Friday, October 05, 2012

On Snow and Rumors of Snow

We get two angles to attack the two-headed winter snow monster that's bearing down on the US here in early December October. Yes, this is a very cold air mass for this time of year, with departures of 20-30 degrees below normal in some cases over a widespread chunk of real estate. Some forecasters and weather enthusiasts have begun to get excited about the potential for snowfall to materialize as a result of this cold push and inbound moisture. I'm here to further Tom's post from earlier this week and briefly temper those expectations.

Upper air and analysis maps from last October's Blizzard.
Credit: Unisys Weather
First of all, what we know: Some computer models, namely the GFS, do continue to spit out minor amounts of frozen precipitation. It can certainly snow in October, as the image to the right will attest to (upper air maps from last October's blizzard). It's going to get quite chilly. Frankly this is all we really know right now. Let me discuss a few things working against accumulating snow...even at some of the higher elevations.

First of all, the calendar. While I admit that it is sometimes the most overused excuse in the world of weather, Sunday is only October 7th. The earliest snows on record in much of Northeast PA have occurred in the first week of October (it has snowed in August before, but for these purposes, that's not really relatable). Various records in Northeast PA around 1,500' elevation going back to well before 1950 show this, and most don't show the earliest first inch of snow occurring until around this October 10-20 timeframe. In other words, it would take essentially an unprecedented event to cause even a few inches of snow. This is cold for sure, and some areas of the interior will push record lows, but I'm not seeing anything that screams this is an unprecedented October cold shot. Routine? Perhaps not, but not unprecedented.

Monday morning output from yesterday morning's European model
Credit: PSU E-Wall
Let's tackle another big area. If you notice above in the upper air maps from last October (which WAS an unprecedented event), 850 mb temperatures (top left panel) were solid... -6 to -8C and lower in areas that got slammed. Let's compare to Monday morning as the core of the shot of cold sets in. Via the European model, which you see on the left, we're looking at 850 mb temps (bottom right panel) of 0 to -4C. The more optimistic GFS maybe sinks us to -5C. Let me expand on this briefly.

The best example of a true early season snowstorm (that people might remember) to me in the Northeast occurred on October 12-13, 2006 in Buffalo, NY, as lake effect snow developed and buried parts of the area. It took a lot of prodding, but eventually it started snowing and didn't stop til Metro Buffalo got an average of about 12-24" of snow. Based on my recollection and of the reanalysis maps from that event, it took 850 mb temps of roughly -5-7 to bring snow down to Buffalo...a low elevation obviously. It also had the support of a 62 degree Lake Erie allowing 25-30,000 foot thunderstorms of snow to develop. That would add a localized cooling effect and air temps during the snow dropped to 32 degrees or lower.

The bottom line is that for it to snow really anything above an inch or two before October 15th, outside the highest elevations or lake effect snow belts of NY, it takes a truly unprecedented weather event. I will not rule out snowflakes at fact, I think there's a good chance we see snowflakes in some of the higher terrain in PA, MD, WV, NY, or New England. However, I would hold off on predictions of anything more than a coating right now for those higher terrain locations at this time. Not all precip that falls into cool air is snow...and not often does it stick before Oct 15, especially after multiple days in the 60s and even some 70s...unless it's an unprecedented event.

Just a brief comment on actual snow...falling in the Upper Midwest. Up to 14" has fallen in parts of Minnesota (Roseau County). Given Tom's post the other day on the firestorm around The Weather Channel's unilateral winter storm naming plan, you may be wondering what name was given to our first winter storm of the season. Well, sorry to disappoint you, but Athena remains unused. This greatly illustrates the problem with this plan. It's entirely subjective and unfortunately cannot be used to accurately show how a winter did or did not behave. One foot of snow, even in Minnesota is a significant deal, especially in October. But since "no one lives there," it's not worthy of a name. To put it more lucidly, this is the equivalent of not naming a strong tropical storm that makes landfall in between Corpus Christi and Brownsville, TX because it's sparsely populated.

So it will go when you don't set parameters by which to assign names or scales to things.