Analyzing this winter’s observed patterns and looking ahead
Long Range: Detailed analysis of this winter’s patterns (updated Sat 2/26)
While the use of teleconnections alone can’t tell us exactly how an individual storm will track (there are many factors in play with the precise look to the weather board), they can allow us to understand large scale patterns. So let’s take a look at the composite image below. Fig. 1 provides a rundown of the types of patterns that we’ve experienced so far this winter, with each individual pattern numbered from 1 to 5.
Figure 1: Five major teleconnective patterns observed this winter
Now, let’s go over each of these individual patterns in more detail:
Back in the month of December, we had a –AO/–NAO/–PNA pattern.
This type of pattern was indicative of a rather unfavorable set up for winter storms/cold based on what was going on in the Pacific, but strong high latitude blocking pressing up against it resulted in a shorter wavelength pattern. Also, with the AO very strongly negative (characterized by strong upper level high pressure/positive anomalies in the higher latitudes) for much of December, counteracting negative anomalies developed in the mid-latitudes. This allowed for very cold weather, with a mean temperature averaging 4.5°F below normal for the month. But blocking was so strong that storms were suppressed to the south of NYC. As we began to come out of this pattern, our major Post-Christmas blizzard made a surprise voyage up the coast.
The start of the new year marked the beginning of a new pattern featuring a –AO/–NAO/neutral PNA, which was favorable for snowfall.
As such, we experienced a significant snowstorm (major for north/eastern areas) during early mid-January and a couple smaller snow events.
Next up we saw a –AO/neutral NAO/+PNA pattern.
This was the first time that we started to have precipitation type issues, but it was still okay because despite the beginning of the breakdown of blocking we still had plenty of cold air focused in the mid-latitudes (compliments of the –AO) and a very favorable Pacific-North American pattern. Our major late January snowstorm occurred in this pattern, along with numerous light/moderate events. With a monthly total of 36 inches recorded at Central Park, January 2011 went into the record books as the snowiest January ever recorded. But January was also a big transition month for the pattern, as evidenced back in Fig. 1. Big changes were evident in mid-late January as we were in a continued shifting pattern.
We then adjusted into a +AO/+NAO/+PNA pattern for a short time in the very end of January and very beginning of February.
We began to have major precipitation type issues because although the storm track was decent, there was a lack of cold air and blocking. In this pattern we saw our freezing rain/rain storm.
Finally, upon the start of February, the PNA took a dive, and starting in early February and continuing well into the future, we remain in a longer wavelength +AO/+NAO/–PNA pattern, with a clear La Nina ridge in the Pacific.
This is about as bad as it gets for snowstorms, and we’ve seen as such with only very little snowfall compared to earlier in the season when we basically had what felt like snowstorm after snowstorm.
The problem is that energy keeps on coming around a semi-permanent ridge which continues to pump up near the eastern Aleutian Islands. This keeps a persistent mean trough in the western US with ridging much of the time in the east. Fig. 2 provides a general overview of the persistent pattern that we continue to experience.
Figure 2: Persistent 500mb pattern
This can all be attributed to the fact that the La Nina is now west-based, as evidenced by the change in configuration of recorded anomalies in the equatorial Pacific, shown in Fig. 3 below.
Figure 3: SST anomalies in the equatorial Pacific
The anomalies in the eastern portion of the equatorial Pacific (NINO 1+2 region) have warmed so much that they are above normal for the first time since early last spring. So the lack of convection above the area of the Pacific with the coldest anomalies (NINO 3.4 and 4, the only regions still lower than -1°C below normal) allows a mean ridge to remain in place in this region. This continues to support my theory that the change in orientation of La Nina out of east-based in late December caused a major pattern shift in mid-late January that resulted in the depletion of high latitude blocking and the gradual turnaround to a milder, less snowy pattern. Only five of the winter seasons featuring a La Nina in the past 60 years were truly east-based, which is how our current Nina was based during the first part of the winter when we saw our major snowstorms. But with the La Nina now full-fledged west-based, I’m doubtful that we’ll see any significant pattern change before the end of winter. We have finally entered into a typical La Nina, and we can see why major snowstorms don’t normally occur for NYC in a typical La Nina. The storms are tracking farther and farther west. Once again, this doesn’t mean that we won’t see any more snow or cold, but it does mean that the pattern will keep on trying to return to a milder one.
So far, Central Park has recorded 60.9 inches of snowfall this season. This gives us the bronze metal; we are in third place for the highest recorded snowfall season ever. We only need 2.4 inches more and we get silver, which I think is still very possible since another light/moderate event or two may still remain in our future. However, gold is a long way’s off, set by the 75 inches we had in '96. I feel confident that the historic winter of '96 will keep its special place as our snowiest year ever. One more 12-18 inch snowstorm would have done it, which wouldn’t have seemed like too much earlier this winter, but those just don’t happen in the pattern that we’re in now. I think we will sneak in a little more snow, though. Of course, I’d like nothing more than another nice big snowstorm before winter’s end, but unfortunately I don’t foresee that occurring.
Updated: 10:14 PM GMT on February 26, 2011
A A A
Rain and wind to impact the area on Friday
Light rain has overspread the area ahead of a warm front. A southern branch shortwave will cruise on through tomorrow with a low pressure system intensifying just ahead of it:
The shortwave will be sharp, but more importantly the storm will be located in the left front quadrant of an approaching 120kt 300mb jet streak:
This will lead to plenty of dynamic lift associated with divergence aloft. The nose of a strong low level jet will provide the area of low level convergence:
The intensifying storm system will also create a sharp pressure gradient and lead to strong winds. The steady rain will taper off by around lunchtime as dry air filters in at the mid-levels, but cold air will be moving in aloft on Friday, which could create an environment for convective showers in the afternoon. We will also see a rapid drop in temperatures into the 20s Friday night.
Updated: 6:20 AM GMT on February 25, 2011
A A A
Light snow event expected tonight into early Monday
Tonight/tomorrow morning’s light snow event is a feature of a very interesting pattern. It is a pattern that we haven’t had to deal with very much this winter. This involves a southeast ridge. I was concerned about a SE ridge like this one popping up for most of the winter (which of course didn’t play out), but now that the La Nina has gone more west-based, the pattern is there for such an entity to assert itself. This storm involves two waves of precipitation.
Figure 1: NAM upper level forecast for 7 am Mon
Now, looking at this map, you may notice that we have warm air over the southeast in association with a strong SE ridge, but we also have cold air associated with a rather weak upper low centered over southeastern Canada. We also have an area of good upper level dynamics back over the Plains. Examining the upper level pattern alone suggests very little going on over our area other than a battle zone between cold air to the north and warm air to the south. However, there is more to the story than this map alone would suggest. In fact, while the boundary between the ridge/trough configuration at the 500mb level is over upstate NY and central New England, the low level boundary is right in our vicinity. This means that cold air is being injected in at the low levels, while warmer air is overriding that cold air in the mid and upper levels. When mid-level warm air glides over cold low level air, we call it an overrunning, or isentropic lift, event. Now, let’s take a look at another panel.
Figure 2: NAM low/mid level forecast for 7 am Mon
The map itself is the surface prog, and the blue arrow shows the low level flow, a general northerly component flow around surface high pressure centered over south central Canada. I also put a yellow arrow indicating the mid-level air flow, which is out of the west. This overrunning will lead to snow developing tonight and continuing into early Monday. Latest model guidance has been trending toward speeding up the northern branch of the jet stream, and slowing down the southern branch. With the N. stream upper low over southeastern Canada farther S/E, this allows for a better injection of cold air into the first wave of overrunning/isentropic lift precipitation. It also permits the second wave of dynamic precipitation to come farther north because the N. branch upper low is already on its way out by the time the S. branch system (over the Plains at the time of Figure 1) approaches. These trends are all very good for us, but I still think the second wave may miss us just to the south. The main reason why I’ve avoided making a definitive forecast on what would occur with this system is because this is a very sneaky snow set up. The overall pattern does not support snow, but the northern branch is acting with the southern branch to allow it.
My main concern is that some of the numbers being spit out by the models may be too high considering that the first wave is a quick-hitting overrunning scenario with little in the way of dynamics (as we saw in Figure 1). Overrunning events often begin a little earlier than one might expect (unless there is too much low level dry air to overcome), and more importantly they are usually in and out fairly quickly. But this does look like an all snow event. Based on the current data, I would expect about 2 to 4 inches.
PREVIOUS DISCUSSION from 2/18
Drastic change in weather as temperatures crash through tomorrow
I hope everyone didn’t get too accustomed to the mild temperatures, because it’s back into the ice house this weekend. Get ready for a major change in weather that is already taking place overhead. Cold air filtering in aloft earlier this evening interacted with the warm surface to destabilize the atmosphere and produce a couple of isolated showers/thunderstorms which moved across the south shore.
Temperatures at 850mb will crash 50 degrees in a day. This afternoon, temperatures in the mid-levels were close to 50°F, which supported highs in the upper 60s. Tomorrow, mid-level temperatures will drop to 0°F by early evening:
What makes this pattern interesting is that we have a strong, broad 588 dm upper level ridge centered in the western Gulf of Mexico, supported by continued troughing along the west coast. But at the same time we have a 510 dm upper low to our north. The interaction between the two is creating a strong and powerful jet raging into the northern mid-Atlantic:
Ordinarily, you might look up in the sky in this situation and note how quickly the clouds are moving. Like right now, upper level winds are moving at well over 100 mph. But luckily the winds are not aligned (as one moves vertically upward in the atmosphere), so we don’t get the transfer of momentum down to the surface. The story changes tomorrow as cold advection ensues, winds align out of the northwest, and the mixing height increases up to 780 mb. This is indicative of a very well mixed atmosphere, and winds at the top of the mixed layer will be moving along at close to 60 mph, so this is the reason why we are concerned that surface winds will gust up to that speed during tomorrow afternoon:
Winds will not decouple Saturday night, so gusty winds will continue with widespread wind chills back into the single digits and approaching zero.
Updated: 10:29 PM GMT on February 20, 2011
A A A
Semi-permanent Aleutian ridging will stream in a taste of spring
All signs continue to point to a very mild stretch of weather over the coming days. This is all traced back to a very strong ridge that is building over the Aleutian Islands.
This ridge is allowing a trough to deepen in the Gulf of Alaska and a corresponding ridge to build over the southwest. As is, this pattern keeps us in a fast zonal flow pattern with numerous disturbances and slightly above average temperatures.
However, an upper low is expected to form in the northern stream over eastern Asia and partially beat down the Aleutian ridge. As it does so, a piece of the ridge will wane and translate eastward, pushing the GOA trough ashore the west coast.
Once this occurs, the ridge and associated building mild air will have nowhere else to go but into the eastern US, and it will meet little resistance as a weakly positive NAO prevails downstream.
It is important to note that a positive anomaly at this longitude in the Pacific is the same as the average La Nina composite ridge position. So this is finally the La Nina influenced pattern that I was afraid would have ruined our chances at snow this year. Luckily, it is coming after we have already experienced an active winter season. Unfortunately, with the La Nina continuing to shift increasingly west-based, I don’t foresee the hemispheric pattern returning that would allow for a negative NAO or a Pacific pattern that would support continuous troughing in the east. We will continue to see brief shots of colder air over the long range, but I think it is unlikely that there will be anymore extended periods of cold or major snowstorms. You can still find my theory on the sudden break down of high latitude blocking here.
Now, we will have a sweet little preview of the coming warmth on Valentine’s Day. It’s interesting to note though that the mild highs on Monday will only be resulting out of a zonal flow regime and ahead of a cold front. This cold front will also make for a strong boundary layer WSW to W flow which will allow the warm up to spread right up to the coast. The cold front and associated upper trough will then usher in some colder air for Tuesday, although the atmosphere will stay decently mixed during the day so highs will still make it well into the 30s. The high moves overhead Tuesday night, and after that, it moves offshore with upper ridging building in, and at that point there will be no containing the mild air. We’re talking about 500mb heights approaching 570dm over southern New Jersey bringing up some very mild air in the mid and upper levels, and a southwest return flow around offshore surface high pressure.
Now, exactly how warm it gets is still somewhat in question. The mild air will be overhead, there’s no question about that, but in order for it to really warm up, that mild air needs to get mixed down to the surface. For that to occur, the winds must be aligned, or out of the same direction. Otherwise we can’t get the transfer down to the surface. I took a look at the forecast soundings from the GFS yesterday, and they were showing better aligned winds connecting the surface up to about 900mb. However, today the GFS is only showing a mixing level of about 950mb, with winds below that level having a more south of west component. The EC is showing highs surpassing 60 degrees on Friday, but this would only be possible if we can mix out that inversion near 950mb.
The surface prog from this morning’s EC does show surface winds that are more SSW, so it looks like we may be capped in the upper 50s on Friday. Regardless, get ready for the mild air. It’s coming.
Updated: 7:05 PM GMT on February 13, 2011
A A A
Change in La Nina orientation may have forced blocking disappearance
Long Range Discussion: Change in La Nina orientation may have forced blocking disappearance
Ensemble predicted NAO forecast
What a difference. Remember earlier in the winter when we saw a strongly negative NAO. Did you ever think we would see a forecast panel like this one, with a positive NAO forecasted straight through the next two weeks? Something has definitely changed. I’ve been doing some research over the past month to try and figure out what might have contributed to the interesting hemispheric anomalies that we saw relative to what you would normally expect in a La Nina. I really liked some of the correlations that I found, and I want to take the time out to share them.
As you probably know, this winter started out with incredibly strong high latitude blocking. You may recall that back in December, we were all excited about the possibility of having our first major winter storm on the same day as last year (Dec. 19). However, we all watched sadly as this storm became one for the “fish.” The blocking was so strong that it shunted a big storm out to sea. Many New Yorkers marveled at how places farther to the south had already seen snowfall, while our roads stayed dry. The blocking remained intense during the first part of January as we saw an upper high actually retrograde across northern Canada. In fact, the NAO had stayed strongly negative from the beginning of winter right through early January.
But then that was it. Sure, we did see ridging start to build in late January in the north Atlantic, but this time rather than expanding it only got pressed back down by polar energy. The pattern has definitely changed. It’s very difficult to say for sure what happened to make the high latitude blocking disappear. However, there is one force in particular that I’ve come across that may at least shed some light on what happened. This all goes back to my winter forecast that I released on November 30th. You may recall that in my forecast, I talked about how we were in an east-based La Nina, but based on the ensemble forecasts at the time, I expected the La Nina to turn more west-based in January. The ensembles were accurate—and let me show you the observations that prove it. Below we have the observed sea surface temperature anomalies in the tropical Pacific:
I would like you to focus on the graph for the NINO 1+2 region (NINO 1+2 is basically the far eastern tropical Pacific). Note how the anomaly went from -1.7 C in mid-December all the way up to its current state of 0 C. The biggest change was in late December when the average temperature anomaly went up more than a degree in less than two weeks. So if that major alteration triggered a change in the overall pattern, we would be looking for a significant adjustment of anomalies to occur between December (when we had an east-based La Nina) and January (when we had a more west-based La Nina).
I have made a comparison between the observed anomalies in December and the observed anomalies typical of other seasons which featured east-based La Ninas. I compiled this image (click on it to enlarge) by using the years that had an anomaly of -1 or below in NINO 1+2 (there were only 5 years, and the most recent one was back in the 70s—these don’t happen often):
Composite anomaly for winter seasons with east-based Nina (left) and December 2010 observed pattern (right)
The thing that I thought was very interesting is how similar the placements of the two positive anomalies actually are, one in the north central Pacific, and the other over southern Greenland. Based on the anomaly positions, this composite also suggests that the development of Greenland blocking is strongly characteristic of east-based La Ninas. So not only did we have a strongly east-based La Nina, but we also observed anomalies in the broader hemispheric pattern that were characteristic of the few observed east-based La Nina seasons of the past.
So once January came, NINO 1+2 anomalies were much higher (like we saw on the observed SST anomalies image above), while anomalies in the other regions remained lower than -1. Therefore, January featured a more typical La Nina that is less east-based and more west-based, and as such, we should have expected to see a change in the Pacific pattern to one that would be more typical of a La Nina. Below on the left we have the compiled image of all the basin-wide La Ninas, and what you would normally expect, and on the right is the observed pattern for this January:
Composite anomaly for winter seasons with basin-wide Nina (left) and January 2011 observed pattern (right)
You can see how last month featured a pattern that was more similar in the Pacific to a “normal” La Nina, with a ridge in the eastern Pacific. The north Atlantic pattern doesn’t look similar at all, with the anomalies essentially flipped, but realize that the strong positive anomaly that we see on this image for January is an average of the entire month—we had the strong high that retrograded across northern Canada early in the month, and then a ridge that tried to strengthen east of Greenland late in the month before getting broken down by polar energy. The average of the two gives us an anomaly for the month near southern Greenland. But that was only a combination of a block that was in the process of moving away, neutral anomalies mid-January, and an east-based negative NAO block that didn’t have a chance to significantly impact the upstream pattern before getting beaten down. You can watch these changes occur in motion on the animation below:
Despite what the average pattern for the month shows, high latitude blocking was well on its way out through most of January.
So if this change in La Nina orientation was the force that altered the pattern like my data hypothesizes, then I wouldn’t expect to see the negative NAO return anytime soon. I guess the question remains—are there other forces at work here? All we can do is wait and see. If high latitude blocking doesn’t return for the rest of the winter, then based on the data I believe the change in orientation of the La Nina may have caused this. The change in orientation of La Nina was basically a shift in the colder anomalies westward in the tropical Pacific.
The point I’m trying to make is that a significant change did occur in the Pacific in late December, and I think we may be feeling the effects of it. This may be the force that led to so much of a change in high latitude blocking, change in Pacific-North American pattern, etc. You’ll also continue to see a longer wavelength pattern as you’re not trying to press blocking up against the Pacific pattern.
Following along these lines, I think we may actually start to see a bit of a warm up as we head into the middle of February. For a while now we’ve seen a ridge off the west coast in the eastern Pacific. This pattern is becoming boring because more often than not, this pattern would correlate to an inland running storm track. That ridge is finally forecast to come ashore next week, and a ridge is expected to build over the central Pacific/Aleutian Islands, starting late next week. You can see the old ridge over the eastern Pacific on the image below, and note the new ridge just starting to build near the western Aleutian Islands:
At first, the new developing pattern will be conducive to a storm coming up the coast late next week; however, as time wears on, the new ridge will slowly translate eastward and eventually lead to a trough dropping into western North America. The ECMWF and GFS ensemble mean have both been consistent over the past few days of having that ridge build over the central Pacific and slowly translate eastward, which leads to reasonable confidence in the forecast:
As a result, a positive EPO/negative PNA pattern will develop, leading to a fast zonal flow flooding the nation with milder air by mid-month:
Ensemble predicted EPO index (left) and PNA index (right)
Based on my discussion above, I wouldn’t expect to see any high latitude blocking infringing on this developing flow pattern as the NAO remains positive. So I would expect to see some milder air come straight across the country and into the northeast.
Another round of arctic chill to deal with before pattern change
Short Range Discussion: Drastic change in weather coming tomorrow (Mon 2/7 Update)
It’s amazing how just a small difference in timing can make such a huge impact on the weather. We see tonight an impressive plume of moisture, but the northern stream cold air and dynamics are lagging behind just a bit too much. This could have been an impressive snowstorm for someone along the east coast if the timing were different.
As the two streams begin to consolidate and the energy combines, low pressure will rapidly intensify over the Canadian Maritimes. As it does so, a sharp pressure gradient will steepen between the departing low and building high pressure. This will make for strong cold advection during the afternoon Tuesday. Thicknesses will drop to near 510dm by nightfall, and arctic air will be ushered in by the lagging sharp northern stream shortwave. Very cold air will stream down from aloft as the atmosphere becomes well-mixed, with temperatures dropping through the day and winds gusting up to 50 mph.
However, like I discussed in my long range outlook, don’t expect the frigid air to stick around for long—a big pattern change is coming!
Medium Range Discussion: Storm potential for late in the week
Over the past couple of days we’ve seen the models trend weaker and significantly eastward with the storm that is expected to develop late in the week. Below is a comparison between last Thursday’s 18z GFS run which was showing a big storm coming up the east coast (left) and this morning’s GFS run which shows a weaker storm that goes out to sea (right):
Look closely (you can click on it to enlarge). Note how the overall teleconnective set up is quite similar between the two. We have a central Pacific ridge, Gulf of AK trough, ridge along the west coast, and a strong polar vortex centered near Hudson Bay. Also, the Pacific jet stream disturbance is in a similar spot over the southeast (denoted in each image by the solid black line). But the danger is in the details, and upon closer observation the progression of the northern stream is quite different in each solution. Last Thursday’s GFS run showed a more progressive polar jet stream, with northern stream energy and associated height falls dropping into the Great Lakes by early Thursday and phasing with the Pacific jet stream and creating the powerful storm. On today’s GFS run, the northern stream energy lags back over southern Canada, and the Pacific disturbance doesn’t have anything to interact with until that northern piece drops down, which happens much too late for us. The northern stream disturbance was also much sharper on Thursday’s run as opposed to the broader polar trough depicted on today’s run.
The other thing that I want to highlight in the graphics is the differences just downstream of the storm system. On this morning’s GFS run we have a 50/50 low with suppression of the jet stream (by the way, the 50/50 low is made up of energy from our early week system), while in Thursday’s run in the same area is a ridge which would allow the storm to intensify and track farther to the north and west. But still, note how even in Thursday’s run, although the ridge is there, it is slightly beaten down and not able to assert itself completely because that strong polar vortex is in the way. This reminds me a little bit of back in December when we had the strong PV over Canada with better suppression. So even despite the positive NAO, I would not expect to see a coastal hugger/rain event with this storm, even in the “best” case ridging scenario (i.e. Thursday’s run).
The new GFS run also shows additional energy coming straight down the chute and into the Plains in a fast moving progressive Pacific jet stream which further helps to keep our disturbance quickly moving along.
It is difficult to say for certain what will happen because we are still several days away, and whenever dealing with Canadian energy the models seem to struggle a bit more. It doesn’t look like the right pattern at all for a coastal hugger/inland runner. I think the models are reasonable in their depiction of a flat wave/out to sea scenario, but I still see a possible slight northwest adjustment of the GFS with a potential light/moderate snow event for the area.
So to recap: The model trend of holding back and broadening polar jet stream energy would prevent the storm from having anything to phase with, so no stream of fresh cold air flowing in and no additional energy to crank up the storm, and it remains a flatter wave. Watch the northern stream disturbance in relation to the Pacific disturbance. The Pacific disturbance is likely being fairly well handled by the models, but the precise orientation of the polar vortex and the associated disturbances rotating around it and their involvement in the storm will be important. My main concern right now is if the timing isn’t right (like the models are trending right now), then we watch a flat wave scoot by to the south (still possibly tracking a bit closer to the coast than being depicted right now). The way the models are trending it doesn’t look like we’re going to have much help from the polar jet stream—this may just turn out to be a Pacific disturbance with a flat surface wave and maybe at best some light to moderate snow.
Updated: 7:04 PM GMT on February 13, 2011
A A A