PRELIMINARY Winter outlook 2016-17 

Posted: 10:20 am Thursday, October 27th, 2016

By Kirk Mellish

Over the past two to three decades Atlanta has had about 7 cooler than normal winters and 15 warmer than normal winters. Last winter was 20% or more warmer than average. What about this winter?

In case you’re new to the concept: I construct a winter outlook based on past weather history, using the analog method. An assessment of current global ocean temperature patterns, recent hurricane activity, and atmospheric signals from a variety of indices like the AO/NAO, PNA, QBO etc. (which you can Google if interested) is made, and then compared to the past to find closest matches to the present. A composite of what the winter turned out to be in the past is then used to project the future.

It’s important to note that long-term or persistent phenomenon like ENSO can be predicted months in advance, most of the other factors which have great influence on what a season will be can only be predicted a few weeks in advance. Hence they are a wild card. See below chart for how the Pacific Ocean influence (ENSO) is modulated by the more unpredictable Sudden Stratospheric Warming events (SSW) and state of the AO/NAO:


It should also be noted that there are never exact matches and so the future weather will never be an exact match to all past years that were similar. As they say in investing, past performance is no guarantee of future results.

Forecasters expect a weak La Niña to continue/develop for this winter, but some of us have reasons to think it may not behave like a typical La Niña.

Modeling of the negative or cool phase of ENSO has vacillated so much that NOAA has put out a La Nina Watch, then canceled it, then brought it back.

“Typical” La Nina winter pattern:




The “typical” ENSO neutral pattern “La Nada” when there is neither an El Nino nor a La Nina:


The cautionary flag we have for using a “weak La Nina winter” template to forecast the coming winter is the lack of past precedent for a weak or neutral Pacific hydrothermal evolution in 68 years of records compared to this year. We are coming off a strong El Nino and in the past every one has yielded a moderate or strong La Nina, yet that is not what modeling shows ending this year and starting 2017.

After the record strong El Niño of 1997-98, La Niña almost immediately set in the following summer and reached moderate-to-strong intensity before finally ending in spring 2001.

A similar thing happened following the strong El Niño of 1972-73.

However, neutral conditions followed three other strong El Niños of 1982-83, 1965-66 and 1957-58.

We have been in a warm or positive +PDO since 2014. A combo of La Nina with a +PDO is uncommon. An Example was 1983-84. The neutral to weak La Nina expected this winter follows a very strong El Nino, and a multi-year El Nino. This factors into choosing analgo years.

While we have a cooling in the ENSO region of the Pacific the big warm blob persists in the North Pacific:


The red box in the SST chart below shows the La Nina region:


The warm SST pattern in the NP like this one was observed in the winters of 1916-17, 1917-18, 1976-77, 1993-94, 2013-14, 2014-15. When you factor in the expected -AO (arctic oscillation) it argues for NW flow aloft.

A foot to 15 inches of rain is needed to get us out of the drought, which is hard to do in the Southeast in a La Nina winter. Atlanta will have to hope the La Nina is very weak or that we have La Nada, that is neutral conditions, which trend toward more normal winter rain.

La Niñas, especially the stronger ones tend to feature an area of high pressure over the Southeast United States which can pump mild air into the SE and Mid-Atlantic. The ridge resists the fridge so to speak. So a La Nina and the warmer than normal western Atlantic Ocean both argue for a SE ridge, which is not a winter lovers friend. But because this La Niña is forecast to be weak, it may allow frigid air in the Northern Plains to plunge southeastward more frequently than the average La Nina due to the other factors on the playing field.

The quiet sunspot cycle and West QBO phase combined with a weak La Nina and a positive to neutral PDO phase argue for a more suppressed SE ridge this winter.

PDO and El Nino/La Nina in the Pacific and relationship to US Temperature patterns in winter:



Higher (+) and lower (-) pressure aloft in West PDO phase years near solar minimum, the line between grey and red roughly shows the primary jet stream path:




Because of the oddities of the ocean temp patterns and atmospheric signals this year it appears anyone using the standard La Nina template for the coming winter in North America may be asking for a proverbial forecast black eye. However, NOAA/NWS clearly disagrees.

For now the best match to the expected North Pacific Ocean temperature pattern is the winter of 1995-1996, which was cold and unusually stormy in the eastern U.S. It is the top analog year with its weak La Nina warm north and east Pacific and low solar cycle with a positive westerly QBO combination.

A weak La Niña is the primary forecast driver, but there are other factors this winter that could favor increased jet stream ‘blocking’ and resulting more cold air transport into the middle latitudes of North America than in a typical La Nina. La Ninas tend to be drier in the Southeast U.S.

An anemic La Niña increases the importance of other winter weather influences, including the sun. We are clearly headed into a ‘down cycle’ with respect to the current Sun activity and this often favors colder and stormier than normal winters over the northern hemisphere.

The problem with using the solar influence is that it occurs over multiple years so nailing the peak winter or string of winters can be problematic as it is not a strait line. We’ve been in a low solar cycle (low sun spots counts) since about 2007 with the last peak in sunspot numbers in 2014 and the low cycle has a ways to go. Last year was a down solar year but cold weather and snow did not exactly rock the world. In other words you can have warm summers and winters even in a low sun output regime.

We are also watching the impressive north Pacific warm SST anomaly and a cooling of the Indian Ocean and its impact on MJO activity through the winter, with the warm water ring in the Pacific and the warm western Atlantic as dominant factors, along with signals for jet stream blocking at higher latitudes.

It does not look like a cold and hold winter, but with shots that come and go with extended thaws in-between. Still with last winter being 20% or more warmer than usual this one should not be that kind of blowtorch.

The ocean SST patterns favor a negative –EPO for winter this year, meaning high pressure aloft over the top of low pressure aloft off the west coast. Downwind this allows for troughing over the East third of the country especially NE.

These are the key teleconnection locations and how the current state of the oceans correlates to winter temperatures based on past history:



Given the warm and cool ocean pool configuration over the northern hemisphere it looks like about 3 suggest cool weather in the Southeast this winter, 1 warm, and 3 normal. So you can see at least by that measure it’s a close call.

Given the ongoing drought in Georgia a solid La Nina is the last thing we want for winter, as that is a signal for very warm and very dry for the winter. That could set us up for water restrictions next spring or summer and lower lake levels.

However, a weak or neutral ENSO would provide closer to normal rain this winter. On the other hand, there is a tendency for more rain and severe weather and stormy episodes in La Nina springs. So there is potential for both a positive and negative impact in the long range.


Dr. Judah Cohen has shown a correlation between Siberia/Eurasia snow cover in October and modeled it to make a winter temperature outlook. Snow cover has been expanding above average in the key regions this year at a rapid pace:



This is a schematic of the process described above and how it works, steps 1-6:



A real world example is below from model projection for start of November, such SW events often precede a change to colder by 2-3 few weeks:


For the coming winter, most models are currently warmer than my outlook, they will update next month:



The analogs from the past show plenty of years that were warm in the south and plenty that were cold, so there is plenty of uncertainty in using the composite.

Preliminary list of analog years used to make my long-range outlook for the coming winter:

1950-51,1952-53, 1958-59, 1960-61,1962-63, 1981-82, 1983-84, 1984-85, 1985-86, 1988-89, 1993-94, 1995-96, 1996-97, 1998-99, 2000-01, 2005-06, 2013-14, 2014-15. Also in the mix are 1959-60, 1961-62, 1966-67, 1967-68, 1971-72, 1968-69, 2002-03, 2003-04, and 2008-09.

The preliminary analog results for coming winter temperature and precipitation:



The preliminary winter outlook then is for slightly cooler than normal conditions, with near normal to a bit below normal precipitation here in the Southeastern U.S. This is the 90-day mean of the winter, not every day or every week.

So it certainly does not look like a rough winter in Georgia, certainly not on average. But not as warm as last winter.

Snow and ice prospects are always the most difficult to determine as it requires moisture and cold air to have perfect timing. That’s hard to do in a part of the country where significant snow is RARE. The average from the analog set includes some years with high snow and many years with a trace or zero.

The average snowfall from the preliminary analog list was 3 inches, 1 above the Atlanta normal. Screening of other La Nina year signals yielded 1.25 inches of snow.

It’s worth noting that there were healthy signals for snow last winter yet despite a handful of close calls, outside of the mountains and some northern suburbs there was little.

If the La Nina turns out stronger than forecast then our winter will be even warmer and drier than the current outlook. The final winter forecast will be issued at the end of November.

On the light side, in the end, perhaps the map below is the best winter forecast ANY year: