Severe storms with torrential rain pummeling the metro, tap here for radar

Joe’s Weather Blog: Research-Do Fall Temps Influence Winter?

This is an archived article and the information in the article may be outdated. Please look at the time stamp on the story to see when it was last updated.

So for the last few days I’ve been compiling graphics and researching the question…does what happens in the fall with regards to temperatures influence the winter temperatures? If I had a dime for everybody who has asked via email or FB or whatever during a cold fall spell (like earlier this week) boy if it’s like this now…what about the winter?

In reality that has been asked a lot. If we get an early surge of summer heat in the springtime people will ask if that means it’s going to be a hot summer. So I thought I’d look through the last 10 years on a national scope and take a look at the fall temperatures then compare them to the following winter. I used October 1st through November 15th as my “fall” time period. In reality I probably could’ve extended this through the end of November but I started down a road and frankly I didn’t want to go back and start again. For the purposes of “winter” I extended the time frame from 12/1 to 3/15.

The reason why I started with 10/1 is that I felt starting on 9/1 brought too much of summer into the equation. Often we don’t get our 1st real fall front till mid September and we’re often still pretty toasty into that time frame. So I eliminated September. Granted I’m taking a lot of assumptions here but it was what I choose to do. I went back 10 years to do all of this and granted as well, it’s a rather short time frame.

With all that said I’ll let you look at the images of the fall vs the following winter temperatures and make your own conclusions. I’ll add my thoughts at the end of each comparison. I’ve also included the snow totals for that particular winter just as an FYI

So let’s start out with the last fall/winter span and work our way down to the last one for this research (10 years ago)

ScreenHunter_01 Nov. 16 12.52

In my opinion in the graphic above there is little to NO comparison. Weak similarities exist in central Canada and a stronger similarity is in the NE part of the country.

ScreenHunter_02 Nov. 16 12.56

This was one of the strongest cases on the YES side for the comparison. We started warm and never really looked back. This was also our lowest snow total for a winter season in KC recorded weather history.

ScreenHunter_03 Nov. 16 12.59

Definitely a big NO in my comparison. With the exception of the eastern reaches of Canada through Hudson’s Bay, for the US at least…not even close and really opposite across the N Plains states.

ScreenHunter_04 Nov. 16 13.01

An argument can be made that this trends towards the YES side…at least in terms of anomalies for most of the US. There is no doubt that the extent of the cooler air isn’t as dramatic in scope in the winter season…but the idea of the cool air certainly is.

ScreenHunter_05 Nov. 16 13.03

This one is more of a NO to me at least. Not as “NOish” as some other years but for the northern US above the I-80 corridor there is NO comparison. There are weak connections in FL and perhaps in the SW part of the country.

ScreenHunter_06 Nov. 16 13.06

Pretty big NO as far as I’m concerned. Some weak connections in the SE and maybe in the N Rockies…and the NE. But there were some major flips from one extreme to the other in the I-70/80 corridors from E>W…so that’s why I’m grading this as a NO.

ScreenHunter_07 Nov. 16 13.08

Another NO in my opinion. HUGE flip from a cold fall to a typical winter for temperatures in the OH/TN Valley region. Also an erosion of the warmth to average in the PAC NW

ScreenHunter_08 Nov. 16 13.11

This is a big YES. Warm in the fall and that followed suit in the following winter season

ScreenHunter_09 Nov. 16 13.12

I’m going to lean YES on this one as well. There was a major flip in the SW part of the country but overall the warm that was present in the N Rockies through the Plains continued the following winter and the cooler NE weather did as well.

ScreenHunter_10 Nov. 16 13.15

The last comparison is sort of a lean NO in my opinion. The strong warmth in the Rockies is mostly gone in the winter while the chill in the winter back east is not to be found in the fall.

So let’s add things up…I have 4 cases of YES and 6 cases of NO. There were some years where it wasn’t even close and others that were very comparable. In reality it’s my conclusion that there is no advanced forecasting conclusion to using the above hypothesis (fall temps=winter outcome) based strictly on the last 10 years of data.

No let’s take things to another level…and slightly more complicated. It’s well known that if you follow certain indices in the climate world, you can get a better hint about forecasting the weather in advance. I wanted to take a look at two indices that are most commonly used when trying to decide int he winter season 1) are the models right in forecasting the cold and 2) with the pattern “lock” in for a particular amount of time. Those two indices are the North Atlantic Oscillation (NAO) and the Arctic Oscillation (AO) values.

Let’s do a bit of explaining before we get into things. The NAO helps to measure the pressure fluctuations between two systems the mostly permanent Icelandic LOW and the semi-permanent Azores HIGH. When the values are positive (+) typically during the winter we trend milder…when negative the opposite holds true.

When the NAO is negative the pressures in the arctic region are higher. This leads to weaker jetstreams up there which allow the cold air to drain into Canada and potentially into the US.

(Graphic courtesy of the State Climate Office Of NC)

When the NAO is positive…the opposite holds. The surface pressures are lower in the arctic regions…this leads to increased jetstream flow from west to east and the colder air tends to remain more bottled up towards the arctic region.

(Graphic courtesy of the State Climate Office Of NC)

So what have the trends been during the 10 fall/winters that I analyzed…click on this image to make it larger and more readable.

ScreenHunter_11 Nov. 16 13.43

The left side of the middle line in the above graphic represents a negative NAO while the right side of the middle line is a positive NAO. Often times weak positives and negatives really don’t do much for us…we’re looking for more significant negatives and positives (or at least I am). The values in the above chart are also from October through March  (6 months) and are monthly values.

My interpretation of the above chart is that there may well be a slightly higher connection with NAO values in the fall season (OCT-NOV) as a predictor to what happens with those values the following winter. The 3 years that were FLIP (fall season goes one way while winter flips the other) years which goes against this were ’03-’04, ’04-’05 and ’10-’11.

Meanwhile the other oscillation that we follow is the Arctic Oscillation (AO). When surface pressures are low in the polar region…the AO is positive. This typically means a faster west to east flow of air through the USA keeping the coldest air locked in the polar regions. The opposite is an AO value of negative. Higher surface pressures in the polar regions mean weaker west to east winds in the middle latitudes…and the better ability of colder weather to move into the mid latitudes.

So now lets look at the AO values over the last 10 fall/winter years…again the same premise as the NAO chart. Left of the middle line is AO- (meaning a cold bias in winter) and to the right is a AO+ (meaning a warm bias in winter).

ScreenHunter_12 Nov. 16 14.10

It’s interesting to note that the fall/winter trends of the NAO and AO seemed to mirror each other and the flip years were the same. Not perfect but these index values have a tendency to go lockstep with each other over the course of a month.

What do I take out of this…when that AO value is strongly positive odds favor less winter snow in KC (at least for the last 10 years). Not exactly breaking ground though with that statement.

The one thing I want all to remember is that these indices aren’t perfect. For example…we just had one of the strongest areas of high pressure build down from Canada earlier this week in the month of November yet the NAO was positive and the AO was strongly positive (opposite of what you’d think for a cold airmass entering the US.

So what is next for me…I’ll be looking at the precipitation aspect of the last 10 years. does a wetter/drier fall mean anything for the following winter. I’ll be opening up the range of dates a bit to include the month of September I think since rain/snow by it’s very nature is not as common as doing daily temperature averaging.

I’m not sure if I accomplished anything over the past 1500 words but it was an interesting exercise for me at least. Perhaps some of our readers will have a different take on things.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s


  • Patrick Trudel (@sedsinkc)

    My goodness Joe, you really outdid yourself this time! A lot of data to compile and analyze, for 10 years. My take on your 10 years’ worth of data is there’s about a coin flip’s chance of the fall temperature trend continuing through the winter. Very simplistic take, I know. Your graphs discuss the AO/NAO and their relation to the strength of the westerlies and how this affects cold air intrusion into the US. My question as a non-met is this. Is there any correlation between the value of the AO or the NAO and the amplitude of the long wave pattern in the westerlies across the US? Amplification of the westerly flow leads to significant storminess ahead of the axis of longwave troughs and aids the southward motion of cold air in the troughs and northward motion of warm air in the ridges, causing larger positive and negative temperature anomalies.

  • Joe Lauria

    Patrick: Yes to your simplistic take. To answer your question Ii would think that IF you had strongly -NAO and -AO values that with the cold air drainage into states the temperature contrasts would initially be very large and that would promote stronger “surface storminess” if you will. The combination typically does indeed lead to more amplification/blocking patterns jl