Long-Range ECMWF ENS Control Member Shows Frigid January; Will it Verify?

 The latest long-range ensembles from the European forecasting agency came in the other day, and while on the whole they produced a generally milder outlook, we did see a surprise forecast from the 'control' member of these ensembles - a remarkably frigid January.

Before we get into the good stuff, let's review what that means.

The ECMWF agency is known for their namesake weather model, but they also provide a critical forecasting tool with their ensemble system. In an ensemble forecast, an initial forecast model is taken, tweaked ever so slightly, and then re-ran. That's one ensemble 'member'. Then, they take that initial data again, tweak it ever so slightly in a different way, and re-run it again. That's another ensemble member, and it continues. The ECMWF's ensemble system has 51 different members - they are all different versions of the ECMWF weather model, but each one has slightly different starting conditions in order to produce a different forecast. 

You might ask why one would do that, since it sounds a lot like all we're doing is intentionally changing the initial conditions and letting it run free. That doesn't sound very productive, does it? In reality, the value comes from how similar all of these 51 members are at each point in time. For example, if all 51 ensemble members showed a winter storm in the Midwest ten days from today, that would be a big boost to forecast confidence. Why? Because if all 51 members have that storm happening, it means no matter what the weather conditions are right now (i.e. even if they're tweaked in all kinds of ways), we still end up with a winter storm on the horizon. As a result, ensemble forecasting is a potent part of any forecaster's toolbox.

But each ensemble set - whether it's the ECMWF ensembles, the American (GFS) ensembles, or any other set - has one member that is different from the rest. This member's initial conditions are the best possible data we have, and all the other members are those slightly-tweaked versions of this one. We call this special ensemble member the control. Indeed, the control member is the raw forecast taken strictly using initial data. Nothing is tweaked, nothing is altered, just the raw initial conditions. Every other ensemble member takes that control member data, tweaks it a bit, and re-runs it, like I described above. 

One may argue that this makes the control member more attention-worthy, since it only uses the best initial conditions data we have and isn't artificially tweaked. Whether that argument has merit is something I have yet to verify on my own.

In any event, now that we understand why an ensemble's 'control' member is worth monitoring, let's get down to the good stuff: the reason why this post was made.

Attached above from WeatherBell (all images on this post are from WeatherBell) is a look at the seven-day average temperature anomaly from the ECMWF ensemble control member over the December 25th - January 1st timeframe. We see that the control member anticipates a deep and formidable reservoir of below-normal temperatures in western Canada, which begins bleeding south and east into the United States as December turns to January. Indeed, seven-day average anomalies of more than 25 degrees Fahrenheit below normal are nothing to shake a stick at!

The outlook continues into mid-January below:

From January 11th - 18th, temperatures are shown by the ECMWF ensemble control member to be much colder than normal across almost all of the country, particularly east of the Front Range and especially in the Ohio Valley. With seven-day average temperatures seen at almost 20 degrees F below normal in parts of Ohio, Indiana and Kentucky, it's clear that the ECMWF Control member expects the Arctic gates to burst open in the first month of the new year.

Teleconnections accordingly show a breakdown in the tropospheric polar vortex as judged by a collapse in the Arctic Oscillation in the blue line (control member's forecast)...

... and a similar plunge in the North Atlantic Oscillation (NAO) from the control member allows that frigid Canadian cold to plow eastwards...

Now, that's all good and fun to look at, but this post wasn't written to get all the cold weather fans out there hyped up. Indeed, as the above two teleconnection charts show, the control member is an extreme outlier in the late-December and early-January period, and the average of all ensemble members just barely turns the NAO negative at the end of this month. The mean doesn't even turn negative at all for the Arctic Oscillation!

My takeaway is this: we're seeing some noteworthy cold signals for the end of December and start of January from the ECMWF control and the CFS ensemble system (a different analysis of that coming later today), but we can't rely on models that far out. To me, it's just not smart forecasting to use them for anything aside from large grains of salt, if that makes sense. Instead, I'm focusing on a changing global weather regime as the Madden-Julian Oscillation changes into colder phases for the end of December and through January. That, I feel, is where the real intrigue lies for Arctic outbreak risks as we enter the new year.

Stay tuned for that aforementioned post later today.

To Summarize:

• The ECMWF ensemble's 'control' member - often seen as the most accurate member of the ensemble group - is expecting much-below-normal temperatures from the end of December through much of January in the U.S.
• At this stage, while it's intriguing, I would not take this signal as anything more than a large grain of salt
• There is more value in monitoring the changing global pattern as the new year kicks off, and a post discussing that will be published a little later today.

Andrew

Long Range Outlook (Made June 9, 2019)

This is a Long Range Outlook made for publication on June 9th. The forecast period will cover the remainder of June. Click on any image to enlarge it.

Observed 500-millibar geopotential height anomalies as of 1pm Central, June 8th.
Source: Tropical Tidbits

A look around the Northern Hemisphere on the afternoon of June 8th shows a rather-busy pattern over North America, with an upper level low positioned over northwest Canada and a secondary trough dropping into the Pacific Northwest while still being influenced by the first system. A strong ridge maximized over the northern Great Lakes into southern Canada contrasts with a closed low positioned in the Southeast U.S., forming a Rex Block-looking pattern despite there not being any actual blocking event taking place. Indeed, even downstream of this pattern into Newfoundland, the flow is rather amplified with an upper level low grinding against a strong ridge that is firmly entrenched into Greenland. Upstream of the United States, weak ridging is evident in the northeast Pacific, helping to diffuse the Pacific jet stream in the process. A strong storm system is seen south of the Aleutian Islands, with additional disturbed weather skirting along the northern Pacific.

Forecasted 500-millibar geopotential height anomalies valid 7am Central, June 12th.
Source: Tropical Tidbits

Even within the five-day forecast window, the evolution of the flow over the Northern Hemisphere becomes rather complex rather quickly. The ridge and closed low mentioned in the eastern half of North America are seen drifting eastward, with model guidance suggesting the ridge will propagate northeastward into Greenland to maximize positive 500mb height anomalies. Additionally, the observed weak ridging in the northeast Pacific is expected to strengthen and entrench itself across the western coast of the continent, thereby setting up a pattern representative of the positive phase of the Pacific-North American (PNA) oscillation, seen by ridging along the West Coast, as well as the negative phase of the North Atlantic Oscillation (NAO), seen by ridging over Greenland.

The result of these two features, as well as a messy picture for the West Pacific Oscillation and East Pacific Oscillation (EPO), is a likelihood of cooler than normal temperatures across the eastern two-thirds of the United States, as well as into Canada. The positive PNA and negative NAO will act to direct air from northern Canada down into the United States. Indeed, the five-day average 2-meters-above-the-surface temperature anomalies below, through the period ending June 14th, shows a substantial and widespread period of below-normal temperatures for the majority of the country.

Forecasted 5-day average ~surface temperature anomalies, valid through the period ending 1pm June 14th.
Source: Tropical Tidbits

Model guidance is showing strong confidence in the evolution of strong ridging over both western North America and Greenland as well:

NCEP Relative Measure of Predictability for 500-millibar values at forecast hour 96 (June 11th 7pm).
Source: NCEP

Given the high confidence in the projected pattern through the five-day forecast timeframe, it does appear that colder-than-normal temperatures are likely for the eastern two-thirds of the country through at least June 14th.

Forecasted 500-millibar geopotential height anomalies valid 7am Central, June 18th.
Source: Tropical Tidbits

Ensemble model guidance is in agreement that the strong ridge over Greenland (again, the negative-NAO pattern) will remain in place through at least the day 10 forecast period, as shown above by the ECMWF (European model) ensembles. Upstream of the contiguous United States, too, model guidance does not anticipate any significant changes in the aggregate pattern, with predictions of ridging continuing along the west coast of North America as an upper level low situates itself just south of Alaska.

Let's now turn to teleconnections and oscillations to ascertain if model guidance is correct in depicting the forecast into the end of June.

Forecasted states of the Pacific-North American (PNA) index, top-left; North Atlantic Oscillation (NAO), top-right; Western Pacific Oscillation (WPO), bottom-left; and the Eastern Pacific Oscillation (EPO), bottom-right.
Source: Earth System Research Laboratory (ESRL)

Viewing model-forecasted teleconnections over the next two weeks or so, we find a broad pattern similar to what was described earlier, with the positive PNA and negative NAO expected to dominate the outlook for North American weather throughout the forecast period. Guidance does weaken the PNA significantly from a strong-positive state to a slight-negative state by June 20th. I personally don't see this as remarkably likely when compared to a forecast that has the PNA remaining positive (even if modestly so) through the period, and I'll discuss why shortly.

As noted earlier, the WPO and EPO are not expected to be strong enough to materially disrupt the +PNA/-NAO combination, with the moderately-positive West Pacific Oscillation forecast actually a boon for helping to sustain the ridge along the western coast of North America, and thereby sustain the +PNA. Model guidance sees the EPO as hovering essentially around levels too weak to have a definitive impact on the weather pattern.

I'd like to briefly discuss a second reason - in addition to the expected positive WPO - why I believe the ridge along the western coast of North America (the +PNA) will sustain longer than guidance indicates here.

Madden-Julian Oscillation (MJO) phase space diagram forecast, from the ECMWF model.
Source: Climate Prediction Center and ensembles.

For a refresher on the concept of the Madden-Julian Oscillation and how to dissect a phase-space diagram, feel free to read this post from earlier. The European model and its ensembles anticipate the MJO staying in a low-grade Phase 3 state over the short-term, before shifting into Phase 4, strengthening somewhat, and then potentially weakening into the 'circle of death'. Given how model guidance in that linked post had expected us to be in the 'circle of death' at the time of this writing, rather than in Phase 3, we will only trust the forecast to take us from a low-grade Phase 3 to a low-grade Phase 4 through the end of June.

We are able to see how the atmosphere typically acts when the MJO is in Phase 4 for the month of June below.

500-millibar geopotential height anomalies when the MJO is in Phase 4 during the month of June.
Source: American Weather

When the MJO is in Phase 4 during the month of June, positive height anomalies tend to be favored over the Gulf of Alaska into parts of the Aleutian Islands, with another area of positive height anomalies over Greenland. Below-normal anomalies are then painted across a good swath of the eastern two-thirds of North America.

If you're thinking that this Phase 4 composite above looks remarkably similar to the forecasted pattern beginning around June 11th, as discussed earlier in this post, you aren't the only one. It doesn't seem coincidental to me that model guidance maintains the +PNA/-NAO pattern through much of the remainder of June, particularly since the MJO moving into Phase 4 would support exactly this kind of pattern to continue.
It is possible that guidance thinks the PNA will weaken and even reverse into a negative state by late June simply because the models are - correctly - trained to forecast the state of a teleconnection by comparing the forecasted pattern with a "textbook" pattern. For example, scientists have developed what the atmosphere should look like when the PNA is said to be positive, and thus models compare their forecast to that "ideal" pattern to determine what state the PNA will be in. However, the atmosphere is not binary like that - in other words, even though the PNA might be forecasted to be negative, the pattern itself may still resemble a positive PNA. Let's discuss this further.

Forecasted 500-millibar geopotential height anomalies valid 7am Central, June 21st, with annotations.
Source of image: Tropical Tidbits
Source of annotations: Author

Shown above is the 500-millibar height anomaly forecast from the GFS Ensembles, valid the morning of June 21st. This is the purported date by which the PNA is forecasted to turn negative, as per the four-panel graphic examined earlier in this article. Looking at the broad forecasted pattern, however, shows the pitfalls that simply taking the index forecast & going with it can present.

First, we recognize why guidance thinks the PNA will flip from strong-positive to a negative state: there is an expectation that troughing will develop along and off the coast of western North America, a part of which I have circled in red. To be sure, stormy activity in this region is indeed what a negative-PNA consists of, so the models are not at fault here. The issue we want to resolve is a lack of context that looking only at forecasts of the indexes results in.

When broadening out beyond the trough along the western coast of the continent, the ridge that is expected to send the PNA into strong positive territory is seen, now having retrograded slightly into the Gulf of Alaska. The movement of the ridge to the west is indeed why troughing is seen developing along the western coast. But take a look at the way the ridge extends into Alaska to see what I mean by 'lack of context'. Instead of merely topping off around Anchorage, the positive anomalies extend well into Alaska, and join up with another ridge based in Canada, the combination of which then pushes into the Arctic Circle. That doesn't mean much for the trough along the west coast of the continent, but it's certainly something we need to note, and here's why.

Take a look at Greenland again. The strong ridge is still forecasted to be positioned over Greenland, keeping the negative-NAO in place, as expected by the four-panel teleconnection graphic from above. Big deal, so what? So, the negative NAO pattern keeps colder than normal conditions in place over the Northeast U.S., Great Lakes and into the Mid-Atlantic. But with that strong ridge in Alaska and Canada, I find it more plausible that the lobe of colder weather will also affect the Midwest, parts of the Plains, and the Ohio Valley. Why? That ridge in western Canada will team up with the Greenland ridge to force the primary lobe of cold air to the south, as is already seen in the graphic above.

But it's the fact that the Gulf of Alaska ridge extends its influence into the Canadian ridge that makes me think those regions will also be affected by cold air. The alignment of the Gulf of Alaska ridge suggests it shouldn't be difficult at all for that ridge to push northeastward into Canada, where the other ridge is based. If/when this happens, the trough along the western coast of North America will be suppressed, and perhaps even displaced to the south towards the Southwest U.S., which could then allow the ridge to build back into the southwest part of Canada or even into the Pacific Northwest again.

All in all, that would allow a more positive-PNA-esque pattern to emerge (with respect to cold air hitting the mid-section of the country and not just the Northeast), even though by definition we should be in a neutral or even negative-PNA state, since that trough will still likely stick around, just a little farther south.

Whether this scenario actually plays out is something only time will tell, of course, but I wanted to emphasize this because when looking at teleconnection forecasts, you need to also know the *context*. I have gotten burned plenty of times by only going off the teleconnection forecasts (for example, only seeing a -NAO or +PNA and basing the forecast off that) and not understanding the context for that forecast (for example, the +PNA might not be as sturdy as the teleconnection forecast shows because of a strong upper level low just to the west of North America, or something along those lines).

---

All told, this provides a good basis for my outlook for the remainder of June.

- Through the middle of June, I expect colder than normal conditions to evolve for the eastern two-thirds of the country. This cooler weather will be widespread, with areas only west of the Front Range likely to see seasonal to above-normal temperatures. In terms of precipitation, any severe weather threats should stay confined to the southern portion of the country, with the jet stream and broader pattern alignment unfavorable for any severe weather outbreaks at this time.

- From the middle of June to the end of June, I expect a continuation of this cooler than normal pattern, though with a reprieve coming for the Plains and perhaps parts of the Midwest as a trough builds in to parts of western North America to return some parts of the central U.S. to more seasonal conditions. However, firmly cooler than normal conditions should persist for the eastern third of the country, especially in the Northeast. Any seasonal conditions will be pockmarked by cooler bouts of weather for those in the Midwest and Great Lakes into the Ohio Valley, though again the coldest conditions should stay further east. Storm chances should increase for the central part of the country with a return to more seasonal weather, but zonal flow aloft and the lack of a ridge in the Southeast should dissuade any significant severe threats from hitting the Central U.S., and likely from hitting the country in general.

Andrew

Special Long Range Outlook: Early to Mid June

This is a follow-up on my previous long range outlook posted last weekend (click here to read). In this post, I'll be using convective activity across the Equator as a basis for the forecast, and then consolidate that with the long range outlook previously published. Click on any image to enlarge it.

**If you do not wish to read this (admittedly technical) discussion, please scroll down until it says to stop.**

Let's first review the concept of the Madden-Julian Oscillation (MJO), as this will prove critical to the rest of this article. The MJO is an oscillation that is based on the location of convective activity along the Equator. Depending on where this convection is, the MJO is said to be in one of eight phases.

Portrayal of enhanced (suppressed) convection around the Equator, shown as green/blue (brown) colors, for each of the eight phases of the Madden Julian Oscillation.
Source: NOAA

In a generalized sense, Phase 1 of the MJO typically sees enhanced convection along the Equator from South America into portions of Africa. In Phase 2, this enhanced convection shifts east into the Arabian Sea and marginally into the Bay of Bengal, with storms most focused directly south of the subcontinent of India. Drier than normal conditions are then normally seen north and east of Australian into Oceania. Phase 3 again sees an eastward shift of enhanced convection, now firmly into the Bay of Bengal and the broader majority of the Indian Ocean, more focused on the eastern half of that body of water. Additionally, we see further strengthening of that convection, maximized just southeast of India. Phase 4 sees the enhanced convection move into the Timor Sea and Arafura Sea. By Phase 5, the convective activity begins entering the far western Pacific Ocean, with calmer than normal conditions developing in portions of the Indian Ocean. Australia also experiences the enhanced convective activity in Phase 5, mainly over the northern section of the country. Phase 6 shifts the enhanced convection further east, and so on through Phase 7 and Phase 8.

However, a number of scientists have put in long hours of hard work to focus in on how these areas of enhanced and depressed convection vary, if at all, in location and intensity depending on the time of year. The Bureau of Meteorology in Australia has created these composites over three-month windows - for our purposes, we will use the May-June-July window so as to be centered over June.

Portrayal of enhanced (suppressed) convection around the Equator, shown as cooler (warmer) colors, for each of the eight phases of the Madden Julian Oscillation during the May-June-July period.
Source: Bureau of Meteorology

 The location of areas of enhanced and suppressed convection are about the same as in the general composite graphic shown first. However, particularly in meteorology, more data and more accuracy is almost always a positive when making a forecast. As a consequence, I will refer to this image later on in this post, as opposed to the first image.

It's reasonable to be skeptical as to why we should care about the location and intensity of convective activity along the Equator. I can imagine some of you are wondering how some showers and thunderstorms thousands of miles from you and your computer screen can possibly have any substantial impact on the weather pattern, much less be a driver of the pattern as a whole. This article will aim to show why we care about the MJO, and how we are able to use those showers and thunderstorms to make reasonable forecasts for the United States almost a month in advance.

Now that we've reviewed what the MJO actually is and how we can determine which of the eight phases it is in, let's see which phase the MJO is currently in so we can begin building this outlook.

Madden-Julian Oscillation (MJO) phase space diagram forecast, from the ECMWF model and ensembles.
Source: Climate Prediction Center

The above graphic might appear intimidating, but in reality it's rather simple. The diagram itself is called a phase space diagram, and it uses two different mathematical models that analyze the concentration of clouds and the behavior of winds at different levels of the atmosphere to diagnose what phase the MJO is in, and how strong it is within that given phase. Those two different models are labeled RMM1 and RMM2, as shown by the 'x' and 'y' axes of the graph. In other words, they help create the graph to tell us what the MJO is doing.
Reading the graph is relatively simple. The eight different phases are clearly marked, with locations typed out at different parts of the chart to indicate where enhanced convection is located when the MJO is in a given phase. You can confirm this by comparing the phase numbers and location names to the composite image(s) discussed earlier in this article. The graph also gives an idea of the strength of the MJO - the further it is from the circle in the center, the stronger the MJO is within any given phase. If the MJO is shown to be in that middle circle, it means the oscillation is too weak to definitively assign a phase to, and thus may not be a primary driver of the atmospheric pattern at that time. Luckily for us, model guidance has the MJO as being outside of that middle circle for the entire forecast period.

Viewing this phase space diagram, we can see that the MJO was actually inside of that middle circle during the middle of April, just over a month ago. From there, however, the oscillation seemed to come to life and strengthened into Phase 2 by late April. From there, through May up to this point, enhanced convective activity propagated eastward from the Indian Ocean, bringing the MJO through phases 3 through 8, where it now sits. Indeed, the MJO has been positioned in phase 8 for about ten days now.

This particular set of model guidance, from the European model (the ECMWF) and its ensembles, sees the MJO shifting into Phase 1 to wrap up May, and then transitioning back to Phase 2 in time for early June. Indeed, this guidance even sees convection moving eastward enough to set up in Phase 3 by about June 6th, though at that point discrepancies among ensemble members begin to materially degrade the quality of the forecast, not to mention those members have the MJO wobbling close to that circle in the center.
Of course, there is more than just one model that tracks the MJO this way:

Madden-Julian Oscillation (MJO) phase space diagram forecast, from the CFS model and ensembles.
Source: Climate Prediction Center

This graph uses the same chart set-up and parameters as the one before, although this time instead of the ECMWF model and its ensembles, the CFS model and its ensembles are displayed. Even so, the CFS takes the MJO through almost an identical path as the ECMWF during the same time period, dragging the oscillation through Phases 1 and 2 to kick off June before trying to move towards Phase 3 by June 6th.

This concept of the MJO moving into Phase 2 to begin June, after traversing Phase 1 to close out May, is more-or-less the consensus forecast across available models. As such, we will incorporate a Phase 2 MJO state into this outlook for the first five days of June or so. Now that we know what the MJO should do for the first several days of June, what does that translate to in terms of the weather pattern here in the United States?

More hard work by scientists has allowed us to have answers to that question.

500 millibar geopotential height anomalies as observed during a Phase 2 MJO event in June.
Source: Japan Meteorological Agency

The Japan Meteorological Agency (JMA) not only narrows down the typical 500 millibar geopotential height anomalies, 2m temperature, 200mb wind speed etc. patterns that occur by MJO phase, but also divides these into individual months. As a result, we are able to view the typical 500 millibar height anomalies seen when the MJO enters Phase 2 during the month of June. In the top image, cooler (warmer) colors represent below-normal (above-normal) height anomalies, which generally correspond to troughs and colder weather (ridges and warmer weather). We don't need to really worry about the bottom panel; it shows the outgoing long wave radiation (OLR) anomalies for a Phase 2 MJO event in June. It might sound complicated, but it's really the same thing as analyzed in the first two graphics of this post: in that bottom panel, cooler (warmer) colors represent areas of enhanced (suppressed) convection. We already know from earlier that a Phase 2 MJO event corresponds to enhanced convection in the Indian Ocean region, and this bottom panel is just reiterating that.

Back to the top panel in this graphic, it appears that the MJO entering Phase 2 in the month of June tends to bring about stormier than normal activity in Russia and northeastern China, while maintaining a tendency towards ridging over Japan and southeast China. On the other side of the Pacific, a June Phase 2 MJO event encourages a ridge well offshore of the West Coast and southwest of the Gulf of Alaska, while stormier/cooler than normal activity is predominant in most of the Western U.S. Eagle-eyed readers will recognize the positioning of that stormier activity as corresponding to the negative phase of the Pacific-North American (PNA) index (you can read more about the PNA by clicking here). As such, it's not surprising that the Central and Eastern U.S. then experiences a tendency for ridging and warmer weather during this state of the MJO for the month of June.
In other words, when the MJO moves into Phase 2 during the month of June, the western U.S. tends to be stormier and cooler than normal, while the central and eastern U.S. will have a tendency in favor of ridging and generally warmer weather.

But wait a second - that sounds pretty familiar to the forecast I published this past weekend, with cooler weather in the West, warmer weather in the East and storms riding a ridge through the central part of the country, doesn't it?  That's not by coincidence - if you've followed this blog over the years, you'll find that these articles have often made these kinds of confirming conclusions between teleconnections and oscillations, enabling these long-range outlooks to have at least a bit of justification behind them as opposed to throwing darts in the dark. Whether that justification ever ends up being accurate, though, is always a question mark!

Indeed, long-range models see a pattern not dissimilar to that Phase 2 MJO composite evolving for the last bit of May and first bit of June:

Forecasted 500 millibar geopotential height anomalies from the ECMWF model (left), GFS model (center) and CMC model (right), all valid for the period from May 31st to June 2nd.
Source: Pennsylvania State University

Using these three models over the May 31 - June 2 forecast period, we can ascertain the general idea of troughing in the western U.S., signifying stormier weather, as well as a nascent ridge trying to form in the southern U.S. Note, however, that our forecast here seems to be scuttled by that large lobe of below-normal height anomalies and associated cooler weather in Canada. This seems to go right up against that MJO Phase 2 chart we were just looking at. What happened?

We have to go back to something I briefly touched on at the beginning of this post to understand why model guidance isn't showing what we expected. When analyzing the structure of the MJO phase space chart, I went over the circle in the middle, where an MJO phase can't be determined definitely because the oscillation is too weak at that point in time. Consequentially, the MJO may not be a primary driver of the pattern at that time. What we must now remember is that the MJO may not be a primary driver of the pattern at this time, even if it is in a defined phase. The atmosphere doesn't abide by any single oscillation or teleconnection, it is all one big puzzle.

In this case, models don't see North America getting a typical Phase 2 set-up because of Rossby waves sending much above-normal air into the Arctic Circle by way of those stout ridges over Eurasia and even potentially north of Japan. This forces the tropospheric polar vortex - already significantly weakened in the summer months, to be sure - to lower latitudes, as happens in the winter. Adding to that, we see a strong ridge positioned near Greenland (negative NAO pattern), which encourages cold weather in the upper latitudes to flow down to lower latitudes - in particular, the United States. These two factors (and likely more) combine to keep that swath of negative height anomalies locked out in Canada, bringing about that "unexpected" outcome. Additionally, during this forecast period, the MJO will be transitioning from Phase 1 to Phase 2, which makes it a little messy to use composite images. In any event, we now have three things to use in our forecast for early June:

- Phase 2 of the MJO
- Continuation of strong ridging over the Arctic Circle
- The negative phase of the North Atlantic Oscillation (NAO)

We'll make an actual forecast out of those factors at the end of this post. For now, let's move on to the mid-June outlook.

I'll begin by using another scary-looking chart which, in reality, isn't too tough to read.

Hovmoller of outgoing long wave radiation (OLR) between the latitudes of 5º North and 5º South, forecast period from May 22nd to June 19th, via the CFS model.
Source: North Carolina State University / Carl Schreck

First and foremost, let's break down what this chart shows.

For all areas above the solid black line positioned next to the May 22nd marker, all variables that will be described next are observed. For all areas below that solid black line, all variables are forecasted.
Next, we need to recognize that the solid colors here are merely showing the same thing we've looked at three times now: convective activity. As you might guess by now, green (brown) colors indicate the presence of enhanced (suppressed) convection.
The third thing to note about this chart is that we will not be paying attention to any of those red, pink and blue ovals and shapes - they indicate different types of waves and are juxtaposed with different bursts and drops in convective activity, depending on what kind of wave it is, but for our purposes we will disregard all that so it doesn't get too confusing.
The last thing to note about this chart so we can begin our forecast is that we do care about the black oval shapes. Remember back at the beginning of this post, when I explained that the MJO is an oscillation based on the location of enhanced convection along the Equator? Well, in this chart, those green areas show those same enhanced spots of convection at those same locations (see the longitude markers on the x axis) and also along the Equator (remember this chart shows convective activity between 5º North and 5º South latitude). The black ovals here are simply designed to highlight MJO waves (think of how the enhanced convection is a "wave" slowly moving east across the Indian and Pacific oceans).

That's about all that's critical to understanding that very-messy chart: knowing that part of the chart shows observed conditions and part of it shows forecasted conditions, knowing that the shaded colors show enhanced or suppressed areas of convection just like we've already gone over in other graphics, knowing that we can disregard the colored ovals and shapes, and knowing that the black ovals merely point out to you where the MJO wave is.

----

So, why do we care about it anyway? Whereas the forecasts of the MJO made using the phase space graphics only go out to June 6th, the CFS Hovmoller (name of this chart type) above goes out into late June, meaning we can use it for our mid-June outlook.

Let's start out at the June 5th marker on that chart. On June 5th, per this forecast, the MJO wave is forecasted to reach about the 85º East longitude region. If we refer back to the second graphic in this post, we can see that enhanced convection around 85ª - 90º East corresponds to the MJO being in Phase 2, nearing Phase 3. That lines up well with what the phase space forecasts anticipate by this time.

Beyond that, however, we run into some trouble. The forecast here has that convection, and the MJO wave as a whole, dissipating by June 12th, with the enhanced convection anomalies vanishing even before that, around June 10th. That poses a dilemma for us, as it means this model sees the MJO entering that center circle if we were to view this using a phase space diagram.

It seems like we're at the end of the road here. Of course, that's not the case.

Forecasted 200mb Velocity Potential for 1pm June 5th, using the GFS Ensembles.
Source: Tropical Tidbits

One way to identify the location of enhanced convection to track the MJO is by viewing anomalies in convective activity themselves, as we have done up to this point. Another way, however, is by viewing the upper air pattern to identify areas that are conducive for thunderstorm development. Recall that strong thunderstorms will build near the surface as air converges and rises, with the clouds of that storm then plateauing and spreading out high up in the sky when the air in those clouds becomes as cool as the environment, leading to the air high up in the sky to spread out and diverge. This is essentially what we're looking at here, only in this chart we're looking at those convergence/divergence motions only at the 200 millibar level (cruising altitude for most airplanes).
Green shading and associated spreading-outwards arrows indicates divergence aloft. This is a plus for thunderstorm formation, as it encourages air lower down to converge and create thunderstorms. Orange shading and associated spreading-inwards arrows indicates convergence aloft, which piles air into the column and suppresses thunderstorm formation. As such, you can think of green shaded regions as being areas where thunderstorm development is encouraged, and brown shaded areas where it is discouraged.

In this forecast for June 5th, ensemble guidance sees strong divergence aloft over Latin America and between the 120º West and 60º West longitude lines. Additionally, we see some divergence aloft over Africa. Remember that this time period is when the MJO wave is dying out and we expect it to enter that center circle in the phase space diagram, where the MJO is technically too weak to be sorted into a phase.
Of course, the atmosphere could care less about what a bunch of humans say about phases - just because the MJO is seen as too weak to define doesn't mean the atmosphere will shut down convection, and certainly doesn't mean that convection in certain areas won't impact the U.S. until the MJO re-emerges into a given phase. So, since we see convection-favorable dynamics over Latin America and Africa at this timeframe, even though the MJO is technically now too weak to track, we are going to go into the mid-June forecast presuming that the pattern will reflect a Phase 1 MJO state, because the convection-favorable dynamics will encourage storms to form over Phase 1 locations. Again, the MJO will technically be too weak to track, but that definitely doesn't mean convection over the Phase 1 area won't produce Phase 1-like conditions.

So, what happens in a Phase 1 MJO event in June?

500 millibar geopotential height anomalies as observed during a Phase 1 MJO event in June.
Source: Japan Meteorological Agency

A Phase 1 MJO state during June typically sees a strong upper level low positioned over the Bering Sea and Aleutian Islands of Alaska, with weak ridging from Hawaii towards the Pacific Northwest and a corresponding tendency for stormier weather in the West and South United States. Some ridging is seen in New England.

200 millibar zonal wind anomalies as observed during a Phase 1 MJO event in June.
Source: Japan Meteorological Agency

What I find more interesting, however, is what typically happens to the jet stream during a Phase 1 MJO event in June. Looking at the north Pacific, we see quite a few contours with pink shading stretching from Japan all the way into the Gulf of Alaska. Those high positive contours and elongation from Japan to the Gulf of Alaska signal a strengthening and extension of the Pacific jet stream, a feature that acts to strengthen low pressure systems which drop into the Western U.S. on the back of this jet stream. I referenced the impact an extended Pacific jet can have when discussing the high severe weather threat earlier this week (click here).

Of course, we already saw earlier how the MJO is only one factor in a forecast, and that model guidance can (and often does) reflect this combination of teleconnections and oscillations in a way that is far from any one oscillation's "textbook" composite output (i.e. the composite for Phase 2 that we ran into trouble with earlier). Model guidance again is showing a different outcome by mid-June:

Forecasted 500 millibar geopotential height anomalies (left) and "spaghetti" plot of all members (right) from the GFS Ensembles, valid 1pm on June 8th.
Source: Pennsylvania State University

This is indeed a 384-hour model forecast, one of the big "no-no's" when it comes to making a forecast. While I agree that one should never base a forecast off of 16-day model guidance, I do believe that even these ultra-long-range model outputs can provide valuable hints about what the overall pattern will eventually be.

The GFS ensembles expect the pattern over the Arctic Circle to not vary too much by June 8th, with strong ridging still forecasted over Eurasia into the Arctic Circle, as well as a stout ridge now moving into Canada. Still, we do see some hints of that Phase 1-esque pattern, with tightened isohypses south of the Bering Sea seeming to signal a stronger Pacific jet stream and a valley in isohypses over the Western U.S. reflecting troughs still positioned over that area.

**If you scrolled past the technical discussion, STOP HERE.**

By now I've either put you through a lot of pain reading this, or you've gotten a satisfactory fill of technical forecasting for the day. You may also have decided to skip the technical discussion altogether. In any event, it's time to put these pieces together and make the forecast.

For the first several days of June, I expect a continuation of the current pattern. Strong ridging in the upper latitudes will keep the upper-level low over Canada in place and keep northern portions of the United States cooler than normal. Storm systems will continue to be deposited in the Western U.S., which will lead to a ridge to build in the East U.S. and allow those storm systems to ride northeast along the ridge, bringing repeated severe weather threats to the Central U.S. The presence of these troughs in the West will maintain predominantly cooler than normal and wetter than normal conditions. The pattern for the Southeast should remain relatively quiet, with this broad set-up not favorable for significant weather phenomena (i.e. big severe weather outbreaks).

From about June 8th through June 15th, I expect marginal shifts in the pattern, but for the broad North American set-up to be relatively unchanged. Ridging is preliminarily expected to continue in the Arctic Circle, which will keep the Northern U.S. under the gun for cooler than normal conditions. An extension and strengthening of the Pacific jet stream, especially relative to the weakened and highly-meridional flow seen to end May, seems plausible in this timeframe, which will then have the potential to re-introduce higher-end severe weather threats. At this point, the ridge in the Southeast should become at least a little less stout, which may allow more severe weather threats in the eastern-third of the country and perhaps give a reprieve to the Central U.S. Broadly-seasonal temperatures may be expected outside of the North and parts of the West (cooler than normal), but this could change if that ridge projected over Canada for June 8th becomes dominant and establishes a well-above normal temperature pattern for the central part of the country.

The longer-range forecast (i.e. beyond June 8th) will undoubtedly change, but hopefully this article gave you a look at how we can in fact use thunderstorms in the Equatorial Pacific to forecast the weather here many days in advance!

Andrew

Significant Arctic Outbreak Threat Rising for Late January

Model guidance is combining with the long range tools commonly used on this blog (namely the Typhoon Rule) to portray the country's best shot at true wintry weather to end the month of January.

We'll start off with the Typhoon Rule.

Tropical Tidbits
Click on any image to enlarge

The above image shows the GFS forecasted 500mb geopotential height anomaly forecast in the West Pacific region, valid for the morning of January 17th. In this image, we see warm colors, indicative of positive height anomalies / typically warm and quiet weather, but we see a much different picture over Japan. In Japan, a strong upper level low is impacting the country, sliding to the east as it does so. Notice how the center of this upper level low appears to go directly over far northern Japan.

The Typhoon Rule can be used to predict weather phenomenon in the United States. The rule of thumb indicates that weather phenomena impacting Japan is reciprocated in the US about 6-10 days later. So, if we see a strong (and very cold) upper level low crossing northern Japan on January 17th, we can extrapolate this upper level low to "re-appear" in the US on a January 23-27th timeframe.

Does this solution have any support from ensemble guidance? Let's find out.

Weatheronline

The image above shows 500mb geopotential height values (not anomalies) from the prestigious ECMWF ensembles set, valid on January 25th. These types of charts are generally used to identify the presence of large-scale ridges or troughs. In this image, we see a very pronounced depression in the contour lines, bringing the 520gpdm line all the way down just south of the Canada/USA border. This depression of contours shows the presence of a strong upper level low, which is actually centered all the way up in northern Canada.

The strong upper level low is being forced southward due to that strong ridge of high pressure developing along the West Coast. Other guidance eventually does much bigger things with that ridge even further down the road, but we'll get to that a little later on in this post. For now, the takeaway from this forecast is very cold conditions might be on deck for the late January period. I have discussed this end-of-the-month timeframe for some cold weather in earlier posts, but its significance is quickly becoming more realized.

ESRL

Next up, we'll analyze the ensembles off of the ESRL agency, a physics-based modeling branch of the NOAA body. In this image, showing 500mb geopotential anomalies for January 26th, we see a similar layout as the ECMWF ensembles showed. Strong negative height anomalies are developing in the Central and East US as the ridge along the West Coast continues to build and push northward. In these sorts of situations, the ridge may force itself so far north that it becomes a 'blocking' mechanism.

What is 'blocking'? My favorite example is to imagine a highway, with traffic moving along at an even pace. That's a good representation of the atmosphere during 'normal' flow. Now, imagine something happens on the highway that forces the cars to stop - a back-up, perhaps. The cause of this 'back-up' is analogous to the blocking ridge of high pressure; the ridge forces itself towards the North Pole and blocks the atmospheric flow from moving things along. This has been known to produce flooding, long-lasting cold, or intense warmth, depending on the season and who is affected.

ESRL

The upcoming pattern is very similar to that shown by the negative phase of the West Pacific Oscillation (WPO). In the image above, we see typical temperature (bottom panel) and 500mb geopotential height anomaly (top panel) values for a positive WPO phase. Notice how stormy conditions persist just west of the Bering Sea during a +WPO phase, leading to warmth across most of the Central and East US.

To see the typical conditions during a negative WPO phase, just flip the color scale. We then see intense ridging just to the west of the Bering Sea, leading to sustained cold flooding the Central and East US. The eventual alignment of this ridge into the waters near the Bering Sea, as the GFS ensembles are indicating, would lead to this -WPO pattern.

The final (and what I consider the most surprising) part we have to go over is the analog forecast.

CPC

This image shows projected 500mb geopotential height anomalies, valid for about 11 days from today, based on the top ten analogs (dates with the atmospheric pattern similar to the one forecasted to occur 11 days from today) produced at the Climate Prediction Center. In this image, we can see that strong ridge along the West Coast of North America into Alaska, resulting in below-normal anomalies over the Central and East US, eastern Canada, and into Greenland. Just for kicks, let's see what the top analog of 20090118 (January 18, 2009) shows, since it has been deemed the most similar to the forecast down the road.

ESRL

Above, we see temperature anomalies from January 16, 2009. I went back two days as the cold wave had already pushed east by January 18th; we want to get a diagnosis of the cold wave itself. This graphic is in units of Kelvins, but we can easily convert to Fahrenheit.

The core of the cold extends from southeastern Minnesota into far western West Virginia. A spot of values below -17.5 degrees below normal (Kelvin) shows up along the border of Indiana and Ohio. Let's see how -17.5 degrees below normal in Kelvins translates to Fahrenheit. Doing the math, that area saw temperatures 32 degrees (Fahrenheit) below normal for this cold wave, deemed the most similar to what we could see to end January. That's very impressive; for a scale of how impressive it is, consider that the average temperature in Richmond, Indiana (very close, if not inside that -17.5 Kelvin anomaly) to end January is 37 to 38 degrees F. In this cold snap, temperatures likely dipped down to just above zero, with even colder conditions to the north.

By no means does this mean we'll see something as cold as what happened on January 16, 2009, but it gives you an idea of the type of pattern that's coming down the pipe to end January.

To summarize:

- Model guidance and the Typhoon Rule are in agreement concerning a potentially significant cold blast at the end of January.

Andrew

Christmas Eve Potentially Significant Winter Storm

I'm watching the increasing potential for a significant winter storm in the Central US, created by what could be one of the strongest low pressure systems in the last few years, if not longer.

Instant Weather Maps

The above image shows 500mb geopotential height contours over the United States from the prestigious ECMWF model, valid for December 23rd. In this image, we see two pieces of energy in the Central US. One is a closed low, placed in the Northern Plains, while the other is a deepening trough in the Southern Plains, shown by the depression in contour lines. In a situation similar to that I described in my December 10th post, we see this trough begin to lift northward into the Central US, but not before the "bomb" goes off...

Instant Weather Maps

Just 24 hours later, on the morning of Christmas Eve, we find that the two pieces of energy have combined, and the trough as a whole has now attained a negative tilt, indicating it has reached its mature phase. As a result, the storm undergoes rapid strengthening, very near the criteria of bombogenesis. Bombogenesis is a meteorological term used to describe the phenomenon when extratropical cyclones rapidly strengthen, and their minimum central pressure values decrease by 24 millibars in a 24 hour period. Switching between December 23rd and 24th, we find that pressure values from the ECMWF lower by about 19.3 millibars, only a little ways off from being a true 'bomb'. Even though this storm doesn't fit the criteria, the strengthening is nothing to shake a stick at.

This solution is a tricky one. I don't have access to pay-to-view weather model graphics (yet), so I cannot see the precipitation pattern from this system. However, from what others are discussing, it appears that not only will the nearly-due-northward movement of this system foul up the precipitation shield, but temperature profiles are above freezing in many spots that would otherwise see snow. This could be placed on model error, or it may be a legitimate forecast. For now, it's just too far out to tell one way or another.

Weather Online

What is interesting, however, is the difference between the ECMWF model and the ECMWF ensembles. The above image shows mean sea level pressure values, valid on Christmas Eve (the same time as the second image we discussed). By simple comparison, note how the ECMWF model places the center of this storm somewhere in southern Ohio, while the ECMWF ensembles put northeastern Indiana in the center of this cyclone.

A further east track of this storm system could result in more of a snow impact to the Great Lakes, instead of primarily wind-driven snow in the Midwest and Ohio Valley from current runs of the ECMWF. But as I said earlier, we are still quite a while away from nailing down these details.

To summarize:

- Model guidance is beginning to sniff out a very strong storm system impacting most of the Central and East US on Christmas Eve into Christmas Day.
- This would severely impact travel.
- Snow would be confined primarily to the north-central Great Lakes into Canada.
- Very high uncertainty still exists.

Andrew

November 22-25 Potentially Significant Winter Storm

A powerful storm looks to present itself to the US around the November 22-25 timeframe.

Tropical Tidbits

Since this post will primarily be a model analysis, we'll begin with the ECMWF-Ensembles. This image shows the forecasted 500mb geopotential height values in the color shadings, with sea level pressure contours and high/low pressure demarcations superimposed. In this graphic, valid for the evening of November 24th, we see a storm system placed on the border of northeast Illinois into northwest Indiana. The minimum pressure is about 1003 millibars, which is a rather weak storm system. Despite its weakness, the mere presence of a storm on this graphic tells us there is some confidence in a storm occurring within this timeframe. Confidence is low to begin with, but it is there.

Tropical Tidbits

We now move on to the ECMWF model, also showing 500mb geopotential height values and SLP contours. For future reference, those two parameters will show up on all model graphics we analyze here today. The ECMWF model favors a very strong 987 millibar storm striking central Illinois on the morning of November 24th, with windy conditions overtaking the Ohio Valley and East US within the storm's warm sector. The sub-540 geopotential height values in the Midwest tell us that a widespread accumulating snow event would likely occur, particularly in the central and northern Plains into the Upper Midwest.

Tropical Tidbits

Pushing ahead, we now analyze the GFS ensembles forecast, valid here for the evening of November 24th. We find a weak low pressure system over central Lake Michigan, with a minimum central pressure reading of about 1005 millibars. Ensemble systems typically tend to be weaker than their operational counterparts (as observed with the ECMWF-Ensemble and ECMWF graphics above), since the ensembles take into account double-digit forecasts and average them all out. Regardless, the mere presence of a storm system is re-assuring to confidence.

Tropical Tidbits

Continuing on, we arrive at the GEM model forecast, valid on the evening of November 24th. The GEM, made by the Canadian meteorological service, shows a deep low pressure system of about 995 millibars right over Chicago, Illinois. This would support some stormy activity from the Southern Plains to the East Coast, but due to the retracted 540 line into the far Northern Plains, I'm not confident that this model is supporting a snowstorm for the Upper Midwest and Central/Northern Plains.

Tropical Tidbits

We now arrive at the GFS Ensembles, which paints an interesting picture for this storm. We see a storm system of minimum central pressure 1001 millibars placed just south of Chicago. This graphic, valid on the evening of November 24th, also draws the 540 line south into the Midwest and Central Plains. Again, since this is an ensemble forecast, the storm is not as strong as individual model forecasts. However, since it is showing up in the first place, confidence continues to rise in a substantial storm in this timeframe.

Tropical Tidbits

We've saved the best for last: The above graphic depicts the GFS-Parallel model forecast for this storm. We see a 979 millibar storm striking the northeast Illinois-southeast Wisconsin border, while 500mb geopotential height color shadings tell this storm to be a closed low. With very windy and cold conditions extending across the Midwest and Great Lakes, an accumulating snow event would likely be expected in the Upper Midwest and Northern Plains.


To summarize:

- Model guidance is confirming the possibility of a substantial storm in the November 22-25 timeframe.
- Some guidance supports a powerful storm system, possibly bringing both accumulating snow to the Upper Midwest, and severe weather to the East/Southeast.
- Confidence in this solution remains low due to the extended timeframe of this storm.

Additional updates will be posted in the coming week as more information becomes available.

Andrew

November 25-29 Potentially Significant Winter Storm

The storm system around the November 25-29/Thanksgiving timeframe continues to look like a significant storm system.

OPC

On the morning of November 8th, we saw the remnants of Typhoon Nuri reach peak strength via Ocean Prediction Center observation. As the chart shows here, the massive storm bottomed out at 924 millibars. This means the storm, located at about 170E and 55N, didn't break the record for strongest extratropical cyclone on record, but it certainly came close.

NWS

The graphic above, produced by the National Weather Service in Alaska, shows the observed mean sea level pressure of the storm at that 924mb reading, while the NWS office measured it at 930mb on the chart on the left, where strongest extratropical cyclones in the North Pacific are depicted. For multiple reasons, including the possibility that past storms may have been stronger than that 924mb reading, this storm was not declared the strongest on record in the North Pacific. Despite this, it's quite clear this storm was a historically-strong one, relative to storms in the last 60 or so years.

NOAA

Purely for comparison purposes, the team at KOPN Weather identified a strong storm in the Bering Sea on April 7th, 2011, bottoming out at 936mb, that was about 10 degrees W of where this very strong storm was observed yesterday in the Bering Sea. If you recall what happened about 2-3 weeks after the date of April 7th, we saw a certain tornado outbreak strike the South US, devastating thousands across the country, and causing millions on millions of dollars of damage. If we look at where the resultant storm ended up in mid-late April, utilizing the Bering Sea Rule, we find the storm situated in the Ohio Valley.

WPC

Now, keeping in mind that this sort of correlation is a tough one to use at best, not to mention all the caveats associated with long range forecasting, we could theoretically juxtapose the remnants of Typhoon Nuri and this storm in the Bering Sea on April 7th to get an idea of where the consequential storm in the US may end up. Recalling that the storm in 2011 in the Bering Sea was at about 180 degrees longitude, and almost the exact same latitude as the one observed in the last day or two, we find the remnants of Nuri placed about 10 degrees west of that 2011 storm. If we take the location of that storm system in late April (pictured above) and move it west, like the remnants of Nuri were west of that 2011 Bering Sea storm, we end up with a map like this:

WPC

Continuing this correlation, just to see what would happen, we notice that the remnants of Typhoon Nuri are moving eastward (a bit northeast in the process) in the Bering Sea right now, slowly at that. If this storm somehow does end up in that potential location outlined above, and if enough cold air is available (this will be discussed later), a significant winter weather event may strike the Central Plains, Midwest, Great Lakes, and Ohio Valley. Similarly, if the correlation works out, a severe weather event may strike the South US. Again, many caveats are associated with this method, and this should not be taken as "gospel", or at face value.

Tropical Tidbits

The above image shows temperature anomalies at the 850 millibar level (about 5,000 feet off the ground) over North America, as forecasted by the ECMWF ensembles ten days from today. In this image, we see a large swath of warmer than normal temperatures in the Bering Sea, with colder than normal conditions encompassing much of the United States and southern Canada. This looks to be a persistent pattern in coming days and weeks, as a large block of high pressure looks to set up shop directly over the Arctic, providing for a very cold period for North America. Extrapolating this to Thanksgiving, enough cold air should be in place for at least a modest threat of a significant snow event. Again, bear in mind long range caveats, but such a prognosis is favored right now.

To summarize:

- A potentially significant storm system still looks to evolve in the United States around Thanksgiving.
- Severe weather will be a possibility, namely in the South US.
- Significant snow will be a possibility, predominantly in the Central Plains, Midwest, Ohio Valley, and Great Lakes (for now).
- Thanksgiving travel may be severely hampered by this storm, if it does come to fruition as currently projected.

Andrew