How we forecast the weather | News
Well, the New Year is but three days old as I write this and we have already experienced colder conditions than were observed at any time in November and December.
As some of you know, one of the services I provide to clients through my forecasting company, Weather-Track, is mid and long range weather projections.
Part of what I am trying to do in these blogs is give you, our viewers, a look at some of the thoughts and the challenges associated with attempting to project weather trends over a time span of weeks and months.
While computers do guide us, they are not as important or useful as recognizing global patterns and trends, some of which I addressed and defined in our first blog, back on December 1.
As you read through these blogs, I encourage you to email me your questions at firstname.lastname@example.org -- I will do my best to respond, as time permits.
About a month ago (December 7), I penned a blog that addressed our research which dubbed the winter of 1994-95 as an analogue season candidate (among several). (Recall that an analogue season is one where various current and projected global weather factors – teleconnectors, are aligned to what was being observed at some point in the past.)
I also warned that no two seasons are identical. And if forecasting the weather were as simple as finding a similar month in the past, I would be out of a job. In addition, there were some teleconnectors that did not fit neatly into this little hypothesis.
But utilizing Rochester data and data from Chicago (to more fully represent conditions in the Great Lakes) we discovered how similar November of 1994 was to November of 2011. They were near perfect matches, in fact, fitting in nicely into the global pattern.
So, one might wonder how December of 1994 and December of 2011 compared in these two cities?
In Chicago, with an average temperature of 34.9 degrees, December of 1994 ended up the 17th warmest in more than 100 years of record keeping. Meanwhile, December of 2011 was just slightly warmer featuring an average temperature of 35.4 degrees (a stunning 13 degrees warmer than last December). The total snowfall was 1.7 inches, about 7 inches below normal for the month.
In Rochester, with an average temperature of 34.8 degrees, December of 1994 was 5.4 degrees warmer than normal and there was just 7.6 inches of snowfall, a third of the normal. In December of 2011, the average temperature was 35.2 degrees and total snowfall was just 4.9 inches.
Clearly, these were very, very good local matches.
With that said and with the most recent global indices in mind, we have no reason to change our general long range thinking for the Rochester area, as addressed in that December 7 blog. Here is a portion of what I wrote:
“This winter will bear little resemblance to last winter (consistent cold, lots of snow, no really big storms). In fact, it will be much different with “inconsistent” cold, near to below normal snowfall and a higher probability of a big storm or two (but with Great Lakes-centered storms favored over east-coastals). Additionally, December will have a couple of arctic incursions complete with wind and lake snows, but will likely still average above normal in temperatures and below normal in snowfall. The late January to February to early March period will be coldest (relative to normal) and the snowiest in absolute terms and will turn harsh if the AO/NAO couplet goes strongly negative.”
As I conclude, Rochester has just experienced the coldest air of the season to date along with the first widespread lake snows. Indeed, the Halcyon days of December are over and Old Man Winter will be making his presence felt in the weeks ahead.
But while January will finally be bringing our area spells of fairly impressive cold and snow, there will likely still be a thaw or two to frustrate the local winter weather enthusiast and challenge the forecaster attempting to gauge precipitation types. And depending on the future state of the AO/NAO couplet, if there is to be a period of longer lasting, extreme winter weather, it will likely occur in late January and February, at least that is what our analogue method would suggest.