Forecasting A Winter Season (Part Three)

Our next look at the upcoming winter season, this post concentrates on the potential usefulness of "analogous years" in anticipating future snowfall.

This post builds upon the work of the previous in this series on the upcoming Winter Season Forecast.

In the previous blog post in the series, we talked about the possible storm tracks we will be looking at that represent high-precipitation events for the eastern half of the country that involve our region.

Today we'll examine a seasonal forecasting concept called "analogous years".

This means, after first understanding the larger-scale sky/ocean patterns that are in place, and further having a good account of how the summer/fall seasons have gone within that framework, the meteorologist would then identify other years in the 100-or so that exist in recorded data in which those conditions were also similarly seen. The idea here is to assess how the Winters went in those analogous years to get a handle on how the anticipated weather patterns will translate into snowfall.

This can be a tricky exercise, because the temptation is to choose analogous years with one eye peeking at the winter weather that followed. For snow lovers, this means highlighting big snow years that as-closely-as-you-can-find matches the weather of the Spring-Fall before it. I don't want to do that though (but I'd like to). What I'm looking for is the overall flavor of this past late-Spring through early Fall as best represented in a previous year. I'd also like to further winnow these potential analogous years down further by seeing which ones of those have similar climate teleconnection patterns (I mean, similar El Nino / La Nina climate stuff as explained in the first blog post on this subject).

The main elements that I'd like to use are the following:

  • An ENSO pattern that features a weak El Nino (actually, a slightly to the El Nino side of "neutral")
  • A widespread drought that occupies the entire summer (and even beyond)

With these two criteria, two good years jump right out at me that are relatively recent: 2007 and 1988. Both years featured a hot/dry summer with many 100+ degree temperature days, and 1988's summer drought is pretty much the drought of comparison to the one we had this summer. They both were also similar years in the El Nino category.

So, when you look at those years, how did they do for snowfall…?

City Name:

Dec Snow(in.)

Jan Snow(in.)

Feb Snow(in.)

Total Snowfall
(Whole Season)

Huntington 1988-9

2.5

0.3

2.0

5.0”

Charleston 1988-9

6.9

1.7

4.7

15.0”

Huntington 2007-8

1.4

2.0

      10” est.

     13.4” est.

Charleston 2007-8

3.6

7.0

7.9

22.4”

Examining this data, if it is assumed that the weather patterns from these years will similarly repeat themselves in the coming season, then we can expect a weaker than normal snow season, but one with a stronger finish after the turn of the year.

(Oddly enough, the National Climate Data Center and the Northeast Regional Climate Center both have missing data for February in 2007-- convenient. But, fortunately we've got Tony Cavalier's records for the same period).

Nice idea, but...

One of the biggest pro's to the "analogous" year technique is that it can give us an actual data point or two into how the mass movings of the atmosphere and oceans during the Spring-Fall seasons translated into snowfall. The idea being, if these large-scale circulations and oscilations do influence seasonal weather, then we should be able to entertain their predictive power in years that have already happened.

The unfortunate reality of such an approach is that there is little data, even over 100 years, to place a whole lot of stock into the method. The National Weather Service has weighed in on this point during a recent conference call on the upcoming winter. They said that in order to have a statistically robust estimation of the upcoming winter, they would prefer 1000 years of data to go on. This would give each year a slew of 'analogous' years so that a trend can be spotted-- rather than only a few data points over the last century.

So, there may be several other years that other meteorologists have come up with that can be said to be 'analogous'-- I suspect this approach can have a lot of variability to it, and each meteorological eye may spy something different. However, when it comes to forecasting the upcoming season, I'll take every tea-leaf I can read ;-)

Don't forget, our "Winter Weather Preview" is coming up this very week. Thanksgiving Day, at 5:00pm (and then replayed a few other times I'm sure). There is where you will first see our final 'snowfall forecast' numbers, among other things. However, I do want to let you know that I will be expanding upon that in a blog post that will be here as well, probably the weekend after Thanksgiving, once I'm done sleeping off the turkey :-)

Any questions and comments are well appreciated, and can be inserted below. I will get to them whenever time permits.

 

Read More Blogs
Comments are posted from viewers like you and do not always reflect the views of this station.
powered by Disqus
WSAZ NewsChannel 3 645 Fifth Avenue Huntington, WV 25701 304-697-4780 WSAZ Charleston 111 Columbia Avenue Charleston, WV 25302 304-344-3521
Gray Television, Inc. - Copyright © 2002-2014 - Designed by Gray Digital Media - Powered by Clickability