Today’s post outlines the drivers at work behind the kind of winter weather we might expect this year and explores 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. Currently the best numerical weather prediction models that are run several times daily (GFS, GFSx, ECMWF, NAM, CMC) 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, 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 a specific number of damaging ice storms or a specific number of snowstorms simply cannot be done.
An added challenge in forecasting snowfall totals lies in the location of our region. Eastern Kansas, Missouri and southern Illinois lie south of the main snow belts and parts of southeast Missouri and southern Illinois border areas where snow events actually become uncommon. Average annual snow ranges from as little as 6″ south of the district SSW of Cape Girardeau to as much as 24″ west and north. These totals are well within the range of one or two powerful winter storms.
That brings up another difficulty in winter weather prediction. With yearly snowfall totals so low, one or two exceptionally powerful winter storms can literally bring an entire season’s worth of snow. Such was the case with SE Missouri and SW Illinois’ early December storm last year. That means that the winter can be “warm and dry” overall with just one event possibly resulting in above average snow.
What a “typical” winter looks like
Let’s start with the averages. Meteorological winter is considered to run from December 1st to the last day in February. Snowfall is tallied from July 1 of the current year to June 30 of the following year. This captures any early fall snows and late spring snows. The graphic below shows the winter weather averages for KC and STL (click to enlarge). Being further west and north, Kansas City is colder and much drier than St. Louis, as moist southerly flow tends to turn east before reaching the KC area and southern stream weather systems tend to pass south of the KC area. Despite this, KC has just slightly more snow than STL, although just by an inch or so. Snowfall peaks in January in STL while there is a peak in February for KC. Despite being on the edge of the Central Plains, heavy snow is not at all common in Kanas City, there are on average (1888-2013) 15 measurable snow events per season. 88% of the events are light snow events (4″or less). Only 8% produce storm totals of 4″ or greater and less than 4% produce storm total snow considered “major” of 6″ or greater. Since 1888, there have only been 17 events where the storm total snow has been greater than 10″.
The coldest temperatures ever recorded in KC and STL were -23 and -22 respectively, set in the frigid December cold outbreak in 1989 in KC and 130 years ago in STL.
For eastern Kansas, Missouri and Illinois, average snowfall ranges from as little as less than 6” in Missouri’s Bootheel to as much as 42” along the shores of Lake Michigan. Topeka, Kansas City and STL all lie within a belt where average winter snows run between 12 and 18”. Click on the graphic to make it more readable.
6-12” is common for SE MO/SW and S IL. In this area, locations from Annapolis MO to Rockwood and Du Quoin Illinois, around 9” is typical with 9-12” north of that line and 6-9” to the south. Click on the graphic to make it more readable. This low annual average total puts this area, in particular, in a zone where it is possible to get a season’s worth of snow in one storm. That was the case in the December 2013 storm when 10-16” fell in one event. The higher annual totals of STL/KC mean that a single storm strong enough to dump a season’s worth of snow is less likely.
Before we look forward, let’s look back at the last few seasons. Some clues to this year lie in those cold seasons.
2013-2014: Cold with above average snow
Looking back at 2013/2014: 68% of all days of the winter were below average in Kansas City and 65% were below average in St. Louis. The winter season came in at 5.41 degrees colder than average in KC and 4.7 degrees colder than average in STL. 37 days never made it above freezing in KC and 32 days in STL. KC had 11 days with temperatures at 0F or lower while STL had 5 days. -11 was the coldest temp in KC and -8 in STL. Click on the graphic to make it more readable.
In terms of snow..it was above average. +11.6” in STL and +7.3” in KC. These were not top ten snow totals. The winter snow total was around and inch more than 2012-13 in STL and about 5” less than in 2012-13 for KC. Click on the graphic to make it more readable.
Next, let’s take a look at how the snow fell across the cool season. These graphs show how much snow fell per month at KC and STL. In both images, the purple line is the actual total while the faded blue line is average. In KC, snow was below average in October & November, then slightly above average in December and January with the “biggest” snow month February. Snows continued above average in March. In St. Louis, snow was below average in November and then above average December through February. The “biggest” snow month in STL was January. Click on the graphic to make it more readable.
Finally let’s compare regional snow to the averages. Last year (2013-2014) all areas finished above average. If you look at the actual snowfall map, you can see one area of higher snows from south central MO through east central MO & SW IL. That’s the southern storm track of the big December/January storms. The higher snow area from east central KS to NW MO is the track of the systems that impacted KC/Topeka, especially the heaviest storm in February. The big snow totals from IA into N IL and into the Great Lakes is from the multiple snow producing Alberta Clippers which tracked NW to SE just northeast of the district. Southwestern Missouri, while still above average, was in zone bypassed by a majority of the northern & southern stream snowstorms. Icing was a problem in this area from the southern storm track, as it typical.
The reason for the cold and snowy winter? A lot of it had to do with ocean temperatures. A warm pool of water developed in the north east Pacific, which helped create and maintain a strong and warm ridge of high pressure in that area. That caused the jet stream to bulge north into the Arctic. The jet responded by buckling south over North America, which brought Siberian air across the pole and Arctic air south into the U.S. This was a predominantly dry and cold pattern for Kansas City with numerous very light to light snow events. There were 16 snow events of 2” or less, 2 events of 2-3” and only 1 “major” event of 7.5”. In St. Louis, there were also numerous light snow events – 15 of 2” or less, 3 events of 2-4” and one major event of 10.8”. In SEMO/SW IL major events struck like clockwork at the beginning of each month, 12-16” in early December, 8-10” in early January, 4-6” in early February and an ice/sleet event in early March. Compare this with Chicago, under the track of the Alberta Clipper snows, which had 32 snow events of 2” or less, 11 events 2-4” and 4 major snows of 5”+. 82” fell in the city through April.
2012-2013: A warm and dry start; significantly colder and snowier from February into early Spring
Going back another year to 2012-2013, the snows at both cities were well below average until February. Most snow that season fell in February and March. Snow totals were also “bigger” that season. Click on the graphic to make it more readable. That winter turned out to be snowier than average and most remember it as a “hard” winter. The actual record shows it to be very warm, nearly 4 degrees above average in November (Autumn), December and January. There were periods of cold temperatures, but these were transient and quickly reverted to warmth. February and March the pattern suddenly flipped and those two months were as much, if not more below average than the months before were above average. Because meteorological winter is Dec-Feb, we only include temperatures for those three months and winter was officially mild. The February cold was not enough to offset the very mild Dec, Jan and early Feb. A majority of the season’s snow (remember we measure snow from July 1 to June 30) fell February and into March. It should be noted that Southeast Missouri and SW IL saw a heavy wet 4-6″ snow December 28, 2012 which missed STL/KC.
These graphics show the temperature and snowfall departure (or difference) from average over the Midwest. Temperatures were 1-3F above average Dec-Feb with the warmest readings in STL/SEMO/SWIL. In terms of snowfall, areas from Columbia MO through KC and Topeka KS had 175-200% of normal snow, while western parts of the STL area had 100-150% of normal. Western SEMO was a bit below average (75-100% of average) but SEMO and SW/S IL had 125-175% of average.
2011-2012: Very warm, record low snow totals:
2011-12 was also a record year- but for the opposite reasons. Ocean temperatures made a big difference. Remember the mild and nearly snowless winter? The ocean temperatures were nearly opposite that year from those in 2013-2014 That configuration led to this type of jet stream pattern:
These graphics show the temperature and snowfall departure (or difference) from average over the Midwest. Most areas in the district were 4-5F above average. Snowfall was scarce, with only 10-50% of the average annual totals.
The warmer waters were back over the northwest Pacific. Cool water was over the eastern Pacific, the tropics, off the coast and in the North Atlantic. A low latitude ridge took shape over the subtropical Pacific and a broad trough over the north Pacific . This funneled a strong and fast moving jet stream off the Asian coast and across the Pacific. A broad, but persistent ridge was over North America. The result, Alaska was bitterly cold and the lower 48 states were flooded with mild Pacific air. A record setting lack of snow in was observed in KC and extremely low snow everywhere else. It was also a very warm winter and a very early spring. No winter storm warnings were issued all winter for the KC, STL or SEMO/SW IL areas. The overall jet stream pattern was similar to the graphic below (click to make it more readable) along with some of the stats for that winter.
2010-2011: Cold and Snow:
The Winter of 2010-2011 was a cold and snowy winter with several strong storm systems and a large blizzard which impacted areas from KC to Chicago in February. St. Louis was on the southeastern edge of that storm and SEMO/SWIL was on the warm side of the storm, with rain. Click the graphic below for more information and to make it more readable:
These graphics show the temperature and snowfall departure (or difference) from average over the Midwest. Temperatures were 3-5F below average that winter. Snowfall ranged from 10-30″, mostly above average.
Overall, with the exception of 2011-2012 we have been in a multi-year period of heavier snows, since 2009-2010. From 2004-2005 to 2008-2009 snow was near or below average while temperatures varied. Here are the specifics:
In 2004-2005, it was a mild winter with below average snow. The following 9 years are listed in this chart:
With the notable exception of 2011-2012, snow has been heavier than average for the past 5 winter seasons. When we expand back to the first winter of the 21st century, we find with the exception of 2003-2004, snow was average or below average from the winter of 2000-2001 to the winter of 2008-2009.
Winter 2014-2015: Clues to the season
Now, we’ll take a look at the current conditions around the northern Hemisphere. First the sea surface temperature pattern, and specifically, what is called the anomalies (difference from the long-term average by being warmer or colder).
As of mid month, that warm pool of water continues over the northeast Pacific. The difference this year compared to last year is that the warmth is more along the coast from Alaska down the west coast and into the subtropics. Last year, it was centered over the open ocean and it was slightly cooler along the Pacific coast. The subtropical and tropical Pacific is also much warmer than it was last year. A pool of cold water has formed over the north central Pacific west toward China where it fades away. On the Atlantic side, we have warm water over the west central Atlantic from the Caribbean to Atlantic Canada. There is also an area of warmth between Labrador and Greenland and there is a pool of cold water in the north central Atlantic west of the British Isles. Last year the warmth in the Atlantic was weak and disorganized. Warmth in the oceans tends to help produce warm air and higher pressure aloft while cold areas help create cold air and troughs of low pressure aloft. Warmth in the tropics helps feed moisture into the subtropics and mid latitudes. Last year, we had a persistent ridge of warm air over Alaska, flanked by a trough over Asia and North America. That delivered a sustained flow of cold from Siberia and the Arctic into the U.S. This year’s sea surface temperatures favor a trough in the northwest Pacific, a ridge in the northeast Pacific and another ridge over the warm northwest Atlantic. Between, the atmosphere will compensate with a trough over North America. Once again, that sets the stage for the jet stream to tap Siberian/Arctic air and funnel it into North America. The warm tropical oceans will provide increased moisture to storm systems in the jet stream both out in the southwest and along the east coast.
One thing to keep in mind though. Even though the general tendency may be for a trough in the central or east central U.S. overall, where precisely this sets up will have a big impact on the overall winter; and during the season, the weather patterns will change over the course of several weeks or so and even totally break down at times.
Check out the graphics below which will help illustrate how a shift in the center of the trough from the longitude of Buffalo NY to Kansas City and finally to Grand Junction, Colorado has a big impact on our weather.
Snow cover is also important, aiding in refrigerating the Arctic air at the polar latitudes and keeping it cold as it heads south. Snowpack developed early and extensively across Asia and is now at most locations on land north of 45-50 north latitude. Some research has indicated early snowpack development in Aisa (Russia/Siberia) correlates with a colder winter here.
Another clue to the winter is to look at previous years when atmospheric and ocean temperatures were to some degree similar with what is currently ongoing or is likely to occur. Based on their similarity or lack of, to then weigh the seasons accordingly. Some are weighed once, others two or three times. The winter seasons which seem to best fit current and expected sea surface, atmospheric, solar and other factors are: 1968-1969, 1969-1970, 1976-1977, 2009-2010 and 2013-2014. When blended and weighed together, compared to the current averages (1981-2010) they produce a composite.
The temperature map has a wide area of cold from the northern Plains and Great Lakes through the Midwest, south and east. Across the district, the average winter temperatures were below average by -5F for STL, -4.5F for KC and -4 to -4.5 for SEMO/SW IL. E Kansas was -3.5F below average. A winter like the one depicted below (temperature) would be a winter similar to last year.
The precipitation map shows a broad area of dryness west of the Appalachians to the Central Plains and from there to the Gulf Coast. Winter precipitation (measured as rain and melted snow) averaged 1″ to 1.5″ below average. Remember that below average precipitation does not necessarily equate to below average snow. Kansas City last winter was a prime example, it recorded the 19th driest winter on record with precipitation 1.5″ below average but finished with 7″ above average snowfall. (Light and powdery snow, common in very cold air, can pile up quite heavily and have very little moisture content when melted.) A winter like the one depicted in this graphic (precipitation) would be very similar to last year from STL west through KC and Topeka, but it would be much drier for SEMO/SW IL. The reason it would be much drier was in large part due to the torrential rain that area had in December when 2-5″ of rain fell right before Christmas.
Winter 2014-2015: The outlooks
Here are a look at some various outlooks: These outlooks are from each organization and are not an “official” forecast from this site. The winter forecasts from the private forecasting companies are copywright by those companies.
The U.S. NOAA Climate Prediction Center issues an outlook for Winter. This is a probability outlook -not a forecast. CPC forecasts the probability of an area being above or below average. When there is no strong signal either way, or those signals conflict, CPC assigns “EC” or equal chances. CPC does NOT forecast how much colder/warmer wetter/drier or snowier than average it will be. The deeper colors only indicate higher probability, not greater extremes. CPC forecasts enhanced (slightly) chances for below average temperatures across the district, with higher probability of cold over the far south and warmth out west and in Alaska. CPC also forecasts slightly enhanced to moderately enhanced chances for drier than average conditions over our east — this is something seen in El Nino years (which this may be – although weakly). There is no signal for wetter or drier conditions in the western part of the district. The best chances winter will turn out to be wetter than average are from California to Georgia and up the east coast. Last year, CPC had “equal chances” for temperature and precipitation for the winter season.
The U.S. CPC Climate Forecast System V2 (CFSv2) resolutely calls for warmth over the entire U.S., strongest in the northern Plains. A majority of the country, according to it, will have average precipitation, except for wetness in from Texas to CA and over FL. Dryness is limited to the Portland to Seattle corridor, near Atlanta and the tip of Idaho. Last year, the CFSv2 predicted a warm and wet winter.
The Japan Meteorological Agency (JMA) climate forecast shows signals for a western ridge and eastern trough, however their temperature forecast is nearly universally mild (although only by 1 degree C) except for south Texas, which is one degree cool. Except for the western Great Lakes and far northern Rockies, their outlook is also slightly above average in precipitation.
The JAMSTEC model projects near average temperatures over the district with below average precipitation, especially for SEMO/SW IL and St. Louis. In 2013-2014, JAMSTEC predicted a warm, dry winter.
Accuweather, a private forecasting service, forecasts cold over the district with above average snow for STL/SEMO/SW IL and near average snow for KC. Accuweather has had questionable success in years past. The agency forecast a severely cold and snowy/icy winter in 2011-2012 (when it was very warm). Last year, the agency forecast severe storms in the eastern district for the winter, with snow and ice for the western district. It predicted near average snow in STL/SEMO/SWIL and above average snow in KC.
Weatherbell, a private forecasting service, forecasts a cold winter over the district and above to well above average snows. Weatherbell has had a very accurate record the past two years. Temperatures 4-6 degrees below average would be comparable to last winter. Weatherbell forecasts 167% of average snow for SE MO/SW IL which works out to 15″. 150% (estimated) of normal snow is forecast over the remainder of the district, which works out to 27-28″ in KC, STL and Topeka. This would also be comparable to last year. Weatherbell correctly predicted a colder and snowier winter for 2013-2014.
In summary, the odds seem to tilt to another cold winter for the district overall. Signals present in the week of Thanksgiving point toward a reversal of the November cold entering December, with more seasonable or mild conditions for the first week (or two) of the month with that mild trend returning to a colder pattern by late month, January and February. That cold should be more sustained and intense once again. How low temperatures go this winter can’t be accurately forecast at this point. That will depend upon how much snow cover (if any) is present when Arctic air arrives and how the center of each Arctic high times out across the area. To get subzero cold you need a perfect match of absolutely clear skies, calm winds (near the center of the high) and snowcover. You also need that high center moving overhead in the overnight hours and not in the middle of the afternoon.
In terms of precipitation- there are several strong indicators for below average precipitation. If the western ridge/eastern trough pattern takes hold as expected, that is a predominately dry pattern for us. That favors Alberta clippers in NW flow as the main weather system and those usually track on top of or north through east of our area, with their precipitation also to the north of us. In this type of overall pattern, our best chance for any big and moist weather systems are in periods when this overall pattern is breaking down or is becoming re-established.
With regards to snow: Again-hard to forecast due to low annual average snow totals (see the reasons stated above) but signals are pointing toward above average snowfall, most everywhere, with the greatest probability of above average snow for SEMO/SWIL. Odds remain in favor of above average snow but decrease westward and northward from STL to KC and Topeka. There are reasons for this, despite predominately unfavorable NW flow aloft and the below average precipitation expected.
For SEMO/SW IL: The first of which is that we’ll have increased moisture available to southern stream mid and upper level low pressure areas undercutting the broad ridge out west aloft. These systems, riding from the SW U.S. through the southern Plains and into the Southeast can be very efficient at generating snow in cold surface air — with weak or almost non-existent surface low pressure areas — we saw this several times in 2013-2014. A slight northward shift in the track of these mid or upper level lows could bring some of these snows to SEMO/SWIL. The second reason is that with below average temperatures expected, there is a greater chance for precipitation to fall as snow. The third reason was discussed above: average annual snow totals are low – only 6-12″ in this area.
For STL: Similar reasoning as for SEMO/SWIL, however average annual snowfall totals increase to 12-18″ in this area, reducing the likelihood of one or two snow event reaching the average annual snowfall total. This reduced risk is countered somewhat by the further north and eastward position of STL which puts it closer to the track of the clipper lows/snows should several of these systems track further south and west than is typical.
For KC: Above average snow expected; but odds of this verifying are lower than areas further east. The overall upper level pattern is a dry one for KC, with this area typically too far north and west of the track of any southern stream system to receive significant moisture and too far south and west of the typical track for clipper lows/snows. The expected cold pattern does work in favor of above average snows for the area as cold air aids in very efficient snow production even with limited moisture. As was seen several times last winter, several inches of fine, low moisture powdery snow can be effectively generated even with weak features at the surface and aloft. As history shows, odds favor more small light events than large heavy events. Best chance for major storms (7″+) for KC will be when the upper level pattern is in transition –going from or going back to– the expected prevailing pattern. How often this happens or if things work out correctly can’t be accurately predicted this far in advance.
For Topeka: Similar reasoning as for KC, but odds are just a little better than even that the forecast verifies, due to the even further westward location of the Topeka area.
Looking beyond this year, the recent trends of cold and snowier winters likely won’t last. Indications are for a reversal of the ocean temperatures back to more of what was experienced earlier in the 2000s and in 2011-2012. That may be a multi-year trend as well, all part of a natural cycle of ocean temperatures in the Pacific and Atlantic which extend across a decade or more. A period of snowier and colder winters in the 1970s was followed by milder and drier winters in the 1980s, and so on. Each winter will be different of course, depending upon the exact configurations in the oceans and the atmosphere, but the winter trends begin to look warmer and less severe in the years to come.