Today’s post outlines the drivers at work behind the kind of winter weather we might expect this year and explores 11 different outlooks for this upcoming winter, but before getting into the winter outlooks shown here, a word of caution on these long range outlooks is needed:
Winter is a notoriously tough season to forecast. Long range climate forecasting and modeling still has a long way to go. Currently the best numerical weather prediction models exhibit around a 90% skill (accuracy) rate to 10 days and 80-90 rate at 16 days. Beyond that, accuracy simply becomes nothing more than random noise. Running these types of models out past these thresholds quickly devolves into chaos theory which is not useful for long range or seasonal weather forecasting. This is a big part of why any forecaster predicting a specific event such as a snowstorm or ice storm at a specific date three weeks a month or even two months in the future does so at their own peril. It is also why predicting several damaging ice storms or a specific number of snowstorms simply cannot be done. Pure chance or a roll of the of the dice has an equal opportunity of giving you an accurate number of winter snow or ice storms per any given winter season.
Long range models such as the NWS Climate Forecast System’s Version 2 model, (CFSv2), the European Center for Medium Range Forecasting (ECMWF) model, the IRI and others shown here are not the same models used in forecasting near term weather. They operate on a different set of physics equations and programming.
An added challenge in forecasting snowfall totals lies in the location of our region. Average annual snow ranges from as little as 6″ south to as much as 20″ north. These totals are well within the range of one or two powerful winter storms. A case in point was the winter of 2012-13. That winter was warmer than average and on the whole (three month period) not very snowy. One late December snowstorm dumped an entire season’s worth of snow and then some over SE Missouri and S Illinois. That one storm, the only major storm, ensured those areas went down in the record books as a snowier than average winter. Kansas City and St. Louis fell victim to two to three major snows in February and March which dumped an entire season’s worth of snow and in the case of Kansas City, an entire season’s worth plus 10″.
The point being that even with a warm and dry winter prediction, it can only take the right combination of factors to come together once or twice in that multi-month period to bring an entire year’s worth of snow. That’s not the same as for areas further north where average snowfall is 2 to 4 feet plus, well out of the range of almost all individual winter storms. Below is a graphic showing the average annual snowfall total for the greater Midwest based upon the 1981-2010 climate period.
Average annual snowfall across the region. Snow totals as little as 6″ for the Missouri Bootheel and extreme southeastern Illinois, with around 9″ from Joplin through West Plains Missouri, Chester and Mt. Vernon, Illinois. Average annual snowfall of a foot is typical from Springfield, to the Lake of the Ozarks and Hillsboro, Missouri across into Belleville and Vandalia, Illinois. Kansas City, St. Louis and on into central Illinois. Northern Missouri and north central Illinois typically see totals from a foot and a half to two feet with totals steadily increasing to the 2-3 foot range further north,
In this post, we will take a look at some of the factors we can look to for reliable clues for the winter ahead, and then take a look at several differing winter ideas.
For cold air to develop, you need darkness, ice and snow. Darkness at the north pole is always a given in winter and nothing will change that, so we’ll look first to the extent of the ice cap heading into Winter. The chart below shows the extent of Arctic sea ice. After a record low in 2012, 2013 is running well ahead of that year, but still on the low side of average.
This graphic shows the extent of Arctic sea ice. The grey zone can be thought of as an average range while the grey line is the 30 year average. 2012 saw a record minimum sea ice extent while 2013 has had much more ice than 2012, within the average range but below the 30-year average.
Once cold leaves the Arctic, in order to stay cold, it needs snow and ice on the ground. Otherwise it begins to modify or warm up as it heads south. The graphic below shows the mid October extent of snow and ice for this year as well as 2012, 2011 and 2010 for comparison. This will become more important later in the season; late November, December and January, but at least we can compare where we are starting from.
This graphic shows the mid October snow and ice cover from this year (2013) back through 2012 (before a mild winter early and cold and snowy winter late), 2011 (before a very warm and dry winter) and 2010 (before a cold and snowy winter). The amount of snow and ice over the high latitudes are important as they act to refrigerate the cold air building at this time of year in the Arctic. If there is a lack of snow and ice at high latitudes, the cold air warms as it heads south, lessening the intensity of the cold. One thing that is watched carefully is how far snow cover extends into the U.S. as even several hundred miles of bare ground can take the bite away from an Arctic air mass. These maps become more important later in the season as we move into December and January when the strongest Arctic outbreaks will occur, at this early remove it just gives us an idea of our starting point.
Perhaps one of the most important factors outside of the cold and snow in the Arctic is the state of the ocean, and specifically how warm or cool it is and where those warm or cool areas are. The graphic below shows the sea surface temperature departure – or difference – from average in mid September of 2013, 2012 and 2011. A quick glance shows how different the ocean temperature is this year from that in 2012 and 2011. The location of the warm and cold pools is important as it plays into the strength and location of warm upper highs & ridges, as well as cold upper lows and troughs. Cold polar air contrasting with warm equatorial waters (El Nino) creates a strong temperature contrast which can create a stronger jet stream. That can result in a jet stream that remains strong and pulled back to the north, while cold polar air has less of a contrast with cooler equatorial waters (La Nina) for a weaker, and waiver jet stream.
Another important factor in winter weather is the ocean. Covering 70% of the earth’s surface, it has one of the greatest influences on the weather. This graphic shows mid September sea surface temperature anomalies, which basically mean how warmer or colder the ocean’s surface is compared to average. When comparing 2011 to 2013, notice how the Pacific north and south of the equator is cool in 2011 and much warmer in 2013. The north Atlantic was cooler in 2011 and is warmer in 2013. Warm water favors blocking patterns in the jet stream. Warmth in the north Atlantic favors a Greenland Block, and warmth in the north Pacific favors a ridge in the eastern Pacific. Along the equator, the cool blue tones in 2011 represent a La Nina weather pattern, while there is no such organized pattern present in 2012 or in 2013.
At this point an El Nino or La Nina is not expected this winter so they will not be the overriding drivers of the winter weather. Whether a warm or cool neutral trend develops and where it develops may be a contributing factor.
A major player in winter weather is always El Nino or La Nina. This year conditions are forecast to be neutral during the winter months (red box) One group of models favors neutral warm while another neutral cool. This means El Nino or La Nina are not expected to be a dominant factor in the winter forecast.
As we have seen in winters past, the Atmospheric Oscillations (Arctic-AO, North Atlantic-NAO, Pacific North American-PNA) can play a really large role in the prevailing winter weather pattern. Note: These are notoriously hard to predict more than a few weeks ahead. In the past three winters, they’ve tended to remain stable for long periods each season.
Atmospheric Oscillations (Arctic-AO, North Atlantic-NAO, Pacific North American-PNA) can play huge roles in prevailing weather. This graphic shows one computer model’s suggestion for these three oscillations. Note: These are notoriously hard to predict more than a few weeks ahead. In the past three winters, they’ve tended to remain stable for long periods each year. We have two essential votes for colder conditions(NAO/AO) and one vote for warmer (PNA).
Winter outlooks for 2013-2014 as of Mid October.
The first outlook is the U.S. National Weather Service’s Climate Prediction Center (CPC) outlook. This outlook focuses the greatest probability of warmth over an area from Louisiana to Arizona and from the Gulf and U.S./Mexico border north to the AZ/UT-NM/CO-KS/OK borders and then southeastwards through SW Arkansas and into Louisiana. A second warm area is forecast over northern New England. An increased probability of dryness is expected over an area centered over far west Texas and New Mexico and as far west as SE Arizona and as far east as the middle of Texas. A second area of dryness is placed over the southeastern states including Florida. The remainder of the country has equal chances for warmer/cooler or wetter/drier conditions. Signals are not strong enough in any way to tilt the odds one way or another. The warmth has been scaled back from the September outlook. In September, the local area was included in the area given a greater chance for above average temperatures. Several other models used in making these predictions have switched over to cold signals in the past month for our area. The intensity of the shading does not indicate how much warmer and drier these areas will be, but rather how confident that the forecast of warmth and dryness will verify.
THE NWS CPC verdict: Near average or no strong signals either way.
This is the NWS Climate Prediction Center’s temperature and precipitation forecast for meteorological winter (Dec 2013-Feb 2014) which indicates no strong signals for warmer/colder or wetter/drier conditions locally. Last month’s outlook had indicated a slight tilt toward warmer than average temperatures. This has been removed in this current outlook. In their discussion issued October 17th, they indicated this was due to conflicting signals between colder weather as shown by European and international models and warm signals still forecast by U.S. and Canadian Models. These colder signals were not present last month. The CPC will update their winter outlook one more time, Thursday, November 14th.
One of the models used in the above forecast is the Climate Forecast System’s Version 2 model, (CFSv2). This model’s most recent run expects temperatures to average nearly 2 degrees F above average over the region this winter. This model has been persistently warm over the past year.
The CFSv2 verdict: Warmer than average and wetter than average.
This is the latest run of the NWS Climate Forecast System’s Version 2 model, (CFSv2). It favors above average temperatures over the region.
Here is that model’s forecast for winter precipitation. It forecasts above average precipitation.
This is the latest run of the NWS Climate Forecast System’s Version 2 model, (CFSv2). It favors above average precipitation over the region.
This model is the Japanese Meteorological Agency’s (JMA) model of the prevailing upper air pattern for winter. It forecasts a high pressure ridge over Alaska and NW Canada to off the U.S. west coast and a low pressure area over Hudson’s Bay. The colors represent upper height departures from average and not specific surface temperatures. This type of pattern would favor cold for the central and eastern U.S. and warmth out west. It does indicate an occasionally active southern jet stream which could act to bring rain and or snow to some areas. This is overall a drier pattern for the region but one in which significant snows could occur given the right setups. It would also favor an increased chance for Arctic outbreaks due to cross polar flow. The prevailing pattern would be a positive Pacific North American (+PNA) pattern.
The JMA’s verdict: Colder and drier overall but increased potential for Arctic outbreaks and snow.
This is the Japanese Meteorological Agency’s (JMA) upper air forecast for Meteorological Winter (December 2013-February 2014) . It forecasts a high pressure ridge over Alaska and NW North America and a low near Hudson’s Bay. This is a cold forecast for the central and eastern U.S. for most of the winter. A southern stream split flow is also shown which would increase the risk for cold air mass-skirting winter storms.
The following two graphics are generalized representations of a +PNA pattern:
This is a simplified classic depiction of a Positive Pacific North American (+PNA) weather pattern and some general effects. This is just a generalized description as each pattern’s configuration will be different.
This is a generic representation of weather at the surface common with a strong positive Pacific-North American (+PNA) pattern. A deep trough in the central and eastern U.S. keeps a steady flow of cold high pressure areas dropping southward into the central and eastern U.S. The plains will see wider swings in temperatures as Arctic air alternates with warm and dry downslope flow, while areas further east remain consistently colder. This is a dry pattern overall, but there are exceptions. An active subtropical jet stream can undercut the western ridge and bring storm systems through the southern tier of states producing heavy snows for the Plains and Midwest occasionally. The patterns depicted here are just general representations, as each +PNA setup is different.
A private forecasting company called AccuWeather issues a specific winter forecast each year. Here is the latest current forecast as well as the past two years. Also shown is how these verified for our local area. This blog does not endorse AccuWeather’s forecast nor dismiss it, these panels are shown for information only. AccuWeather owns the rights to their forecasts.
AccuWeather’s verdict: Snow and Ice for Kansas City with above average snowfall for the season and severe thunderstorms for St. Louis, SE Missouri and S Illinois and near average snow.
A private weather organization called AccuWeather is famous for its much anticipated winter outlooks. Shown here are the current and last two winter outlooks. Please note: this blog does not endorse or dismiss their forecasts, they are simply shown for informational purposes. For 2012-13, the forecast for above average temperatures verified, but the below average snow prediction did not. For 2011-12 neither the temperature or snowfall forecast verified.
The next winter forecast is from the International Research Institute (IRI) which runs a number of differing climate forecasting models to come up with a winter outlook.
IRI verdict: Near average or no strong signals either way.
This is the IRI (International Research Institute) multi-model forecast for winter temperatures. It calls for greatest probability of above average temperatures over northern and eastern Canada, New England (north) and the southern Plains and Mexico (excluing Baja California). Over the US, it closely mirrors the CPC outlook.
This is the IRI (International Research Institute) multi-model forecast for winter precipitation. Dryness has an increased chance over the southern Plains and Southeast. The pink areas are regions where a dry season occurs in this period.
The next outlook comes from us from Japan, the JAMSTEC (Japan Agency for Marine-Earth Science and Technology) long range climate model. Their latest forecast is a bit dated, from September 1st.
JAMSTEC verdict: Warmer than average with near average precipitation.
This is the JAMSTEC (Japan Agency for Marine-Earth Science and Technology ) temperature outlook for winter issued September 1st (the latest available) It forecasts a warm winter for most of the continental U.S. except for the western Great Lakes.
This is the September (latest available) JAMSTEC (Japan Agency for Marine-Earth Science and Technology ) forecast for winter precipitation departures. Near average precipitation is expected for our region with dryness southwest in SE Kansas and slightly more moist conditions northeast in the Great Lakes,
Here are some additional models and what they expect:
NMME (National Multi-Model Ensemble) verdict: Warmer than average temperatures and no strong signal for precipitation.
The NMME (National Multi-Model Ensemble) forecasts an enhanced probability of above average temperatures and no strong signal for precipitation.
NCAR (The National Center for Atmospheric Research) verdict: Cooler than average west and warmer than average east. No precipitation forecast is available.
NCAR (The National Center for Atmospheric Research) expects cooler than average weather west and warmer than average weather east.
The Canadian Meteorological Center’s verdict: Warmer and drier than average.
The Canadian Meteorological Center’s forecast results in warmer and drier than average conditions.
The US Navy model’s verdict: Slightly warmer and slightly drier than average.
The U.S. Navy’s GFDL model expects slightly above average temperatures and slightly below average precipitation.
A private forecasting company called WeatherBell issues a winter forecast each year. Forecasts are based upon a comprehensive detailed meteorological analysis of many different global models, sea surface temperature, temperature of the stratosphere and solar cycles. Here is the latest current forecast. Over the past few years, Weatherbell’s forecasts have verified quite well for our area. This blog does not endorse Weather Bell’s forecast nor dismiss it, these panels are shown for information only. Weatherbell owns the rights to their forecasts.
WeatherBell’s verdict: Colder and snowier than average.
WeatherBell, an independent private forecasting agency forecasts below average temperatures by 1-2 degrees C over the region and 133% of average snowfall. That would amount to roughly 25″ in Kansas City, 23″ in St. Louis and 7-15″ for SE Missouri and Southern Illinois. Note: WeatherBell owns the right to its winter forecast and this graphic neither supports nor dismisses its claim.
Here is a simplified scorecard of all the differing forecasts presented here. Cold and snowy is in first place with three votes, followed by warm with 2 votes, then near average and warm and dry with 2 votes. warm and wet and cold round out the last two categories. NCAR is split in to halves, one cold vote for the west and one warm vote for the east.
Winter outlooks will be updated; and this post will be updated in mid to late November.