It’s that time of year once again where we break down the winter forecast. As always, everyone is excited to see what Old Man Winter has in store for us. Will the winter be filled with winter storms or unusually above normal temperatures? Maybe we’ll have both! So let’s break this winter forecast down and get rolling.
First let’s talk about how I come about making a winter forecast. I believe if you are going to read this forecast that you should know HOW I come to my forecast. I use a combination of factors revolving around observations that are available and how those evolving observations will potentially interact with each other. I do NOT put a lot of weight on long range model guidance from either the CFS 1 or 2 nor the ECMWF as they tend to be highly volatile from week to week or even day to day. As such, if you base your forecast on a model and that model changes, what really can you say about the winter? This was a key issue last year when stratospheric warming did not evolve as expected. Why is this important?
Well, if the winter forecast starts to go wrong, you and I will both know why and then adjust accordingly. That’s important because if you know the physical reasonings of a forecast then you can be on the look out and have clues IF the forecast is not panning out.
This forecast will be broken down into four sections. The first section will discuss the observations of sea surface temperatures in the Pacific and Atlantic with a focus on El Nino Southern Oscillation and the Atlantic Multidecadal Oscillation. The second section will discuss stratospheric influences including the development of stratospheric temperature anomalies, stratospheric wind anomalies, and influences on high latitude blocking. The third section will go over the cryosphere development in the higher latitudes with focus on Siberia, Alaska, and northwestern Canada. Finally, the fourth section will feature the forecast for the winter as I put all of the factors together along with storm track ideas.
SEA SURFACE TEMPERATURE ANALYSIS:
We start with an overview of the sea surface temperature anomalies for the entire globe. We can clearly see well defined areas of cold and warm temperature anomalies throughout the Atlantic and Pacific Oceans as of October 15, 2012. Of special note, I want you to consider the cold anomalies over the northern Pacific around the Gulf of Alaska, warm anomalies over the western Pacific near Japan, and very warm anomalies over much of the western and northern Atlantic.
Right now we’ll start with the Pacific with special focus on the El Nino Southern Oscillation and also the Pacific Decadial Oscillation or PDO.
PACIFIC OCEAN TEMPERATURE ANOMALIES:
I know everyone wants to know what is up with this El Nino. As I’m sure many of you remember, there was a lot of talk about El Nino back in the Spring. The usual sources were already claiming that this up coming year was going to be the hottest ever (yawn) due to the impending El Nino. Then of course, the Earth’s atmosphere and ocean currents threw a curve ball to show all of us once again how little we really understand about ocean and atmospheric dynamics and the forecasting of those factors. So what is happening here? I think we need to understand why the El Nino Southern Oscillation zones are basically averaging out to neutral before we can forecast for the winter.
The map to the left from NOAA shows a cold pool of water in the Gulf of Alaska. This is called a negative PDO where cold water pools in the northeastern Pacific Ocean. The ocean currents associated with the PDO drives colder water from this pool down the western coast of North America right into the El Nino regions, specifically around NINO 1+2 and at times NINO 3. We can see this influence especially in the Sub Surface sea temperatures. I should note that the PDO measurements last year and this year are on par, although slightly stronger this year so far. This same influence is what enhanced last year’s La Nina development in NINO 1+2 and NINO 3 leading to an east based La Nina evolution. This year, this cold influence has helped blunt the development of an El Nino forecasted by several models and has had a fundamental impact on the development of thunderstorm development in the Pacific and also mid and upper level wind patterns that were expected to develop. Instead of a robust El Nino development, the atmospheric conditions of the past two months have featured a neutral to extremely marginal El Nino characteristics.
Considering this factor, I don’t see any indication of the influences of this strong PDO subsiding any time soon and thus the idea of a strong or even moderate El Nino has to be taken off the table at this point. Frankly, I had this idea off the table back in April, but no point beating a dead horse.
What would jump start El Nino? Glad you asked!
There are several factors to consider for jump starting a weak west based El Nino pattern. The first factor I have been studying is the sub surface sea temperature anomalies. These anomalies can give us a good idea of what might be uncovered moving forward as lower and mid level tropospheric (where we live) winds develop over the next several weeks and months.
As you can see with this analysis here, we’ve seen a clear focus of sub surface temperature anomalies around the date line or 180 degrees Longitude for the past several months. As such, we’ve also seen a correlation (does not mean causation just yet though) of sea surface temperature anomalies remaining near or slightly above normal in NINO 4 and 3.4 for some time. This is the western portion of ENSO and would suggest a very weak west based El Nino.
The above normal sub surface temperature anomalies here though are VERY impressive with anomalies approaching three degrees Celsius above normal under NINO4 and only 150 meters below the surface.
However, this is all a mute point if you can’t get the warmer water to rise to the surface and for that you need a developing easterly component to the surface and mid level winds.
Here we have the latest anomalies at 850 MB, keep an eye on the winds around the equator to five degrees north latitude. You’ll notice a growing development of easterly winds from just east of NINO 4 to NINO 3, but still lacking a defined or strong easterly component around NINO 1+2 or around the Central America coast. Now, as I stated earlier, we can’t get that warmer water to rise to the surface without a sustained easterly wind. That easterly wind is clearly starting to develop over the tropical Pacific and that is why we are starting to see a rise in sea surface temperature anomalies in NINO regions 4 to 3 as seen below.
A very important factor to consider going forward is the state of the Madden/Julian Oscillation or MJO. The MJO is basically a statistical analysis and graphical representation of the intensity and location of tropical convection. When the MJO is in phases 6 through 1, the tropical convection tends to be focused around the 180 degrees longitude to 150 degrees Longitude. This location of the convection leads to a higher potential of a trough around Japan, a trough around the Aleutian Island, and also a trough over the eastern two-thirds of the United States. There are numerous studies showing that when the convection is focused around the date line that the Eastern United States tends to have cold and active winters. That doesn’t always mean above normal snowfall, but at the very least you are in the “game” when convection is focused in these locations.
When the MJO is in phases 2 through 5, the opposite is true and that is when we typically see a focus of convection well to the east of the date line near, southeastern Asia and the Indian Ocean. In these cases, a ridge is typically observed over the eastern United States while the West coast gets all the snow and cold action.
Now remember, this is only one piece of the puzzle so you can’t go making a complete winter forecast based on these rules. However, if you know the phases of the MJO over the next 45 days, then you can get an idea of the inter-seasonal pattern transitions. What I mean by that is no pattern locks in forever. As much as we all want to see one pattern and only one pattern through the winter (it would make my life easy then!), the atmospheric pattern is not a static state system. The pattern changes just like the climate changes. However, if you know what drives the changes then you can develop an idea of what to expect in the weeks ahead. Based on a study of stratospheric anomalies (I’ll go into that later) and a break down of torque analysis over Asia along with the Global Atmospheric Angular Momentum Budget (technical but you can study this feature here), the MJO will be the strongest in phases 6, 7, 8 and especially 1 this year. At the time of these phases, winter storms are the most likely along with periods for significant below normal temperature anomalies in the eastern two-thirds of the United States. The rest of the winter will feature neutral to very weak phases in 2 and 3. I think, unless a fundamental change occurs at the stratospheric level to influence tropical convection anomalies, that there will be very few periods if any of MJO phases in 4 or 5. This will suggest a growing influence of an eastern wind in the tropical Pacific and as such a growing likelihood for a very weak west based El Nino to neutral ENSO conditions this winter.
A weak EL Nino to neutral conditions will lead to the increased potential for troughs in the eastern two-thirds of the United States and tends to support prolonged negative EPO phases in the Pacific (trough around the Aleutian Islands), negative AO phases (Polar Vortex towards the Canadian/US border), and negative NAO phases (trough around 50N/50W and ridge over Greenland). These winters can lead to above normal precipitation and below normal temperatures over the East coast if other factors work together with the current Pacific state.
ATLANTIC OCEAN TEMPERATURE ANOMALIES:
While the Pacific Ocean has features below normal temperature anomalies, the Atlantic has been anything but cold through the Fall. As you can see from the map from the left from NOAA, the majority of the Atlantic Ocean is averaging above normal. This state of the Atlantic Ocean is not unusual but is due to a positive Atlantic Multidecadal Oscillation which as of September was measured at 0.487 degrees above normal. I should note that this is a very warm anomaly and mirrors the state of the Atlantic back in 2010. So what does this mean for the overall pattern for North America?
Well, first let’s start with the North Atlantic Oscillation. The North Atlantic Oscillation or NAO is a key factor in forecasting the impacts for the up coming winter as the pattern for a positive and negative state has a significant influence on storm tracks over the eastern half of North America. When the NAO is positive, the weather conditions are generally mild with a storm track shifted well to the west of the East coast. When the NAO is negative, storm tracks are focused more along the coast leading to an increase potential for winter storms. The extremes of both indices can lead to dry conditions as the storm track is either shifted well to the west or significantly suppressed. In fact, in the famous book Northeast Snowstorms by Kocin and Uccellini, we find that when the NAO is in a state of change that most major storms develop.
The current sea surface and sub surface temperature anomaly dynamic strongly supports a feed back pattern is developing in the Atlantic to support a prolonged negative NAO pattern over the northern Atlantic. There is a text book thermal gradient developing from the north Atlantic with well above normal temperatures to the central Atlantic with near to below normal temperatures and then above normal temperatures over the Caribbean Sea and Greater Antilles Islands. This type of sea surface temperature anomaly strongly suggest that storms will be developing in and around 50N/50W that will produce periods of intense troughs or upper level lows while also supporting ridging or building heights over Greenland and northeastern Canada. We have already seen the influence of the Atlantic sea surface anomalies on the weather pattern this October.
The next aspect to consider with the above normal sea surface temperatures in the Atlantic is the influence on storm development and intensification. Note that Sea Surface Temperature Anomalies (SSTA) are well above normal along the East coast, especially around the Carolina coastal waters. Naturally we expect water temperatures to be warmer around the Carolinas and coastal waters beyond 100 nautical miles over the Northern Mid Atlantic and New England due to the Gulf Stream. However, in this case, we are observing SSTA well above normal even over immediate coastal waters. This is an extremely important observation as this environment supports the development of rising air and thus naturally falling pressures. Cold air masses being more dense will naturally attempt to fill the void of air rising along the coast leading to significant clashes of air masses along the East coast, specifically over the Mid Atlantic. We can also find similar although not as intense development of above normal SSTA in the Gulf of Mexico as well. These observations suggest physically that colder, dense air is going to make a run towards the Gulf Coast and East coast leading to the potential for powerful storm development and an increase likelihood of cyclogenesis or low pressure development along the coastal plain of the East coast. This factor does not dictates the eventual track of these low pressure systems however as we have other factors to consider. Again though, this is one very important piece to the puzzle that we are putting together.
One of the factors that influences the winter weather pattern is the development and orientation of the lower stratospheric thermal state and it’s influence on the troposphere, where we experience the weather. Forecasting for the stratospheric environment for the winter is a very difficult task with a high level of volatility. Of all the aspects of the forecast, the development of major stratospheric warming events can have and has in the past had a significant impact on cold season 500 MB patterns.
Last year was a perfect example where the stratosphere had a key influence on the winter weather impacts throughout the northern hemisphere. As warm stratospheric anomalies focused over Asia and northern Europe, these locations experienced some of the harshest winters seen in some time. In fact, the Black Sea even froze as a result of the powerful cold and snowy weather pattern with several news outlets covering the event. Meanwhile, much of the United States experienced a mild and quiet winter with only one moderate snow storm along the East coast and even that storm generally underperformed.
Last year I gave a stern warning that the developments in the stratosphere were uncertain due to a mixed signal in the Quasi-Bennial Winds at 30 MB and 50 MB that conflicted with each other that had 30 MB with negative or easterly anomalies and the 50 MB having positive or westerly anomalies. The idea was that given the trends at the time of the forecast period that a gradual shift at 50 MB to an easterly component would lead to the support for weak high latitude blocking and a focus of warm stratospheric anomalies over North America. This did not occur and I saw this by early December, leading me to change the winter forecast to a far milder forecast.
This year the stratospheric environment is different on many levels. Let’s discuss the finer points here. First, let’s discuss what is the Quasi-biennial Oscillation is a periodic oscillation of the equatorial zonal winds between easterlies and westerlies in the tropical stratosphere in a 29 month period. Changes in the QBO state occur from top to bottom, which is why last year I though the QBO at 50 MB would eventually go negative. They did, but not until the winter was nearly over. The QBO was discovered in the 1950′s and was explained by excellent research by Richard Lindzen and James Holton in the 1970′s. The QBO when above 10 m/s leads to a mixing of the stratospheric ozone and transports warmer air from the tropical stratosphere to the polar stratospheric level. When the stratosphere features an easterly wind or a negative QBO, the transport of warmer air and more ozone from the tropics to the northern latitudes is far stronger and thus stratospheric warming is far more likely. The opposite is true when the QBO is positive.
This year, the QBO at 30 MB and 50 MB are both negative and strongly so. At 30 MB the QBO is down to -26.61 which is a very strong wind anomaly. At 50 MB the QBO has fallen to -11.42. So this year we don’t have to wait for a QBO wind shift and in fact we have the support from a terrestrial point of view for a warm stratospheric state. However, what about solar influences?
Yes, the sun plays a major role here in the winter forecast. Just a few days ago in fact, the sun produced a powerful solar flare that impacted the Earth’s atmosphere. This impact led to a rapid decrease in ozone at the stratospheric level which caused a rapid cooling of the stratosphere. This rapid cooling though appears to be a temporary impact as temperatures at 1 MB are already rapidly rising in a short period of time. The temperatures have also leveled off at 10 MB to 70 MB as well, showing the influence of the sun is starting to taper off. However, while the influence of this solar storm is coming to an end, we can learn some interesting impacts of the stratosphere on the 500 MB pattern which we have seen has lead to limited impacts on the tropospheric weather pattern, specifically the 500 MB pattern. Why is that?
The answer as always is location. Much like when studying the impacts of El Nino and La Nina, there seems to be a disconnect in understanding that location of thermal anomalies do matter and on a significant scale. For several weeks at the EPO, NAO, and AO have remained negative, we’ve seen a constant concentration of ozone and above normal temperature anomalies over the Barring Strait, central and eastern Canada, the northern Atlantic, and eastern Asia. These anomalies have to be studied at 30 MB as not to have a contaminated influence from conditions at 500 to 100 MB, which one could argue a feed back effect for the cause of thermal anomalies. At any rate, what we have here is a constant support for the 500 MB pattern to be forced into the negative EPO, negative AO, and negative NAO pattern and this pattern to be enhanced when the stratosphere as a whole warms, which we will see in significant stratospheric warming events.
Given these factors, I do expect the stratosphere this winter to average near to slightly above normal with an increase threat for major stratospheric warming events. Should these events occur, than watch out for a pattern that is very stormy and very cold. The key reason for this statement is again the negative QBO observations and trends and the resulting high latitude blocking setting up over much over the higher latitudes the past 6 weeks. We’ve seen the stratosphere run on the cold side and yet there remains a constant feature of high latitude blocking and a negative phase to the Arctic Oscillation (AO) and North Atlantic Oscillation (NAO) going back through mid September.
Let me say this about the stratosphere though. Influences from the stratosphere have been and continue to be a hotly debated topic in the world of meteorology. There are still many questions and even more theories on how the nature of stratospheric thermal anomalies and the QBO state impacts the troposphere. Even more interesting, gravity waves or changes in the mass of the atmosphere at a specific location can have unforeseen and significant impacts on the winter pattern as a whole, which is why this aspect of the winter forecast is always full of uncertainties and is a huge volatility threat for this or any winter forecast this or any year.
One of the popular observations throughout the internet is the study of the cryosphere or snow pack over the northern latitudes, specifically over northwestern Canada, the Arctic, and Siberia. These locations are important because the more snow and ice cover that develops in October over these locations, the stronger the Polar and Arctic air masses can get as the snow pack begins to reflect what ever sun shine that is left in these locations back into space. Think of the following of this observation in the terms of low temperatures on a winter night. Let’s take two locations with the same sky coverage, wind speed, and air mass. However, one location has snow and the other is bare ground. The area with snow is already going to be colder because more sun shine is reflected back into space than in areas where there is no snow cover. Now, night falls and the temperatures with the snow cover are falling more rapidly at a lower level, thus ending up with colder night time lows. The same idea follows with the snow coverage idea. With the winter fact approaching in northern Canada, the Arctic, and Siberia; there is less and less sun light. With this important fact, if snow is on the ground in these locations, the source of the cold air for our winters can become more intense.
Now, consider that other observations above about above normal sea surface temperatures in the Atlantic and you can see the potential conflicts in air masses developing as this very cold air interacts with very warm air over the northern Atlantic. Such a development in Siberia with above normal snowfall has shown that a feed back process develops in keeping a negative NAO pattern in place all winter. Further, above normal snow growth in northern Canada tends to lead to Polar and Arctic air masses invading the Plains and East coast when above normal sea surface temperature anomalies are also present in the Gulf of Mexico and western Atlantic as the dense cold air fills the void of the warm rising air along the coastal plains.
So now we know why snow growth is so important in a winter forecast because below average snowfall over the northern hemisphere this time of year can significantly impede on the development of Polar and Arctic air masses. Now let’s look at the boservations as of October 14, 2012.
As we can see on the map to the left as of October 14, 2012; the coverage of snow is rapidly growing with above normal levels in much of northern Canada and Siberia. The snow growth is also rapidly expanding an approaching near normal levels in northwestern Canada and the Arctic as well. Over the past 10 days we have seen a rapid growth in snow coverage, but we should be mindful that the snow coverage may not be deep as many locations are having their first snowfall of the season. Overall for this time of year, North America is basically near normal in overall arial snow cover, perhaps even a bit above normal. However, the snow coverage is currently not supportive of developing powerful Arctic air masses, at least not initially.
Naturally, going forward the state of the snow pack will be very important with the potential for above normal snow coverage in western Canada if medium range guidance is correct. At this point I put this factor as a neutral state influence with near normal conditions in place, nothing extraordinary can be found in the observations to point to a very cold nor warm winter with these observations alone. I will say that the snow coverage in Siberia, Greenland, and northeastern Canada is a very important observation to further support a predominant negative North Atlantic Oscillation over the next five to six months.
Beyond the snow coverage development there is one other issue I want to touch upon and that is the slow rise in sea ice growth over the Arctic Ocean. At this time I will not go into why the sea ice growth has been slow this year, but you can see from the positive state of the AMO why such an observation would make sense. However there is an issue of potential influence of such a dynamic on the high latitude wind fields that could help to enhance the high latitude blocking schemes in and around the Polar regions. Such a dynamic is still relatively new in the world of meteorological research but the theory currently in place is that the slow return of ice coverage over the Arctic Sea will help to enhance high latitude thermal gradients and thus enhance the blocking mechanisms associated with such thermal gradients. For example, note that over the past several weeks we’ve seen a growing support for strong above normal heights over and around the North Pole. Couple the factor of slightly warmer than normal thermal conditions over the Arctic Sea with cold stratospheric anomalies over the North Pole and you end up with above normal heights at 500 MB and more high latitude blocking.
So now it’s time to put all of the observations together and give you a forecast!
As you know, I’m combining a variety of factors into this forecast so let me give you my general idea of what this winter will be like. In one word, volatile and stormy. Note that I didn’t say snowy. Look, it’s a nice catchy theme to say above normal snowfall but at this point I don’t think making such a statement can be supported with the facts unfolding in the atmosphere.
This winter will be under the influence of exactly how the Pacific pattern sets up from week to week, thus the volatile adjective. When tropical convection is brewing around the date line, which will be forewarned by the MJO going into phases 6 and 7, a trough will intensify around the Aleutian Islands. This is called a negative EPO pattern.
When this trough shifts towards the Gulf of Alaska, the pattern is described as a positive EPO pattern. All Fall we have seen the EPO swing from positive to negative with the changing states of the MJO and naturally the intensity of convection around the date line. When the EPO is in a negative state, the pattern is aligned just right to set up a ridge in the Western United States which translates to a trough in the East. This trough in the East will then feature a negative Arctic Oscillation and negative North Atlantic Oscillation. Meanwhile, the Sub Tropical jet stream, enhanced by the MJO phases 7 through 1, interacts with the Polar jet stream configuration leading to an increased potential for winter storms followed by significant cold air advection.
On the other hand, when the EPO is positive, the whole pattern shifts and splits. A trough digs into the Western United States while the trough associated with the North Atlantic Oscillation attempts to hang on. The negative NAO is hard to dismantle because of the focus of stratospheric ozone anomalies over the northern Atlantic and the configuration of sea surface temperatures. As such, you end up with a ridge blunted in the East with warm air dominating the Southeast but not the Mid Atlantic and Northeast. This type of pattern will feature an increasing potential for storm tracks from the western Gulf of Mexico into the St. Lawrence River Valley.
Aside from these two overall pattern themes, the above normal sea surface temperatures observed along the East coast should foster the development of secondary low pressure systems and storm center jumps to the coast. The dominant low pressure system will be determined by the interaction of Sub Tropical and Polar disturbance and focus for the strongest Positive Vorticity Advection for each storm, but the point is that a theme of coastal redevelopment will be seen often this winter.
I suspect given the strengthening of the easterly wind component at 850 MB over the tropical Pacific that we’ll see a very weak El Nino develop and thus foster more frequent negative EPO phases rather than positive EPO phases in the weather pattern. So let’s look at the forecast map!
So here we have the general idea of the winter forecast. When the EPO is positive you can see the resulting storm track with low pressure system crashing into the West coast and driving through the Plains and then cutting up through the Great Lakes. When the EPO is negative, you see a far different evolution in the storm track with a focus more towards the East coast. This winter will feature Miller A and B storm tracks with the potential for significant winter storms for the Mid Atlantic, Ohio River Valley, Tennessee River Valley, and back west through the Southern Plains.
As you can see from the map above, I think the heart of the coldest air will be centered over the northern Plains, Great Lakes, and northern half of the Mississippi Valley. A powerful thermal gradient will waver between the eastern Great Lakes and Tennessee River Valley to the East coast. The most intense storms will develop and organize in these locations. As you can see, when the EPO goes positive, expect a storm track much further to the west but when negative, the East coast is going to under a high threat for winter storms.
Finally, I want to touch on the notion of putting out above normal snowfall forecasts. You might have noticed I’m staying clear of saying that description for the New York City metropolitan areas on south to the Washington Metropolitan areas. The reason why is due to my concern with the boundary layer dynamics. In other words, with water temperatures being above normal along the East coast, a slight shift in the winds will easily turn a snow storm into a mix of sleet, freezing rain, and rain. Still stormy but not snowy. There in lies the issue with issuing statements like above normal snowfall. I guess I’m cutting at symantics here, but I try to be as accurate as possible. This winter will feature far more active storm potential than what was seen last year for much of the eastern two-thirds of the United States. Whether that translates to above normal snowfall will be determined by the interaction and phasing of individual disturbances for each storm environment.