Tuesday, October 25, 2016


 Follow on Twitter                       Like on Facebook

It's that time of year again, the leaves are changing, the presence of hoodies is on the rise and the burning question on everybody's mind seems to be, "What should we expect this winter?" Long range forecasts are used by businesses to manage risk or manage supply of weather-demand products, and they are becoming increasingly important. Some people live for the snow and cold, and some could do without. No matter what side of the spectrum you fall, you likely have an interest of what to expect this winter. Alright, let us dive into the Midwest Weather Winter Forecast for 2016-2017! Full forecast at the bottom of the page.

2015-2016 Recap:

The story of last winter was the extensive warmth across the entire country. This was, in large part thanks to a strong El Nino in the Central Pacific. A strong El Nino is described as a warming episode in the Equatorial regions of the Pacific. In fact, last year was a top 3, possibly top, warming event since 1950. It had a profound effect on temperatures in the Midwest and last season was no exception. Overall, the warmer temperatures favored below normal snowfall, yet some places were caught under a few storm tracks and saw above normal snowfall. Here is a link to how the forecast panned out last year: The Midwest Weather 2015-2016 Winter Forecast

2016-2017 Discussion:

In most winter forecasts the state of ENSO, or El Nino/La Nina is examined and much of the forecast is based on this index. However, this year we are sitting at a neutral state to eventual weak La Nina and are forced to look elsewhere for more clues. Contrary to last year, a lot of the weight was given to the El Nino, this will not exactly be the case this time around. Putting too much emphasis on the projected weak La Nina may prove to be a flaw in other published winter forecasts. We will examine the state of other regions in the Pacific and Atlantic Ocean to help create analog years, or years with similar conditions to use as a proxy for the 2016-2017 season. This process has been proven successful in the past several years for Midwest Weather. And because every year is different, we will look to pick out the major differences and potential factors that will have the largest impact on the upcoming winter. 


The state of ENSO has been hovering around -0.5 since the summer and most models continue to hold steady at this rate. It implies a neutral to weak La Nina, or slightly below normal ocean temperatures in the Central Pacific Ocean. Usually in the transition to a La Nina, up-welling will occur in the eastern Pacific, bringing colder water to the surface. This cooling process has yet to establish itself as the water temperatures in the mid to upper level of the ocean have stayed warm. Confidence is high that we will not turn to a moderate La Nina anytime soon. For our case, we will examine years in which the ENSO index was within 0.0 to -1.0. Below are the top 20 winters that best resemble a 0.5 ENSO value for December through February; their results on temperature are plotted below: 

Pacific Decadal Oscillation (PDO):

The Pacific Decadal Oscillation (PDO) is a pattern of Pacific climate variability similar to ENSO in character, but which varies over a much longer time scale. The PDO can remain in the same phase for 20 to 30 years, while ENSO cycles typically only last 6 to 18 months. The PDO, like ENSO, consists of a warm and cool phase which alters upper level atmospheric winds. Currently we are in the positive phase of the pattern and should last through the winter, its effect on the overall weather pattern is a major factor.

The positive phase of the PDO is described as warmer than average water in the Gulf of Alaska down through the Pacific Coast of the United States. The subsequent result of this warm water is a cold blob of water further off to the south and west, by virtue of ocean currents and areas of up-welling. The plot from Unisys shows the SST anomalies resembles these exact conditions. 

The figure below shows how positive PDO winter have influenced temperature anomalies over the period from 1951-2010. 

A bit contradictory to the state of ENSO right? This is why we will continue to examine other factors around the globe for the sake of our winter forecast.

Atlantic Multi-Decadal Oscillation (AMO):

The Atlantic Multi-decadal Oscillation (AMO) is a mode of natural variability occurring in the North Atlantic Ocean and which has its principle expression in the sea surface temperature (SST) field. The AMO is identified as a pattern of variability in North Atlantic SSTs. We are in the positive phase of the AMO as well. However, we may be making a turn towards the negative phase. The SST anomalies plot from Unisys plot above also shows some pooling of colder water, which is part is due to sea ice melt adding fresh water to the ocean , locally changing the water density is some spots. Given these factors and since the Atlantic is downstream to the United States, the AMO will  be less of a factor on the winter forecast than it has in other years. The winters closely related to the positive phase of the AMO:

Quasi-Biennial Oscillation (QBO):

The quasi-biennial oscillation (QBO) is an oscillation of the equatorial zonal wind between easterlies and westerlies in the tropical stratosphere with a mean period of 28 to 29 months. The alternating wind develops at the top of the lower stratosphere and propagates downwards at about 1 km (0.6 mi) per month until they are dissipated at the tropical tropopause. This wind pattern has an influence on the weather patterns in North America. Coincidentally, we are also in the positive phase of the QBO as well. The winters that best resemble the current state of the QBO are averaged below:

Siberian Snow Cover:

Dr. Judah L. Cohen has done research on the state of the snow pack in Siberia (His research). He has shown a correlation in snow cover to the level of cold air intrusion to North America. When October snow cover is above average, cold air has a better chance of manifesting itself into the Midwest. Below are the top closest years that best resemble the current state of snow cover across Eurasia: 

Putting it all Together: The Best Analogs

Clearly each factor in the forecast gives a different result on temperatures and the subsequent winter forecast. A way to resolve this is to use analog years, using winters that best take into account each factor. The process was to first find years that best held these conditions:

ENSO: 0 to -1 | PDO: Pos (+) | AMO: Pos (+) | QBO: Pos (+) | Sib Snow: Above Ave

Below are the best analogs based on all factors and are given weights that best describes how related each year is to the upcoming winter:

Here is the result on temperature for December through February:

And the result on precipitation: 

And the result on geopotential height at 500mb and surface pressure:

The analogs suggest the midwest is at an increased chance for below normal temperatures, near to possibly above average snow and an increased risk for large winter storms. Yet every year is different, so let's continue...

A Look at the Models:

Models are starting to trend towards a cooler solution for the winter, which coincides with our analogs, but still suggest some spots could still be at or slightly above average. The CFS model, which is a low resolution model that run 4 times a day out to a year, also has an increased chance for slightly above normal precipitation in parts of the Midwest.

State of the Arctic Ocean:

A more organic method in forecasting is to directly look at the current conditions. For a North American winter, the Arctic Ocean plays a major role in how cold the winter can get, and it's actually counter intuitive. 

The extent of Arctic sea touched a record low this week. Furthermore, the ocean water temperatures are well above average and approaching records as well. These two factors have implications on the polar vortex, yes the polar vortex. The strength/magnitude of the vortex, or jet stream around the Arctic Circle depends the temperature gradient, which is the change in temperature from the north pole to areas further south. The colder the North Pole is, the largest the temperature gradient and the stronger the polar jet gets. A warmer north pole leads to a weaker polar vortex by the opposite logic. Consequently,  strong polar vortex spins fast and locks in the colder air for areas further north, while a weak polar vortex spins slower with more wobbles, allowing colder air to reach areas further south, including the Midwest. Think of it as a top spinning, as it slows there is more wobbles, just like the polar vortex. Each one of these wobbles has the potential to bring colder air to the mid-latitudes. In summary, conditions favor better chances for more shots of cold air into the United States than prior years. And it's all thanks to warm Arctic Ocean temperatures! Isn't weather strange sometimes? Here is a look at sea surface temperature departures from average:


Overall, the forecast suggests a much cooler winter than last year with the potential of more snow for many. Expect several cold spells, yet each arctic blast is not expected to last long. There should be more variability in temperatures from week to week than average, setting up a higher potential for large winter storms to develop. Warm waters (5-6 degrees above average) will prolong and enhance the lake effect snow season. There will also be an increase in Alberta Clipper-like systems that drop a quick 1-3"/2-4" snowfall during cooler periods this winter. Spring may come earlier than usual, even with cooler temperatures through the winter. While only Mother Nature will have the final say, it looks to be a good 'ol fashion winter for the Midwest.

Bachelors in Atmospheric & Oceanic Science
University of Wisconsin-Madison

Thank you for reading. You can find more by following on Twitter and on Facebook.

Here is the final forecast:

**Be sure to keep it here to Midwest Weather for updates throughout the winter!


- National Center for Atmospheric Research
- Earth System Research Laboratory
- Zack Labe for his arctic sea ice plot
- Tropical Tidbits for CFS plots
- Midwest Regional Climate Center
- Unysis 
- Climate Prediction Center