Monday, November 11, 2013

Long Range Thoughts & Potential Thanksgiving Major Winter Storm

**This post is dedicated to all former and active soldiers and veterans across the Armed Forces.**

I was taking a look at the long range this morning, and decided it was time to do a post on what my thoughts for the middle and end of November are. However, as I was sifting through the data I wanted to include in this post, I came upon the realization that this data indicated a threat for a potentially major winter storm around the Thanksgiving timeframe. Let's jump right in. We'll start off with the long range thoughts, and then go into the major winter storm threat. At this time, the winter storm discussion remains preliminary. Nothing is set in stone for this potential, and the reason I include 'major' in the possibility is due to certain composites I have at my disposal that indicate this could be a big winter storm. Again, it's nowhere near certain. But let's begin.


We'll start with the teleconnections, with the Arctic Oscillation up first. The image above shows the GFS ensemble's projection for the Arctic Oscillation from now until a couple days before Thanksgiving. As we stand now, the Arctic Oscillation (AO) is in a strong positive state. The positive phase of the AO calls for a stronger than normal vortex in the Arctic (commonly referred to as the polar vortex), and a tendency for warmer weather to prevail across North America. This has been the case in the last few days for some in the United States, but with the big cold snap coming into the heart of the nation starting today, that warm weather trend will be quickly reversed. Bear in mind that the Arctic Oscillation is not exactly what I consider to be a significant influence on the American weather when compared to the other teleconnections. For one, the AO is based in the Arctic, and while the jet stream is influenced by the AO's actions, the Arctic isn't upstream of North America like the Pacific teleconnections are. Because those Pacific teleconnections are upstream of the nation, they wield significant influence on the North American weather pattern. Regardless of which index wields more power, it is agreed on that the Arctic Oscillation does play at least some role in the US' weather, and this is why it bears watching. With the AO heading down to neutral territory about a week prior to Thanksgiving, and possibly hitting negative territory when Turkey Day comes about, the tendency for warmer weather should gradually let up in coordination with the weakening +AO.


Next, we'll take a look at the close cousin of the Arctic Oscillation, the North Atlantic Oscillation (NAO). The NAO has a positive and negative phase, like the Arctic Oscillation. The NAO is determined by pressure anomalies in and around Greenland- if there is abundant high pressure in Greenland, the NAO is determined to be negative. Similarly, if stormy weather is found in the Arctic landmass, the North Atlantic Oscillation is declared to be positive. Generally, the negative NAO is the index that is the most beneficial for winter weather folk along the East Coast. When the NAO goes negative, the jet stream shifts in a way that not only allows cold air to sink down into the Northeast, but also grabs storm systems from the Southern US and swings them up north to ride the jet stream along the East Coast. This set-up usually leads to the infamous Nor'easter coastal storms. Right now, we see the NAO is in a moderately positive phase. Like the Arctic Oscillation, a positive NAO tends to enhance the likelihood of a warm weather pattern in the US. However, as time goes on and we progress into the middle of November, the ECMWF ensembles (pictured above) project the NAO to gradually shift into a neutral phase, and possibly a negative phase as we move towards Thanksgiving. The ECMWF ensembles don't really show this negative NAO shift at the end of the forecast period, but the American ensemble system is set on the NAO going negative at around the same time as the Arctic Oscillation. These two indexes going negative for the middle and end of November spells good news for winter weather folk across the board, and will influence the potential Thanksgiving winter storm threat.


The third teleconnection we will discuss rounds out the 'Big Three' teleconnections that seem to have the most influence on winter weather in the United States. The image above shows the GFS Ensemble forecast for the Pacific-North American (PNA) index, which is based along the West Coast of North America. Like the previous two indexes, the PNA has a positive and negative phase. In the positive phase, high pressure anomalies are found along the West US and into Southwest Canada. This allows the jet stream to buckle south in the Plains and Midwest, giving those regions cold weather. The shifting jet stream also permits the Midwest and Ohio Valley to receive some decent chances for winter storms. In the negative phase, cold and stormy weather is found in the West US, and this alters the jet stream in the opposite way as the +PNA. The buckling of the jet stream occurs in the West US instead of the East, and this then leads to, you guessed it, warm and tranquil conditions for the East US. The PNA has been slightly negative and still is as of now, but is projected to gradually push into neutral territory in coming days. In the long range, ensemble guidance indicates the PNA should push into positive territory around Thanksgiving. This will be a big factor in the potential winter storm I will discuss later on.


As if this long range discussion wasn't long enough already, we're going to dive into our fourth teleconnection, the East Pacific Oscillation (or EPO). There are two phases associated with the EPO; positive and negative. In the negative phase. warm weather tends to arise along the Northeast Pacific, as a result of ridging in the region. Downstream of those waters, colder weather is favored in the Central and East US. In a similar way, the positive EPO comes about from stormy weather in the Gulf of Alaska (same as the Northeast Pacific waters I discussed with the -EPO), and leads to warmer than normal temperatures across a wide swath of the Central and Eastern US. Currently, the EPO is in a deep negative phase, and looks to stay there for the next week or so. Beyond then, ensembles fluctuate but eventually settle the EPO back down in time for a moderately negative EPO in time for Thanksgiving. This fourth index, while not as significant as the other three, will play a role in the long range temperatures and where this potential winter storm will end up.

The final teleconnection to discuss is the Madden Julian Oscillation (MJO), which is more like a whole other index than a teleconnection. The MJO tracks convection anomalies along the Equator, and, depending on the strength and location of the enhanced convection, is sorted into one of eight phases. Each phase has different temperature and precipitation effects for different times of the year. Per the current MJO analysis and forecast above, we see that the MJO is in the white circle in the middle, meaning it is too weak to have a significant impact on our weather. However, towards the days leading up to Thanksgiving, model guidance indicates that the MJO will strengthen again in Phases 1 and 2.


Composites of November MJO phases 1 and 2 reveal that the West, Plains and Midwest experience cooler temperatures in Phase 1, with a very wet pattern for the Midwest, Gulf Coast and Ohio Valley in that same phase. The East Coast ends up warm and dry for Phase 1, but the opposite is true for Phase 2. The Plains experience warm weather and dry conditions in Phase 2, while the East is bathed in cooler weather. Per the latest teleconnection guidance and MJO forecasts, I would expect a cool Central and East US for the days leading up to Thanksgiving, with the wet pattern possibly persisting across the Central and portions of the East US. Because model guidance has the MJO so weak in the long range, I'm hesitant to say it will have a big impact on the weather pattern for the end of November, but I am confident in saying that it will have at least a slight impact.

Based on all of the teleconnections and indexes presented to me, it looks as if the weather pattern will undergo some change in the next week or two, as a new, more winter-like pattern takes hold near Thanksgiving. This will set the stage for what could be a major winter storm in the US, and that's what we will discuss next.

So what's all of this talk about a Thanksgiving winter storm? It's only November 11- how can I know this is actually a possibility? Well, we know that this is at least a possibility due to the weather experienced in the Bering Sea recently, and the weather East Asia will experience soon.

Take a look at the Bering sea in this two-panel reanalysis of the weather on November 8th. If you look closely, you can see that the Bering Sea was in the midst of a decent storm system on this date, just three days ago from today. This storm system passed along those waters in the midst of a mammoth ridge of high pressure just to the south. The 500mb height anomaly chart on the left best reflects this storm system, but its presence is confirmed on the right panel, which displays mean sea level pressure contour lines and denotations, as well as cloud cover. This Bering Sea storm has a connection to this potential Thanksgiving winter storm. Based on research done by Joe Renken, weather in the Bering Sea correlates to weather here in the US approximately 2.5 to 3 weeks after the Bering Sea weather anomaly occurs. This storm system in the Bering Sea happened on November 8th, and extrapolating that out 2.5 to 3 weeks ( 17-21 days) leads us to a potential storm system impacting the US around November 25 to 29. Considering November 28th is Thanksgiving, there does appear to be at least decent potential for a winter storm around the Turkey Day timeframe. But this is only scratching the surface of this storm's potential, and for that reason, we'll continue on with the investigation into this potential winter storm.

This is a forecast from the GFS Ensembles out to the Hour 168 timeframe. While all of the colors and lines may seem a little overwhelming, we're only going to focus our attention on East Asia, which is under the deep blue swath in the top left part of the left panel. Another correlation that has been proved effective is how weather in East Asia can be reciprocated in North America 6-10 days after the weather anomaly in East Asia. This includes the presence of storm systems or high pressure in East Asia. The November 18 forecast from both the GFS Ensembles and ECMWF model indicates that a deep negative height anomaly swath will slide down over Japan. If we extrapolate November 18th six to ten days out, we come up with a potential cold weather and storm system time frame of November 24 to 28. This narrows down the timeframe previously set out by the Bering Sea correlation, and if we match up the two timeframes (November 24-28 for the East Asian connection and November 25-29 for the Bering Sea connection), we come up with a November 24-29 potential timeframe for a storm, which can be isolated into the November 25-28 period, which is shown by both connections. Based on these two connections coming up with a common timeframe for a potential US storm system, confidence is rising that this event will happen.

So, we have tentatively established that there will be a Thanksgiving winter storm. Where would this storm go? Well, we can get a decent idea from not only the aforementioned teleconnection discussion, but also from a set of composites I outlined for the Midwest and Northeast for major storm systems.

Let's outline what we are expecting in terms of teleconnections for the time around Thanksgiving- more specifically, the November 24-29 period for when this storm could occur. The Arctic Oscillation is expected to be neutral or negative for this timeframe, meaning there may be cold air available for a winter storm and not just a regular storm system. The North Atlantic Oscillation should be in neutral or negative territory, which tells us the energy from this storm should be kept south in the Plains, following the probably-positive (or neutral) PNA, which ought to encourage this storm system to head into the Plains, Midwest or Ohio Valley. The EPO will most likely be negative, and this helps out the cold air idea that the negative Arctic Oscillation should try to establish for this storm system. So now that we've established that there should be a neutral/negative AO, a neutral/negative NAO, a neutral/positive PNA, and a negative EPO, let's take a look at some composites I made up for the Midwest for major winter storm systems and their relationships to teleconnections.


Legend on left shows number of storms.
By accounting for a handful of major winter storms in the Midwest, I was able to create a composite of when winter storms are most likely to occur in this region. By looking at a total of 17 major winter storms, I compiled data that indicated a negative NAO, positive PNA, negative EPO and negative AO were most favorable for Midwest winter storms. Do those phases of the teleconnections sound familiar? They should, because they match up nearly perfectly with the forecasted teleconnections around Thanksgiving. I also collected data from these 17 storms on the Madden Julian Oscillation, and that told me that an overwhelming percent of the sampled winter storms involved the MJO being in Phases 5-8, with emphasis on Phases 7-8. This composite doesn't match up as well as the teleconnection information, but as I showed in the MJO discussion earlier, Phase 1 still has a very wet pattern for November for the Midwest and Ohio Valley.

I do have snowstorm composites for the Northeast as well, but the teleconnection composites aren't nearly as definitive on which phase is more favorable as the Midwest composite is. What I did draw from the MJO Northeast winter storm composite, however, is that the latter MJO phases (Phases 7-2) are most favorable for Northeast winter storms. The negative NAO and favorable MJO composite would seem to lend a hand to the Northeast in terms of this winter storm potential, although right now, I personally see the Midwest, Plains and Ohio Valley at risk for this system. But as is with these long range projections, it's not set in stone.

WARNING: This graphic is unlikely to verify as is shown. I am showing it because the GFS may have caught on to this idea for one run, but I am NOT endorsing it in any way, other than just to point it out.
For what it's worth (and it's worth very little), the GFS model caught what could be this possible snowstorm on its 0z run, but has since dropped it. This may give you an idea of my current thoughts in terms of storm placement, but not necessarily the precipitation type.

Andrew