce! 

I want to make something clear... I'm not saying that a big storm is impossible in the 2nd weekend of February. All I'm saying is that it's way too early to make any sort of prediction. Hopefully, now that you've looked at the highly variable model data with me, you can see why it's too early to make such a forecast, and you understand why there aren't any reputable weather sources that are making such a bold prediction!