Wednesday, December 26, 2012

Polar Vortex Breakdown Will Lead To Frigid January-February

Confidence is growing that the polar vortex will sustain enough damage to break down in some levels of the atmosphere and lead to a pretty darn cold January into February.

We'll start out with observations. Above is the observed temperatures in the 70mb layer of the stratosphere, known as one of the lower layers in the stratosphere. We are looking on the far right side of this picture to see current values. If we do so, we see that two warming events have already taken place. The first warming event brought temperatures well above normal in the 70mb layer, and there is an ongoing warming event taking place that is leading temperatures to skyrocket in the lower stratosphere.

[Animation of 50mb]

This animation of observed temperatures in the 50mb layer also shows such recent warming (for those unfamiliar with the millibar scale, lower numbers = higher height in the atmosphere). Both warming events previously described on the first graph are shown in this animation, with the first warming event propagating from East Asia into the North Pole. That warming event then died down and slipped back into East Asia. From there, that warm body of air moved across the North Pacific and is now in Canada, where it may be trying to make a move into the Arctic Circle. If such an event happens, this warming will become much more significant than the last.

The main subject of this post is something called the 'Polar Vortex', or PV for short. The Polar Vortex is a low pressure system stationed over the Arctic, with varying strengths. The PV is actually what drives the well-known Arctic Oscillation. When the vortex is weak, the AO is negative. When the vortex is strong, the AO is positive. We'll talk more about the AO later on. But it is important to note that the polar vortex is weakened when significant warming events, like the two shown above, happen.

Now, let's go back to the first chart. See how we are already well above normal for this time of year? That means that the polar vortex is theoretically weaker than normal at the moment. If that warming in Canada continues and/or strengthens, the polar vortex will only get weaker and weaker. As long as those above normal temperatures continue, the polar vortex is indeed vulnerable to collapse, if the right parameters come together.

There is something else used to see how the stratosphere is being impacted by warm air, and it is called the Eliassen-Palm Flux, or EP Flux for short. The EP Flux measures the strength by which air from the lower atmosphere is being forced into the stratosphere. Considering the word 'flux' means movement, one can think of the EP Flux as showing the strength of movement of warm air into the stratosphere. Shown above is an 8 day observation of the EP Flux. In recent days, we have seen the EP Flux strengthen, as the second warming event continues to hold its ground. Note the new presence of colorful arrows in the area above the number '10' on the lefthand side of the Dec. 24 image. This means that warm air is now penetrating the upper stratosphere, effectively infiltrating the entire stratosphere. This is a major blow to the polar vortex, as warm air is now pushing into the entire vortex, not just the lower part.

The ECMWF forecast for EP Flux is at the bottom of this multi-image forecast. Looking at that EP Flux, we see that the European model is forecasting the flux to rapidly strengthen as the New Year approaches. This would mean warm air from the troposphere would be shooting into the stratosphere at a rate more than triple the strength we saw with our first warming event in early December. That first warming event is shown as a cluster of arrows above the word '1DEC' in that same bottom image. Should such a forecast verify, it would be the highest EP Flux values we have seen this winter season, and would certainly induce very heavy stratospheric warming- even more than we have seen to this date. If any further warming occurs (and if the EP Flux forecast shown above verifies), we have a shot at breaking the warmest value in the 70mb stratosphere on that date (shown in the first graph as the top light gray line)- something that means chaos in the polar vortex.

Continuing with observations, we visit an index called 'Mountain Torque'. The Mountain Torque, also called MT, is a fairly challenging index to decipher. However, I have come to understand it as a component that can help warm the stratosphere and weaken the polar vortex. In high values of mountain torque, it is expected that the stratosphere will warm rapidly in a phenomenon called the Sudden Stratospheric Warming (SSW). (The Sudden Stratospheric Warming got its name due to the rate at how the stratosphere warms, hence the word 'Sudden'. There is also a weaker sudden warming event called a Major Stratospheric Warming event.) Considering the mountain torque values are at values not seen since late fall, I find it very plausible that the next 5-15 days will bring another good warming to the stratosphere, a crippling blow to the vortex as the current warming continues.

Now we get to the forecasts of what the models say will happen.

We start with the GFS model and what it says about warming in the stratosphere. Forecasters commonly use two types of graphs when dealing with the stratosphere: Temperature forecasts and pressure forecasts. The former option shows warmings in the atmosphere, the latter shows the strength of the polar vortex (remember, the polar vortex is a low pressure system just like ones we see in the US, just bigger and more permanent). We'll start with the temperatures. Below are 5 images of 5 layers of the stratosphere; I'll annotate each image appropriately.

1 millibar 10 day forecast

2 millibar 10 day forecast

5 millibar 10 day forecast

10 millibar 10 day forecast

30 millibar 10 day forecast
You might be wondering why you should care about a bunch of reds and blues and fancy colors. Well, in each of these images, significant warming in the form of very warm colors is shown next to very cool colors, a.k.a. the polar vortex. Anytime you see those reds and grays, that defines very warm air finally reaching the upper stratosphere.

Something very interesting I saw in those 5 images, something you may have noticed as well, is the steady westward progression of the warm colors with each image. If you can picture 3-dimensional things in your head, you may have thought of a spiral, in the sense of the warmest areas of each respective stratospheric level rotated as each image showed a lower level of the stratosphere. If you want to compare it to material objects, think of rotini pasta or how you get bubblegum from the gumball machine; how it goes down the spiral to your waiting hand.

That spiral analogy leads us into the next significant forecast piece- displacement. You may hear this word and think back to 8th grade chemistry and water displacement, or even high school, but this is a different displacement. This is the displacement of the polar vortex.

100 millibar 10 day temperature and pressure forecast

1 millibar 10 day temperature and pressure forecast
Shown above are two pressure and temperature forecasts from the ECMWF model. The letter 'L' defines the center of the polar vortex for each level in the atmosphere, and the letter 'W' shows the warmest values in each respective layer of the atmosphere. If you notice a plus symbol, that is the center of the Arctic, quite literally the North Pole. And if you happen to see the letter 'H', that symbolizes the highest pressure for that atmospheric level.

I said we were looking for displacement here, and that is what we will do. We are trying to find how far apart the letter 'L' is from the 1mb image to the 100mb image. If we look in the top image (100mb), we see that the polar vortex is centered just to the south of the North Pole, a.k.a. just north of Europe. The 1mb image on the bottom has the polar vortex just east of Greenland, and high pressure building in over Eurasia.
While this is not a classic example of extreme displacement, it is apparent that there is some good displacement. The top image shows two apparent centers of low pressure- one with the 'L' that I described in the above paragraph, and one not marked but still shown in the top right corner of the image in what appears to be East Asia. On the 1mb chart, high pressure has built in right over that second low pressure area that I told was in East Asia. As for the deepest centers of the polar vortex, they are relatively close to each other, meaning the displacement is not that extreme, but far enough apart for me to introduce my Jenga analogy:


Think of the game Jenga. You must try to take out pieces of the tower to get as many pieces out, but still have it standing. The more pieces you take out on one side, the more unstable that side becomes, and eventually the tower falls down because of the difference in stability of the tower (a.k.a. where more blocks are placed). In a similar situation, when you have the polar vortex far apart in different levels, like what is forecasted above, the polar vortex becomes more unstable and weakens, possibly to the point of collapse.

One more analogy for you: Think of the polar vortex like a cylinder full of cold air. If you cut the cylinder in half and move one half away from the other, the cold air will sink. That is exactly how the stratosphere. If you have displaced parts of the stratosphere, the cold air that is held in the polar vortex will be released and flow down into lower latitudes. If the PV is displaced enough, it may collapse altogether, leading to an icebox solution over parts of the world.

The GFS model shows a similar situation as far as that high pressure/low pressure difference in Asia, but has a lot more displacement in the actual polar vortex as shown below:

1mb 10 day pressure forecast

100mb 10 day pressure forecast
The GFS dislocates the polar vortex far away from itself in between these two levels, which theoretically would enhance the probabilities of a weaker polar vortex and thus higher potentials for a strong cold snap in January and February.

So, we have big warming in different levels of the atmosphere per the GFS big pressure differences in the polar vortex in the GFS, as well as good displacement of the vortex in the ECMWF. But there is something else showing up in the GFS- complete collapse of the upper stratosphere polar vortex.

1mb 384 hour pressure forecast
5mb 384 hour pressure forecast
10mb 384 hour pressure forecast
This is where you will need to read these images closely. We'll start with the top image at the 1mb level. In this top image, we see a large circle covering much of the Arctic. If you read closely, you will see that the numbers go up as the lines get smaller, meaning that this is a gigantic high pressure system. The polar vortex has collapsed in the 1mb layer at Hour 384! The 5mb layer shows a big high pressure system over the Bering Sea splitting up the polar vortex into two pieces, and a similar forecast resides in the 10mb image.

You weather aficionados are most likely wondering why I'm showing forecasts at Hour 384, the longest of the the long range, and the high point for forecast failure rates. The reason is simple: Trends. I went back across several GFS runs for the 1mb forecast and found that the high pressure system was still covering the Arctic in as many as 5 other forecasts. That said, my faith in this forecast increases dramatically, and, while it's still quite a ways out, I am encouraged by this forecast of a polar vortex collapse in the very upper stratosphere.

Remember how I said the polar vortex is the driving force behind the Arctic Oscillation? Let's keep that front and center. This is a multi-model (and ensemble) forecast of the Arctic Oscillation. In this forecast, we see that the GFS, GFS Ensembles and GGEM Ensembles project the AO to be in the positive range come the 2nd week of January. Based on all the evidence I reviewed above, do you think that will actually happen?

This 8-10 day 500mb height anomaly forecast from the ECMWF (left) and GFS (right) shows no striking presence of a positive Arctic Oscillation, which would be shown as below normal height anomalies across the Arctic Circle. While we do see some low pressure anomalies creeping into the North Pole, nothing defined is showing up, meaning I believe this positive AO forecast is bogus. Considering the warming already happening, the warming forecast to happen, and, potentially, the collapse of the upper stratospheric polar vortex, I have a hard time believing the models are correct with a +AO in the future.

In summary:
-Significant warming is already ongoing in the stratosphere.
-EP Flux values are now pushing warm air into the entire stratosphere.
-EP Flux forecasts show dramatic and significant warm air being forced into the stratosphere.
-Mountain torque highlights more stratospheric warming in the future.
-Model forecasts support heavy damage being sustained to the polar vortex.
-The positive AO forecast does not look correct in my opinion.

All in all, what I have described here today bodes well for a cold episode that may span the months of January and February. If the polar vortex has a collapse, severe cold could very well overtake the nation. Prepare for cold times ahead.

Andrew

37 comments:

Anonymous said...

Andrew what does this mean for winter snows in Iowa? Will it be too cold for snow?

Anonymous said...

Can you make a map for this? It doesn't have to be exact, but where in general are you expecting the cold/snow to be?

Rod Houston said...

Didn't you said that the polar vortex could break last year but it never happened?There's not going to be a big cool for the U.S.It will just keep being warm and almost no cold at all.How do you know that the polar Vortex will collapse?Stop hyping stuff up already.

Andrew said...

Iowa Anonymous: This cannot be forecasted for a particular location.

Andrew said...

Anonymous at 12:56: A map cannot be made for this- this is a general thing, not for any particular location.

Andrew said...

Rod: I provided hard evidence for my claim, how about you provide evidence for yours?

Anonymous said...

Rod doesn't have any evidence because he doesn't know what he is talking about. He only wants to cause trouble. If there is a big cool down in January/February we will not hear a word from him.

Rod Houston said...

Stop hyping the even to get people's hopes up.It's not gong to happen.The Low in the 8-10 range will be centered over Greenland which will set a ridge over the U.S.The Euro and the GFS says the low will be in western Greeenland and it looks like Alaska will get very cold again.The MJO says it will be warm in the second week of January.Unless it's an error by the MJO which is most likely not,it will be warm across the U.S.Phase 4 of the MJO=warm for the U.S.

Andrew said...

Rod: The only one hyping is you- The MJO will still be fairly weak, so it's effects will not be as strong. Also, the MJO is one of dozens of factors that influences the weather, you can't take one and run with it. You aren't one to say whether the polar vortex will break or not without providing sufficient evidence like I did.

Anonymous said...

lets hope it breaks down and gets annihilated. i hate it..stay gone forever!..let it come snow!!

Anonymous said...

The question is this: Whose reading of the tea leaves is right. Meteorology seems like political polling, in many ways more an art than a science. I want snow (not frigid weather, though -- 32 degrees is cold enough for me), so I want Andrew to be right.

Anonymous said...

Hi Andrew do u think the chances of this occuring are pretty likely or do u think the models will start to show otherwise accuweather isn't showing any real cold temps for the next 25 days where i live in east tn and they are showing warm temps for michigan too. well warm for the standards and the nao is going positive and the ao is showing it going WAY positive where is this cold coming from and will it come

dillon said...

very last anony. i checked 3 models..ALLL 3 are showing a very big warmup to take over the nation in 7-14-20 days. i have ALSO checked accuweather and i didnt really believe what they was saying. but i see where their getting it. the models are showing jet stream going up north past INDINAPOLIS!!..very warm is what they are showing for the coming weeks..sucks i know.

Rod Houston said...

Dillon your are right. It looks at like the models have a zoneral flow on the 8-14 range with the NAO positive and the AO is very positive. I think Andrew is hyping this up.The 540 line will be well north of Chicago.This stratosphere warming event and the Polar vortex breaking down is just hype.Nothing else.

Andrew said...

Rod: I'm not going to respond to someone who's here just to start trouble.

Andrew said...

Dillon: I disagree- models are showing a cool down starting off January and well into January. I'm not sure what models you're using, but there are more than 3 models and also ensembles.

dillon said...

andrew im not here top start problems..and i know u wasent point what u said to me..but i was jsin. but rod im not gonna argue with andrew because im sure he knows 10x as much as me..if not..more. but i do know how to read models....i know enough by looking at the models to say to other people what their saying to me. and idk what the heck the ensembles even are so yeh. but this is the link to what im using.http://www.wunderground.com/wundermap/?zoom=4&rad=0&wxsn=0&svr=0&cams=0&sat=0&riv=0&mm=1&mm.mdl=GFS&mm.type=SURPRE&mm.hour=0&mm.opa=100&mm.clk=0&hur=0&fire=0&tor=0&ndfd=0&pix=0&dir=0&ads=0&tfk=0&fodors=0&ski=0&ls=0&rad2=0
it has a model section u can control urself. i was using the models *GFS* *NAM* and *ECMWF*...the gfs is the model that can go out the furthiest and its showing warm 40-50 degree weather pooling well into east central america.
i am probaly reading them wrong,hopefully i am. but i highly doudt it considering it has the temperature gaudges and everything. and i also looked at the jet stream and the 925 mb wind. and the win(kts) to help show me the dips in jet stream.

Andrew said...

Dillon: I was talking to Rod when I said I wouldn't respond. Using the GFS and ECMWF is good, the NAM is a more short range model; it doesn't go out as far as the other two. As for ensembles, it is a group of forecasts from one model, typically giving a better forecast than an individual model.

dillon said...

i know u ment him..but i was jsin.
i also dont think this site has ensembles. ive been looking on google and trying to find good sites for the models and long range forcast. cuz im kinda of a minor predicter for my county.i base my idea off of everyone else's idea.
but why would ur GFS model be showing a cool-down...when my GFS model i use, is showing a warmup??

dillon said...

i meant this site as in the link i posted

Anonymous said...

Rod, it sounds like you have something against Andrew. I wouldn't consider it "hyping" when he (on his own weather blog) writes that he thinks the data is pointing to a cold January and February. It's not like he is blogging about it daily and telling us it is going to be record breaking cold. And he is not alone on this. I have seen several other meteorologists and bloggers say the same thing. Also the most recent CFSv2 forecast has each month cold and the Weather Channel's most recent forecast for those months were also cold (especially February).

Anonymous said...

Dude, some of the images he is showing are stratospheric. These maps are not on Wunderground. While they do mean stuff, they are still in the experimental phase right now because meteorologists don't have a complete grasp on them quite yet. The only non-stratospheric maps that he showed were an 8 to 10 day forecast for the northern hemisphere. Unknown to most forecasters, that's all they show. Lots of folks use them as a 'long range prognosticator' when they really only show the pattern between days 8 and 10. I believe that the maps that you were using may have been in a different time frame than Andrew's. Like I said, Andrew's model only shows days 8 through 10, while yours might have shown Day 14 for example. That's probably what is going on.

Anonymous said...

My previous comment was directed at dillon

dillon said...

to the 1st anony. u are right. the only people that are the complete opposite. NOAA - their saying MUCH of the nation will be alot warmer. and as time goes by..it will only get warmer.
2nd anon - i wasent basing what i said off his maps. i was basing it off what he has said. as in the first few weeks ..few as in (1-2) a.k.a 14 days..he said the models was showing him a much cooler pattern. when the SAME models i use..show me warmer.

Anonymous said...

Well then maybe you two are using different runs. A lot of folks still haven't grasped the fact that long-range meteorology isn't exactly a science and that the stuff models spit out can vary significantly between each model run.

Rod Houston said...

Anonymous I have nothing against Andrew.The models I have looked at shows warmer temperatures and with the NAO positive and the AO way positive I just don't see to much cold for the U.S in the long range.If the models end up showing cold then he's right but most models shows warmth.I not trying to cause any trouble.I want cold and snow as much as anyone does but models shows warm for the future.

Andrew said...

Rod: I understand your viewpoint after looking over several analysis tools. What I failed to mention is that this freeze would take place later on, as it takes time for cold air to propagate down from the stratosphere. I expect mid January to solidify cold. My apologies if I offended you or anyone else.

dillon said...

thats same i show. and andrew, im hoping for that artic freeze....sounds like its gonna be..cold.lol,and i wouldnt know ANYTHING past hour 384 which is around january 9th or 10th. thats the furthiest the models ive ever seen will go. if u have models that go further. please share!

dillon said...

andrew i have a major question. this artic freeze. the very last time we had a polar vortex breakdown was in 1985. once the polar vortex dropped..TONS of record low temps was set everywhere across the nation. scientiest are saying this polar vortex collasping is no joke and that it could very well lead us into a new mini ice age?true?

Rod Houston said...

If the polar vortex somehow collapse.It will get very cold across the Country.Could the Polar vortex collapsing mean too cold for snow in Chicago?If the Polar Vortex doesn't collapse,will it still get cold for January?

Anonymous said...

Andrew just quit responding to Rod he just wants to argue

Eric said...

Andrew, I first want to say I love your website and I know you work really hard as a forecaster (as I do) in trying to decipher weather patterns. lol, I see Rod Houston stirring things up, he does the exact same thing over at weatheradvance. I remember he was saying last week that "it is just like last year" or "I don't see and cold or snow coming" then next thing you know back to back blizzards strike the US, I think it's a good thing that he doesn't agree with you Andrew, given his history you should feel quite confident in this forecast for this SSWE.

Andrew said...

Dillon: We need another 10 years until the sun gets much less active to even consider such a possibility.

Eric: I must admit I admire your forecasting as well- you are definitely one of the more skilled forecasters in the weather community. Hopefully this SSW means something good for mid January...

Eric said...

@ Andrew
Thank you, and the long post you made about the stratospheric warming event was very impressive and I certainly have learned a few things from you. In fact, if you had not talked about the northern hemisphere snowfall anomaly, I may have never had the idea for the 2nd half of winter to be cold. In case you haven't seen it, here's something I posted as a comment on weatheradvance and weatherbell premium. everything may be coming together towards late January and February. This idea is something I’ve had since at least December 9th, when I realized how the northern hemisphere snowfall anomaly seemed to correlate with the meteorological winter (Dec 1-Feb28 (29)), where we started off December fairly warm over much of the US, then as we got into mid-December things cooled down a bit, but were still somewhat mild, and now we find ourselves in late December with a relatively cold pattern over much of the US, but indications are that this pattern may get even colder as we enter into the first week of January. However, like any good winter there obviously will be periods in which it must be milder (at least in comparison to the other parts of the winter) and this winter is going to be no exception. I actually looked into instances in which we saw a trough near Greenland (+NAO) line up with a trough near the eastern US, and in late January-mid February 1995 this occurred where a large trough was over the eastern US, with a general region of low pressure that extended well to the northeast towards Greenland and Iceland. This kind of trough alignment is quite rare as it is rather difficult for a polar trough and mid-latitude trough to vertically align because they are so far apart, also due to the fact that storm systems move eastward towards the arctic, while storms move westward in the mid-latitudes, and the differences in the amount of distance and difference in direction make it hard for this kind of set-up to occur. However, when it does occur, usually the mid-latitude trough becomes infused with very strong arctic air, and a strong storm system is the result in the mid-latitudes, but I do not think this kind of extreme scenario is going to happen due to all the conditions that have to line up, and the rarity of such a set-up occurring. Interestingly though, that trough alignment occurred during a stratospheric warming event, which is also what we’re beginning to see at the moment. In general, I think we are going to see a much colder pattern evolve over the relatively short-term as this storm system around the 30th will move fairly close to the northeastern US and produce snowfalls from the southern plains to Atlantic Canada, with totals in your area (I agree with Dante) being around 3-6 inches, but there is still some uncertainty and these may need to be adjusted with time. Also, as this storm system passes to the north and east it will get stuck under the transient block near the Hudson Bay, and in doing so, this will force the jet downstream all the way into the Pacific to buckle in response. As this low pressure meanders over eastern Canada it will become much larger and may take on characteristics of a large gyre of low pressure. If this occurs, then like the spokes of a wheel, Alberta Clippers will rotate around the base of this low pressure system as they feed off the differences in pressures and temperatures from western North America to eastern Canada and the northeastern US, and with each clipper system, light accumulations of snow may impact the upper midwest, Great Lakes, and northeast with each system reinforcing the arctic air that will be in place towards the northeastern US as we get into January. Beyond this timeframe, conditions over the US in general may moderate some, as the NAO and AO go positive (indicative of the cold air at the surface loading over the arctic thanks to the stratospheric warming event), and a more zonal flow may take over.

Eric said...

Here's the rest of what I said, see what you think about this, "However, this pattern would be short-lived as the cold continues to build over the arctic, and the breakdown of the polar vortex continues, and as you know, colder air induces pressure rises, which leads to a more -AO look, and with the potential for the polar vortex to completely breakdown, which is something that has not been seen in quite some time (as noted by Andrew of the weathercentre) this could lead to a very deeply negative crash of the AO (and NAO). Based off of what I noted in my post from December 9th, the 500 millibar pattern in December showed some similarities to 1984-85, which also had a complete collapse of the vortex and a major stratospheric warming event, (both of which have the possibility of occurring over the next several weeks), and as a result some of the coldest air of the 20th century moved into the US. Also, this idea for a rather cold period is supported by one of Dante’s most recent posts which made parallels to 2002-03 winter, which also had a major stratospheric warming event, and really became cold in late January through the rest of the winter. Although this type of pattern comparison to 1984-85 is rather extreme, certainly has the possibility of occurring starting towards late January given some of the conditions at hand and the history of these kind of events with the parallels in the overall patterns in other years like 2009-10, 2002-03, and 1984-85, all of which turned out to be very good winters for the US towards the 2nd half of winter (especially the eastern US). My ideas remain unchanged from early December, as I continue to think based off of all of this information above and other info provided in some of my posts, that the 2nd half of winter starting in late January, is going to much colder and snowier than the 1st half. If the conditions at hand and history prove to be true, then this could be a particularly cold pattern that sets up in the 2nd half of winter."

Indndawg said...

I would love to see Jan/Feb 1985 happen again

Indndawg said...

I've noticed true cold air is having a hard time penentrating far south (Memphis to B'ham line)

Seems the westerlies keep the very cold temps more to the north. Am I wrong

I hope what's coming is similar to J/F of 85. Cold and snow for the southland