October Winter Forecast Update

Decaying El Nino pattern with time. We're going into our La Nada winter.

Quick and to the point for people who are interested in just the short version: I am expecting this winter to be 1)slightly below normal temperatures (Appalachian Mountains and west of the Appalachian Mountains), with the remainder of the Southeast being at near normal temperatures 2)slightly below normal precipitation for the Southeast, with above normal precipitation west of the Appalachian Mountains 3)Above normal snowfall for the Ohio River Valley and the Northeast, near normal snowfall for the Southeast.

Hang on, doesn't that conflict with just about everyone else forecast thus far? Yes. However, this is a tricky scenario unfolding for this winter and everyone has their own interpretation of what is going to happen over the next several months. With that being said, lets jump into the heart of the forecasting!

My top 5 most similar winters used were : 1951-1952, 1952-1953, 1966-1967, 1978-1978, and 2008-2009.

The JAMSTEC model (thanks Brad for the link) projection for this winter (temperature anomalies) for December - February. Color scale is from -1.5*C to 1.5*C.

L'Heureux's Weather temperature anomalies (Dec - February) using the top 5 winters method.

The conflict between my output and the models output is with northern Alaska and northwest Canada. For those regions, my output shows this as anomalously cold while the model output shows it as "warm." But this is the only conflict I have run into (and my output is only 1*C off from the model output).

October model projections for precipitation anomalies for this winter. Blue = positive anomalies, red = negative.

L'Heureux's Weather precipitation anomalies. Red = drier, blue = wetter.

I actually did a double take when this popped up. But the more I thought about it, the more sense it made. We have a retreating summertime El Nino pattern and will be entering into a "La Nada" winter (neither warm or cold water anomalies). An index heavily used in my forecast, the Multivariate ENSO Index (MEI) uses more factors than just sea surface temperature anomalies alone.  From the MEI site "These six variables are: sea-level pressure (P), zonal (U) and meridional (V) components of the surface wind, sea surface temperature (S), surface air temperature (A), and total cloudiness fraction of the sky (C)". This in my opinion makes it a more accurate estimation of true ENSO conditions rather than just the water temperatures alone. But for the sake of consistency with other forecast methods, I have factored the water temperature anomalies too.

So in one of these La Nada winters, both the subtropical jet and the polar jet stream have their periods of dominance. You will see that very clearly in the storm track animation below. Lets dive into our drivers first.

Northern Annular Mode (the Arctic Oscillation) impacts on the weather patterns.

The Arctic Oscillation is a measure of the strength of the polar vortex; it does this by measuring pressure anomalies in 1)the mid-latitudes 2)the polar regions. For simplicity purposes, I will break down the polar vortex to something we can all relate to: a coffee cup.

Stir a cup of coffee. For this test to work, we need some coffee grinds in our coffee cup. These coffee grinds represent our cold air reservoir. Now, spin your cup in a counter clockwise direction. Notice how when the circulation grows stronger, the grinds/cold air becomes more concentrated under the center. Imagine the coffee itself is a representation of "pressure height anomalies." The center is lower and the edge is higher. This would be an example of a positive phase of the Arctic Oscillation (my red box above). The Arctic has anomalously low surface pressure and the mid-latitudes have anomalously high pressure. Now, put a spoon in your coffee and block the circulation. The circulation will distort and the coffee grinds will spill out from the center. The circulation may even break in half. Now the center is not as low and the edge is not as high. The distorted flow would be something like a negative Arctic Oscillation.

In real life, when the polar vortex is very circular and zonal (west to east), colder air is bottled up in the Arctic. But when warmer air from the mid-latitudes is driven into the Arctic (as a ridge), it can distort the circulation. If it distorts enough, cooler air from the Arctic is pushed into the mid-latitude regions instead. A zonal flow needs cooler air to the north and warmer air to the south to create its west to east flow (thermal wind). The stronger the contrast, the stronger the west to east flow. But if one were to weaken that contrast, the zonal wind flow weakens with it. February 2009, February 2010, and December 2010 were notorious for the flow breaking down entirely and reversing in the Arctic stratosphere, because the stratosphere over the Arctic was warmer than the stratosphere in the mid-latitudes.

The changes of the storm track and pressure patterns between the positive phase of the NAO and the negative phase of the NAO.

The North Atlantic Oscillation is a branch of the Arctic Oscillation but is important enough on its own to mention. A positive phase has a stronger than normal low pressure near Iceland and a stronger than normal high pressure over the Azores. A negative phase has weaker pressures over both.

The PNA is a little trickier but think of it as an equivalent to what the NAO does for the east coast, over the west coast instead. In a positive phase, there is a ridge over the northwest, a trough south of Alaska, and a trough over the Southeast. This is another helping hand in distorting the Polar Vortex. There is a theory that the PNA is a bridge between ENSO and the Arctic. The literature is still a little muddled and I will not include much here, except that in an El Nino winter the PNA tends to lean positive.

Impacts of the NAO and PNA on the storm track

Watch the wind animation below. Notice that when the height anomalies reverse over the Atlantic, the wind arrows become shorter as well (late December and again in late January/February) Also notice that when the Arctic Oscillation is in a negative phase (lots of red over the Arctic) the storm track goes a lot farther south too, and the zonal flow into the northern United States is weaker. This is good for snowfall across the east coast.

The flagship animation. The 500mb geopotential height anomalies with the storm track on a daily time scale (top 5 winters). The plots are smoothed out via use of a 5-day running mean.

The projections for the Arctic Oscillation this winter; the blue bars represent the top 5 most similar winters and the red bars represent the top 10 most similar winters (bottom North Atlantic Oscillation). The boxes around the graphs represent clear warm phases and cold phases. Lighter colors are equal to the minor warm/cold phases.

500mb wind speed, December 1st - March 31st

My wind animation (above), using the 5-day running mean for the top 5 winters selected.

500mb geopotential height anomalies, December 1st - March 31st

So again, quick and to the point, I am expecting this winter to be 1)slightly below normal temperatures (Appalachian Mountains and west of the Appalachian Mountains), with the remainder of the Southeast being at near normal temperatures 2)slightly below normal precipitation for the Southeast, with above normal precipitation west of the Appalachian Mountains 3)Above normal snowfall for the Ohio River Valley and the Northeast, near normal snowfall for the Southeast. Northwest Flow will play a larger role for the Appalachian Mountains this winter while the rest of the Southeast may have its snow opportunity in 1)Late December 2)February. The battle zone certainly seems to be in the Ohio River Valley and the northern Mid-Atlantic this year.

The Pacific Northwest stays chilly but slightly dry for the winter. The northern plains I expect to have drier conditions for the winter, while the Rockies get a time share deal from a boosted subtropical jet and a polar jet, both of which are competing for the #1 slot. I would wait for those + AO periods for precipitation for the northern Rockies and a -AO period for precipitation in the southern Rockies.

The Deep South will be robbed of moisture until those -AO periods due to the battle zone being farther north (Midwest, Ohio River Valley). My concern for these areas and with the Southeast comes from increased ice storm chances late December (primary risk Midwest, secondary Mid-Atlantic) and February (primary risk Mid-Atlantic, secondary risk Midwest). There was not much amplitude with the trough in respect to the wavelength itself, and weak low pressure systems over the Midwest with a cold air component OR a two-tier low (primary weakens west of Appalachian Mountains, secondary forms along east coast) to them are a good indicator for mixed precipitation.