Winter 2022-2023 Promises Plenty of Mid Season Mayhem

[40/70 Benchmark]

Addressing the Question of Whether or Not Winter is Over

 Matter of Perspective

In the world of weather, and everything else for that matter, perception means everything. Thus when one ponders whether or not winter is in fact "over", the appropriate response is highly subjective in nature and is dependent upon both where one lives, and what the term "winter" entails. It is with this in mind that perhaps the question should be reframed to more "exclusive", as opposed to "inclusive", since describing what the pattern is not is less prone to the distortion that is attributable to viewing through a different lens.

It is evident when viewing the 500mb prognostications over the course of the next week that a sustained spring time pattern is not evident for the duration of the month of March.

 

 

 

This is not to say that it will not feel present on certain days with full sunshine in the absence of a true arctic air mass, such as today, because it will. That is the benefit of being the recipient of late March strength solar irradiance. However, what should be apparent on the chart above valid next weekend is that there continues to be a fair degree of high latitude blocking working on conjunction with impulses periodically ejecting eastward from a very deep western trough. This implies that it should remain seasonably cold and stormy and if this is a pattern that sounds familiar, it should. A pattern consisting of NAO blocking to the north in conjunction with a polar vortex lobe in the vicinity of Hudson Bay Canada is a pattern that would normally support the threat of significant snows across most of the season even at this late juncture. But, as has been the case for most of the season, the potential for harsh winter weather is mitigated due to the ridge of greater heights over the southeast US in response to the very deep western US trough upstream. This not only acts to limit cold air infiltration and encourage a storm track closer to the coast, or even inland, as opposed to offshore, but the compressed medium between the vortex to the north and the ridge to the south causes the system to shear out on approach this weekend.

 

 

 

This means that the answer to the million dollar question regarding whether or now winter is over is dependent upon:

A) Where you live, of course.

What winter means to you...if it means significant snowfall, then the answer significant snowfall, then the answer is likely "no" to the south of the above line and below approximately 1000' of elevation.  Locales to the south of that line and at or above 1000' in elevation, across the northern Worcester hills and Berkshires, could see light accumulations. 

If this sounds familiar, than it should, since these are the locations that have experienced impactful snowfalls all season long and may continue to do so into next week, when a similar evolution may take place Monday night into Tuesday.

 

Rinse & Repeat-

[40/70 Benchmark]

March Forecast Offers Mixed Results

Low Western heights Temper High Latitude Blocking & Active Storm Track 

The Eastern Mass Weather forecast for a very active and more wintry month of March had been well advertised dating back to the release of the winter outlook this past fall. The parallels drawn to the analogs of 1956 and 2018 were valid in the sense that there was in fact a second major disruption of the polar vortex.

 

 

 

And this did in fact lead to another period of major late season high latitude blocking that ultimately led to a major mid month winter storm that resulted in over 3 feet of snow across the hills of northern Worcester county and the Berkshires.

 

 

However, as evidenced by the 500mb chart above, the NAO block was largely offset by extremely low and persistent troughing over the western US. 

 

 

 

While lower western heights attributable to residual cool ENSO influence were anticipated, this was clearly more extreme than forecast. 

This obviously resulted in lower heights over the western US, and slightly elevated heights over the eastern US relative to the forecast, which may have played a role in the delayed phase of the major mid month winter storm that greatly mitigated snowfall impact over the coastal plane relative to expectation.

 

Eastern Mass Weather Forecast H5 Composite

 

Unsurprisingly, the March 1956 analog was plagued by the same issue greater western heights relative to 2023, which negated the fact that March 2023 featured even greater NAO blocking than 1956.

 

 

The month or March was never expected to be as cold as March 1956 or 2018 across the eastern US due to the fact that the PV was displaced onto the other side of the globe.

 

But was still milder than forecast over the eastern US due to persistence of very low western heights, which actually encompassed the western half of the AO domain space and also acted to bias the AO positive for the month.

 

 

 

 

 

 

March temp forecast near normal for region:

 

 

Reality was a few degrees above average:

 

The precipitation forecast was decent in reflecting near normal precipitation across the region:

 

 

But the active subtropical jet that has been pummeling California was missed.

 

 

 

[Damage In Tolland]

On 4/3/2023 at 12:27 PM, 40/70 Benchmark said:

March Forecast Offers Mixed Results

Low Western heights Temper High Latitude Blocking & Active Storm Track 

The Eastern Mass Weather forecast for a very active and more wintry month of March had been well advertised dating back to the release of the winter outlook this past fall. The parallels drawn to the analogs of 1956 and 2018 were valid in the sense that there was in fact a second major disruption of the polar vortex.

 

 

 

And this did in fact lead to another period of major late season high latitude blocking that ultimately led to a major mid month winter storm that resulted in over 3 feet of snow across the hills of northern Worcester county and the Berkshires.

 

 

However, as evidenced by the 500mb chart above, the NAO block was largely offset by extremely low and persistent troughing over the western US. 

 

 

 

While lower western heights attributable to residual cool ENSO influence were anticipated, this was clearly more extreme than forecast. 

This obviously resulted in lower heights over the western US, and slightly elevated heights over the eastern US relative to the forecast, which may have played a role in the delayed phase of the major mid month winter storm that greatly mitigated snowfall impact over the coastal plane relative to expectation.

 

Eastern Mass Weather Forecast H5 Composite

 

Unsurprisingly, the March 1956 analog was plagued by the same issue greater western heights relative to 2023, which negated the fact that March 2023 featured even greater NAO blocking than 1956.

 

 

The month or March was never expected to be as cold as March 1956 or 2018 across the eastern US due to the fact that the PV was displaced onto the other side of the globe.

 

But was still milder than forecast over the eastern US due to persistence of very low western heights, which actually encompassed the western half of the AO domain space and also acted to bias the AO positive for the month.

 

 

 

 

 

 

March temp forecast near normal for region:

 

 

Reality was a few degrees above average:

 

The precipitation forecast was decent in reflecting near normal precipitation across the region:

 

 

But the active subtropical jet that has been pummeling California was missed.

 

 

 

What was the final grade you gave yourself? Could not have been above a D- right ?

[40/70 Benchmark]

14 hours ago, Damage In Tolland said:

What was the final grade you gave yourself? Could not have been above a D- right ?

I haven't done that yet....but it will be higher than that for the season. My seasonal outlook encompasses more than how much snow falls at 990' in Tolland, CT.

[Damage In Tolland]

23 minutes ago, 40/70 Benchmark said:

I haven't done that yet....but it will be higher than that for the season. My seasonal outlook encompasses more than how much snow falls at 990' in Tolland, CT.

Should probably be lower for entire region 

[40/70 Benchmark]

16 minutes ago, Damage In Tolland said:

Should probably be lower for entire region 

If you don't know how to formulate one, I wouldn't expect you to have any clue on how to properly assess an outlook, so this isn't surprising. 

Hope you enjoyed your March 2012.

 

[40/70 Benchmark]

The snowfall portion of the outlook was clearly flawed in a major way for the majority of SNE...not all, but most of the area. Would not debate that. All I am saying is that there are many other factors to consider rather than snowfall for SNE and on a grander scale, the work was not nearly as bad as it appears.

[tamarack]

When the pattern is basically how it was forecast but the pieces just don't come together for snow, one should grade the forecast, not the snow stake.  

[40/70 Benchmark]

15 hours ago, tamarack said:

When the pattern is basically how it was forecast but the pieces just don't come together for snow, one should grade the forecast, not the snow stake.  

I factor everything in....it wasn't a total fluke that snowfall underperformed my expectations in most of SNE and it was warmer than forecast. There were issues that I will assess and then grade the work accordingly...all I am saying is it isn't as bad as SNE snowfall would make it appear.

[Great Snow 1717]

On 4/10/2023 at 6:21 PM, Damage In Tolland said:

What was the final grade you gave yourself? Could not have been above a D- right ?

Nearly every forecast for the winter of 22-23 failed miserably.  Raindance had a great handle on the winter. He has 3 things going for him.  He doesn't live anywhere close to SNE, and he isn't biased towards cold and snow, and he doesn't over analyze the situation.

The Pope also had a great handle on the winter. Some people ridiculed him for his thoughts and opinions but he was right far more often than he was wrong.

Every forecast that called for some combination of average/below average temps and average to above average snowfall deserves an F- grade. And there are many forecasts that deserve a grade of F- because essentially it was a non winter. 

Here in Methuen I give the winter an F- grade...well below normal snowfall and well above average temps. I think I wore a winter coat on perhaps 8 or 9 days the entire winter. Often times when it was snowing the snow was already in the process of melting. The warm ground limited accumulations and the time snow remained on the ground.

The downplaying of a 3rd year La Nina and the warmth in November led to many forecasts going bust. The warmth in November was a harbinger of things to come.

[40/70 Benchmark]

20 hours ago, Great Snow 1717 said:

Nearly every forecast for the winter of 22-23 failed miserably.  Raindance had a great handle on the winter. He has 3 things going for him.  He doesn't live anywhere close to SNE, and he isn't biased towards cold and snow, and he doesn't over analyze the situation.

The Pope also had a great handle on the winter. Some people ridiculed him for his thoughts and opinions but he was right far more often than he was wrong.

Every forecast that called for some combination of average/below average temps and average to above average snowfall deserves an F- grade. And there are many forecasts that deserve a grade of F- because essentially it was a non winter. 

Here in Methuen I give the winter an F- grade...well below normal snowfall and well above average temps. I think I wore a winter coat on perhaps 8 or 9 days the entire winter. Often times when it was snowing the snow was already in the process of melting. The warm ground limited accumulations and the time snow remained on the ground.

The downplaying of a 3rd year La Nina and the warmth in November led to many forecasts going bust. The warmth in November was a harbinger of things to come.

I'm not sure how you can consider a multifaceted seasonal outlook a complete failure because of one aspect, though.....while I agree that raindance had the best outlook, and have stated as much many times, I don't think its entirely fair to ignore the fact that he did whiff on the NAO....that needs to be factored. I hit on many aspects of the outlook, which I will demonstrate in my post analysis, so it doesn't make much sense to claim the work is a dumpster fire because of mid Atlantic and SNE snowfall. I wasn't that far off in CNE/NNE and did I a pretty good job identifying early and late periods of blocking. The middle portion of the season was a big whiff in terms of temps, but I did get the active pattern right.

There were certainly some big issues and it definitely wasn't an A or B....I agree with that. If you are speaking specifically as it pertains to SNE snowfall, then yes...that is probably a D- or F.

Sometimes I feel like you have to actually attempt one of these before you really have any insight on what goes into it and how to assess. 

[40/70 Benchmark]

Even though snowfall was comparable for a lot of SNE, this season was definitely not a 2011-2012 type of dumpster fire pattern....if it were, then I would grade my effort an "F". The pattern evolved pretty similarly to how I postulated it would, but unfortunately the result in terms of SNE snowfall was still a dumpster fire. The main issue I see with my work is that heights out west ended up being lower than I expected and the middle portion of the season remained void of blocking (I thought January may bounce back some after a big thaw), so its not a total fluke that it was warmer with less snowfall than expected....but the fact that it ended up this bad was also due to some synoptic misfortune.

[40/70 Benchmark]

Just doing a quick calculation, I actually nailed the DM NAO value....was within my predicted range. Satisfying since the polar domain is where I have struggled a great deal since doing this.

[40/70 Benchmark]

Ironically enough, this season may have been my best effort in terms of teleconnections Only thing I really whiffed on was the EPO, as I expected it to be positive and it was actually pretty negative. While I hit on the ONI, I definitely blew the modoki index this year, but the site it down right now. I don't like to do the full write up until May due to random snow events, but I am confident this forecast effort was actually pretty good, SNE snow not withstanding. A lot of the cold ended up on the other side of the globe, though, in which case a neg AO is actually a mild signal. This is why I was off on temps IMO...that and the bouts of +PNA, most notably in January, were heavily west biased, which essentially acts as a -PNA...so while it looks like I essentially nailed the PNA, western heights were lower than forecast, which is evident when comparing the H5 chart to the forecast composite. This in conjunction with the cold being in Eurasia also played a role in the season ending up milder than forecast. Its details like this that do not show up on a grander scale in the teleconnection tables that can turn an ostensibly serviceable winter season into a rat.

December-March 2022-2023 Teleconnection Forecast

 

 

Index Value	Predicted '22-'23 DM   Value Range	Actual  '22-'23 DM Value	Forecast Error
Pacific Decadal Oscillation (PDO)	-1.27 to -1.57	-1.39	Verified
Perennial North American Pattern (PNA)	-.35 to -.65	-.68	.03 too positive
ENSO	SON -1.0 to -1.2 EMI: -.3 to -.5 (slightly east-hybrid)	SON -1.0 (Verified)	Verified
(J-M) East Pacific Oscillation (EPO)	+.30 to +.60	-.43	.73 too positive
Arctic Oscillation (AO)	-.15 to -.45	-.37	Verified
North Atlantic Oscillation (NAO)	+.15 to +.45	+.23	Verified

 

[WxWatcher007]

Strongly agree, Ray. While the past “winter” is something I’ll loathe to the day I die, my misery doesn’t necessitate a failing grade on any outlook. Like @tamarack said, it’s not just about looking at the snow stake, even if that’s what most only care about.

What you’re doing is far more complex than that. I respect it.

[40/70 Benchmark]

9 minutes ago, WxWatcher007 said:

Strongly agree, Ray. While the past “winter” is something I’ll loathe to the day I die, my misery doesn’t necessitate a failing grade on any outlook. Like @tamarack said, it’s not just about looking at the snow stake, even if that’s what most only care about.

What you’re doing is far more complex than that. I respect it.

Thanks alot....believe me, I don't disavow my errors.....quite the contrary. The correct parts of a forecast provide validation of some of the methodologies, which is gratifying, but its the errors that serve as the real impetus for growth and act as a beacon to guide future seasonal efforts through that fog of ambiguity that is long range forecasting.

[40/70 Benchmark]

I wish there was a better way to forecast not only the predominate seasonal state of indexes, but also their character/orientation....ie, the heavy west bias of the PNA and the cold loading into the eastern Hemisphere. It is this that is the smoking gun IMHO and it is why I feel like misdiagnosing ENSO is so often fatal. I got the peak ONI correct.....BOING.....but I don't think its a coincidence that la nina shifted to modoki and so much went wrong. I expected a continued slight eastward lean....had I realized that this was going to shift into a modoki signature for winter, I would have anticipated a lot more of these details to conspire against us.

[Great Snow 1717]

2 hours ago, 40/70 Benchmark said:

I'm not sure how you can consider a multifaceted seasonal outlook a complete failure because of one aspect, though.....while I agree that raindance had the best outlook, and have stated as much many times, I don't think its entirely fair to ignore the fact that he did whiff on the NAO....that needs to be factored. I hit on many aspects of the outlook, which I will demonstrate in my post analysis, so it doesn't make much sense to claim the work is a dumpster fire because of mid Atlantic and SNE snowfall. I wasn't that far off in CNE/NNE and did I a pretty good job identifying early and late periods of blocking. The middle portion of the season was a big whiff in terms of temps, but I did get the active pattern right.

There were certainly some big issues and it definitely wasn't an A or B....I agree with that. If you are speaking specifically as it pertains to SNE snowfall, then yes...that is probably a D- or F.

Sometimes I feel like you have to actually attempt one of these before you really have any insight on what goes into it and how to assess. 

Uhmm did you not read the part about well above normal temps??? Both below normal snowfall and well above normal temps were the 2 major contributors to my grading the winter as an F-.

[40/70 Benchmark]

9 minutes ago, Great Snow 1717 said:

Uhmm did you not read the part about well above normal temps??? Both below normal snowfall and well above normal temps were the 2 major contributors to my grading the winter as an F-.

Oh, your subjective grading of the winter.....ie your experience of it was unfulfilling. I hear that. I thought you were speaking of my forecasting effort. I would give the winter a D from my perspective...I have seen worse, as frustrating as it was. I would honestly take this past winter over 2009-2010...that was much more frustrating for me. I would take the 2009-2010 synoptic landscape again...but not the result. lol

[Great Snow 1717]

8 minutes ago, 40/70 Benchmark said:

Oh, your subjective grading of the winter.....ie your experience of it was unfulfilling. I hear that. I thought you were speaking of my forecasting effort. I would give the winter a D from my perspective...I have seen worse, as frustrating as it was. I would honestly take this past winter over 2009-2010...that was much more frustrating for me. I would take the 2009-2010 synoptic landscape again...but not the result. lol

I had exceedingly low expectations going into the winter so it wasn't unfulfilling to me. It met my expectations. 

[40/70 Benchmark]

2 minutes ago, Great Snow 1717 said:

I had exceedingly low expectations going into the winter so it wasn't unfulfilling to me. It met my expectations. 

Yea, I mean in an absolute sense....meaning not how you would have liked it to turn out. I do remember you being on the meager train.

[tamarack]

2 hours ago, Great Snow 1717 said:

Uhmm did you not read the part about well above normal temps??? Both below normal snowfall and well above normal temps were the 2 major contributors to my grading the winter as an F-.

If I graded only on temps, my grade would also be F, as DJFM here managed to eclipse the horrid 2015-16 for warmest in our 25 years here.  However, snow total alone would rate a B/B- and the AN pack plus December bomb kick it up a notch or 3.  And the irony of the AN snow/record warmth in the same winter isn't lost on me.  Both the weirdness and the firewood savings add to my satisfaction.

[40/70 Benchmark]

Horrendous job on the December-March snowfall forecast this season...my worst in 9 years of doing this. However, I don't rate the forecast in its entirety as the worst. Diagnosis of the overall pattern, while leaving something to be desired, was not as poor as the great dearth of snowfall would imply. Primary issues were misdiagnosing the orientation of la nina and underestimating what turned out be a historic season Pacific cold phase.
Give this one a "D".
Time to flip the page and focus on the developing el nino-
https://easternmassweather.blogspot.com/2023/05/winter-2022-2023-seasonal-review.html

[40/70 Benchmark]

Winter 2022-2023 Seasonal Review & Forecast Verification; Tamer Than Expected Results

Bookend Blocking Negated by Near Record Cold Phase of Pacific

Month of December Largely Well Forecast

December 2022 Review

December Analogs: 1970, 1971, 1973, 1975, 1984,1992, 1996, 2005, 2007, 2008, 2013, 2016 

 

The month of December was forecast remarkably well with respect to the Pacific, as a predominately negative phase of the PNA prevailed, which favored cold loading to the west.

 

 

In fact, the December PDO value of -2.21 was the fourth lowest value for the month of December on record dating back to 1950, only exceeded by December 2021, 1994 and 1990.
Thus given the historic nature of the Pacific pattern it is not surprising that precipitation on the west coast was more prolific than the sensible weather composite had indicated for the month of December:

 

Actual: 

 

 

Forecast Composite:

 

 

 

This highly anomalous surplus of precipitation on the west coast is of course very reminiscent of the -PNA style extra tropical Pacific pattern, which featured 3/5 primary ENSO analog seasons in 1955-1956, 1970-1971 and 2010-2011, respectively. The one notable difference being its placement over California instead of the PNW, which will be addressed in a moment.

 

 

 

 

It is fair to say that snowfall over southern New England was lighter than anticipated due to the nature of the Pacific pattern, however, in general the forecast narrative for the month of December favoring the deep interior and elevations, as well as northern New England for snowfall was correct given the historic prominence of the cold phase Pacific pattern (-PDO). 

 

"The month will gradually turn colder as the high latitudes become more disturbed, and while the mid Atlantic may find snowfall scarce, New England should not...especially north and interior. However, coastal areas/lower terrain may have frequent precip type issues during coastal storms due warm SSTs. Some elevation events are certainly possible. There will also be southwest flow events/warm air advection events, which will be more fruitful for the coast in terms of snow. A white Christmas is likely across the interior and northern New England, and possible along coastal southern New England. It is very unlikely for at least the lower terrain of the mid Atlantic". 

 

 

This is in large part due to the fact that the early season high latitude blocking that was anticipated materialized even more fervently than forecast, otherwise the majority of storms would have tracked into the mid west and great lakes.

The was a slight deviation from the forecast, as the most anomalous episode was postulated to instead take place in January.

 

 

 

However, as illustrated in the above annotation, clearly the massive episode of blocking materialized during the month of December instead of January. While this did not affect the December portion of the forecast much for the primary area of focus here in southern New England, it did mean a slightly cooler outcome relative to the expectation across the mid atlantic and southeast, with a slightly warmer outcome than forecast across northern New England, despite near climo snowfall for those locales.

 

December Temp Anomaly Forecast vs 1991-2020:

Actual December Temp Anomalies:

 

 

The stronger than anticipated high latitude blocking during the month of December is represented very clearly in the H5 anomaly plot:

 

Versus the more modest degree of high latitude blocking present in

 the composite December forecast, which

 allowed for some southeast ridging and thus a slightly milder outcome across the southeast:

 

The historic magnitude of the PDO, which favored very anomalously low heights across the western CONUS, undoubtedly kept snowfall scarce across the southern New England forecast area. This especially played out with respect to the major storm threat just prior to Christmas, as the displaced polar vortex was forced to phase with the storm system well to the west, underneath the NAO block, which resulted in a mild rain storm for the region. 

 

50mb H5 (Stratosphere)

This nuance forced a very wintry ostensible pattern into a not so wintry outcome for the region and would set the tone for the mild winter of 2022-2023.

Although the month of December featured even greater high latitude blocking than expected, the opposite was true after the New Year, as a shift in the tropical Pacific accompanied a consolidation and intensification of the polar vortex sooner than anticipated. This rapid easing of high latitude blocking after the new year was evident with respect to the solar/cool ENSO analogs, as well as some of the seasonal guidance.

Very Active January Much Milder Than Forecast Due to a Shift in ENSO

 

January Analogs:1971, 1976, 1985, 1986,1996 2011, 2014, 2021, 2022

The month of January verified several degrees above average across the eastern half of the US:

 

 

 

This was clearly much warmer than the January forecast, which was near depicted near normal temperatures:

 

One reason for the terrible January forecast was the very active Pacific jet stream, which although it was expected to be prominent at certain points over the course of the season and thus supply a very active storm track, it was more of a factor than anticipated because of the notable absence of the polar jet. This is at least partially attributable to earlier than expected changes with respect to the character of la nina, which acted to augment the recovery of the polar vortex faster than anticipated.

 

Here is the la nina event entering into the month of December:

 

 

Note the basin wide character of la nina, which was less resistant to the window of polar vortex disruption and subsequent episode of high latitude blocking that occurred early in the season.

 

DM H5 Basin Wide La Nina

 

 

This of course played out as a major early season disruption to the polar vortex during the latter portion of November and into the month of December 2022, as previously alluded to.

 

December 2020 50mb Stratosphere Height Anomalies

However, the translation of the la nina event westward was very evident by the new year.

 

 

Which is a trend towards a configuration that is very resistant to the redevelopment or maintenance of the mid season high latitude blocking that was forecast.

 

 

Note that this also favors a flatter Aleutian ridge, which makes it more difficult for any cold loading over the western CONUS to bleed east and in conjunction with a strengthening PV favors a very mild weather pattern.

Accordingly, this coincided with an extended period of mid winter intensification of the polar vortex.

 

This is apparent when looking at the actual H5 composite versus what actually transpired.

 

Cold Unavailable for Active January Storm Track

 

It is very evident from the Eastern Mass Weather January 2023 H5 forecast composite that the polar jet was expected to interact with the aforementioned Pacific parade of storm systems to at least some degree.

 

However, the early season high latitude blocking never rematerialized as forecast. While there was in fact cold air available, as evidenced by the mean negative AO value for the month.

 

 

 

It was largely concentrated on the other side of the globe given that the polar vortex consolidated over Siberia.

 

January 50mb H5 (Stratosphere)

 

And whatever cold managed to spill into the western hemisphere leaked westward, given the persistent western trough owed to a western biased +PNA working in tandem with the -PDO. Given the dearth of poleward Aleutian ridging present mid season, there was no mechanism to drive whatever cold was available to the east.

 

 

 This pattern also acted to pin the storm track inland , thus it is easy to discern why much of the east coast experienced little to no snowfall, despite such an active storm track. It was only the higher latitudes of northern New England and the higher elevations above about 1,000 feet where snowfall was fairly abundant, given the greater correlation to precipitation than temperatures in said regions.

 

 

 In fact, much of the region received in excess of 5" of precipitation with very little snowfall, especially across the southern half of the southern New England region.

When considering why that was the case it is important to remain mindful of the fact that nothing within the global atmosphere occurs in a vacuum. It is apparent from the January 2023 annotation above that the NAO remained decidedly positive throughout the month of January.

 

 

This combined with some subtle nuances with respect to the Pacific pattern acted to ensure that storms tracked inland from the coast.  And perhaps more importantly, trap the main hemispheric reservoir of cold in Siberia, thus preventing it from being entrained into the active parade of storm systems. 

Here is the PNA plot for the month of January:

While it did in fact average positive, as forecast, the higher heights were focused in the western portion of the PNA domain and off of the west coast, which only served to augment troughing along the west coast. This configuration largely mimicked the RNA pattern of December. This is why the active pattern normally present across the Pacific northwest during a negative PNA extra tropical Pacific pattern, which was discussed in the winter outlook this past fall, was actually displaced to the south, over California.

 

 

Unlike December, January featured a +NAO, which worked in conjunction with the western biased +PNA to ensure an inland storm track across the east. And given the fact that there was no Aleutian ridging or cross polar flow to tap the cold supply in Siberia, there was very little antecedent cold to allow for southwest flow events or front end snowfall for the coast. 

 

January 2023:
 

 

 

This is not at all unlike the evolution of January 2002, which was a double weighted year in the seasonal analog forecast composite, but clearly not emphasized enough in the final  forecast product.

 

January 2002:

Thus while the month did feature some el nino like qualities, as anticipated, the cold air was never able to link with the active storm track. 

The latter was an unexpected occurrence during the majority of the month of January, however, it was not at all unexpected during the month of February.

Mild & Well Forecast February Both Aloft & At The Surface

February Analogs: 1955, 1956, 1975, 1976, 1984, 1986, 1989, 1990, 1997, 2001, 2008, 2011,2017, 2018, 2022

 

 

While the transition to a modoki la nina took place earlier than forecast for the month of January, its persistence throughout the month of February was entirely anticipated. The pattern during the month of February was expected to vacillate between a traditional "modoki" la nina like pattern, which is very mild over the east, and a gradient type of pattern. The latter features a pronounced reservoir of cold over the source region despite a lack of blocking, similar to the 1973-1974, 1975-1976, and 2007-2008 seasons. This is precisely what took place.

Note the similarity between the MJO filtered forcing from the month of February and the modoki la nina like regime.

 

This makes sense given that the la nina had shifted into more of a modoki event by this point in the season.

 

 

 

Note that the early season orientation was more basin wide:

 

And the mid season progression towards that of a modoki event:

 

 

And accordingly, the forcing took on a similar appeal:

 

 

 

Here is the February H5 anomaly plot, which is unsurprisingly similar to both the modoki la nina and February forecast composites:

 

 

Versus the forecast February H5 composite plot:

 

Note the similarities between the actual and forecast save for two differences. Firstly, while there was not expected to be any high latitude blocking this month, the polar vortex ended up even stronger than anticipated, which was likely due to the mid month SSW that precipitated the anticipated changes for the month of March. Secondly, heights were lower over the southwest than anticipated, likely due to some variations in the western Pacific.

 

Here are the temp anomalies for the month February....the stout positive AO/NAO is apparent via cooler anomalies over NNE and greater positive temp departures over the SE, which is opposite of the blockier December:

 

 

 

Versus the February forecast:

 

 

Note again that while the east coast is mild, as forecast, the southwestern US was cooler than forecast. It is also evident by the slightly warmer departures over NNE that the PV likely strengthened a bit more than the forecast composite implied.

 

Finally, here are the actual precipitation anomalies for the month of February:

 

 

 

And the forecast February precipitation anomalies:

 

It was drier over the PNW and northeast than anticipated.

The monthly teleconnections were all fairly well forecast.

 

The AO and NAO both averaged positive:

 

 

 

The extremely positive AO value is not only a reflection of the modoki la nina pattern, but also due to the flex of the polar vortex that often takes place during and in the immediate wake of a major disruption, which took place mid month.

Note how consolidated the vortex was in the vicinity of Greenland at the outset of the month of February:

 

And became stretched and displaced as it retreated towards Europe in a greatly weakened state due to rapid warming of the polar stratosphere during first half of the month:

 

Which culminated in the vortex being completely displaced by said warming during the second half of February:

 

 

 Since the initial stage of the disruption manifested itself in the form of a stretched vortex, latter February initially only turned colder across the north. This was the "gradient pattern" that was referenced with respect to February. However, the the warming over the polar stratosphere continued propagating down into the troposphere.

 It is this warming during the middle and latter portion of February that set the stage for the prolonged period of high latitude blocking that was forecast by Eastern Mass Weather during the month of March.

And thus while Eastern Mass Weather was incorrect in hedging against an official SSW, which did in fact occur, the forecast of a major disruption of the polar vortex during the month of February was correct.

 

"Obviously confidence at this range is modest at best, but if there is to be another major disruption of the polar vortex and a return to high latitude blocking, it will be this month. And while a SSW warming event is unlikely, if there is one to occur it will be from latter February into March and it will be a race against the ascending sun angle for winter enthusiasts to reap the benefits."

 

Low Western heights Temper High Latitude Blocking & Active Storm Track During March

March Analogs: 1956, 1962, 1965, 1971,1975,1981,1984, 1996, 2001, 2006, 2013, 2018

 

The Eastern Mass Weather forecast for a very active and more wintry month of March had been well advertised dating back to the release of the winter outlook last fall. By this point in the season, the modoki la nina was decaying quite readily:

 

 

This may have primed the atmosphere to be less resistant to the major and protracted period of high  latitude blocking that would ensue as a direct result of the stratospheric warming that erupted over the polar region during the month of February. Whereas the December period of blocking ended up being more fleeting and less effective than was originally modeled, potentially due to the transition of a still flourishing la nina into the modoki configuration.

 

 The parallels drawn to the analogs of March 1956 and 2018 were in fact valid in the sense that there was in fact a second major disruption of the polar vortex.

 

 

 

 

And this did in fact lead to another period of major late season, high latitude blocking that ultimately produced a major mid month winter storm that resulted in over 3 feet of snow across the hills of northern Worcester county and the Berkshires. This did in fact occur within the identified March 1-15 window for a major winter storm from last fall.

 

 

 

However, as evidenced by the 500mb chart above, the NAO block was largely offset by extremely low and persistent troughing over the western US. In fact, whereas the December PDO reading of -2.21 registered as the 4th lowest since 1950, the March reading was even more extreme, -2.35. The only season since 1950 to register a lower PDO value for the month of March occurred during the primary analog year of 1956 (-2.93).

 

 

 

 

While lower western heights attributable to residual cool ENSO influence were anticipated, this was clearly more extreme than forecast. 

This obviously resulted in lower heights over the western US, and slightly elevated heights over the eastern US relative to the forecast, which may have played a role in the delayed phase of the major mid month winter storm that greatly mitigated snowfall impact over the coastal plane relative to expectation.

 

Eastern Mass Weather Forecast H5 Composite

 

Unsurprisingly, the March 1956 analog was plagued by the same issue, however the RNA pattern was not nearly as prominent, which negated the fact that March 2023 featured even greater NAO blocking than 1956.

 

 

The month or March was never expected to be as cold as March 1956 or 2018 across the eastern US due to the fact that the PV was displaced onto the other side of the globe.

 

But the month still verified milder than forecast over the eastern US due to persistence of very low western heights, which actually encompassed the western half of the AO domain space and also acted to bias the AO positive for the month.

 

 

 

 

 

 

March temp forecast near normal for region:

 

 

Reality was a few degrees above average given the increased amplitude of the western US trough relative to the forecast, which was the theme of the season:

 

 

 

The precipitation forecast was decent in reflecting near normal precipitation across the region:

 

 

But the active subtropical jet that has been pummeling California was missed, as was the case during the first half of the winter. 

 

 

 

This is because the historic amplitude of the western CONUS trough displaced this feature further to the south than would normally be expected during a -PNA type of extra tropical Pacific pattern, as discussed in the winter outlook this past fall.

This is what also took place in association with the first about of historic east coast troughing from December into January. But since the NAO blocking was stronger and less transient in March relative to December, likely due to previously referenced changes in ENSO orientation, significant snowfall did occur closer to the coast in March with more excessive totals over the higher terrain and NNE.

 

 

 

 

December-March 2021-2022  Composite Verification & Summarization

Here is the H5 forecast composite for the December 2022 through March 2023 period.

 

It is measured against the protracted climo period of 1951-2010 in an effort to account for the evolution of climate change over the several decade window of time from which the analogs were derived:

 

 

Note the resemblance between the forecast composite and the -PNA extra tropical Pacific configuration data set.

 

 

This is because Eastern Mass Weather correctly posited that of the three extra tropical Pacific patterns, one being characterized by +EPO, one -EPO and one -PNA, this past season would assume the -PNA configuration. And this proved correct; perhaps event too correct. It is important to remember that this season registered one of the most intense cold PDO readings on record, with a mean aggregate DM PDO value of -1.87, which is significantly lower than the forecast range -1.27 to -1.57. Additionally, while the DM mean PNA value of -.68 was essentially within the forecast range of -.35 to -.65, this was comprised of a modestly positive PNA value of .21 during the month of January that was heavily western biased, and thus only served to mimic an RNA configuration by further augmenting the western CONUS trough. Thus the practical PNA value, so to speak, was even lower. Furthermore, the DM numeric mean numerical value of -68 belies and fails to adequately convey just how anomalous the aggregate western CONUS trough was over this past winter season.  This is illustrated well by the actual 2022-2023 DM H5 chart

 

Note how much deeper the western CONUS trough is relative to the forecast, which was reflective of a significant cold phase of the Pacific as it was, and also the -RNA extratropical Pacific data set. Couple these historically low western heights with the fact la nina evolved into a modoki event earlier than anticipated, and it is not surprising that the Aleutian ridge was also flatter than forecast, which creates a very warm winter for the eastern CONSUS with fleeting periods of blocking.

Here is a summarization using the DM 2022-2023 series measured against the most recent 1991-2020 climo period.

 

Also take note of the fact that the area of well above average precipitation along the west coast is displaced significantly to the south of both the Eastern Mass Weather Forecast Composite:

 

And the similarly oriented RNA extratropical Pacific dataset due to the historically low western US heights:

 

Actual DM precip:

 

 

Precipitation anomalies were otherwise forecast fairly well.

 

Here is the DM 2022-2023 Eastern Mass Weather forecast temp anomaly composite, which called for about a +1 to +3F DM mean aggregate anomaly across the region:

 

 

 

Which is again similar, though slightly warmer than the preferred RNA style data set:

 

 

And the actual 2022-2023 DM temp anomaly map vs 1951-2010 climo period reflecting +5F or greater mean temp anomalies. It is apparent that the coolest anomalies are displaced into the southwest with the higher anomalies stronger and more expansive to the NW than forecast due to the extreme western trough.:

 

 

Here is the DM temp map vs 1991-2020 climo:

 

 

 

Note the similarity between DM 2022-2023 temp anomalies:

 

 

 

 And the modoki la nina DM composite temp anomalies:

 

 

This is somewhat in contrast to the primary ENSO analogs of 1955-1956:

 

 

And 1970-1971, which were both basin-wide events.

 

This discrepancy between the forecast for a basin-wide hybrid la nina event and the actual modoki la nina evolution, along with the deeper western CONUS heights than anticipated, are the two reasons why the 2022-2023 winter season was tamer than forecast for the northeast region.

The impact of la nina on the seasonal forcing is evident by the failure of the MJO wave to enter phase 8 anytime from latter November until the second week of March.

 

 

This is consistent with the dual episodes of blocking early and late in the season. 

 

December-March 2022-2023 Teleconnection Forecast

The polar domain was well forecast, however, as referenced in the seasonal review, the primary errors were an underestimation of the Pacific cold phase (PDO/PNA) and miss characterization of la nina. While the intensity of la nina was correctly forecast, it verified as a modoki event instead of basin wide. 

 

 

 It was these errors were coupled with a Rosby wave train sequence that kept the timing of antecedent polar pressure patterns out of sync with the approach of incoming storm systems that ultimately proved catastrophic for the snowfall portion of the forecast. In fact, not only did this combination result in the worst Eastern Mass Weather seasonal snowfall forecast yet, but it also led to one of the lowest seasonal snowfall totals on record for coastal southern New England down into the middle Atlantic region.

Ordinarily, save for the poor timing of polar air patterns, this pattern would not have produced such a dearth of snowfall. This is reflected by the fact that the Eastern Mass Weather teleconnection forecast, aside from the orientation of the ENSO as basin-wide instead of modoki, was fairly successful. The EPO calculating more negatively than anticipated was likely due to the mid-season western biased PNA, as there was not sufficient draining of the cold air to the south and east as is often the case in a -EPO pattern. This western biased PNA that persisted throughout much of January resulted in a largely successful seasonal PNA forecast from a statistical standpoint (.03 out of range). however, it was not as accurate from a practical perspective because this western biased PNA mimicked the appeal of an RNA pattern. This belied the fact that the seasonal PNA value, like the PDO, was significantly more negative than anticipated.

 

Index Value	Predicted '22-'23 DM   Value Range	Actual  '22-'23 DM Value	Forecast Error
Pacific Decadal Oscillation (PDO)	-1.27 to -1.57	-1.87	Biased .30 positive
Perennial North American Pattern (PNA)	-.35 to -.65	-.68	Biased .03 positive
ENSO	SON -1.0 to -1.2 EMI: .3 to .5 (slightly east-hybrid)	SON -1.0 (Verified)	Verified   +1 Modoki Error
(J-M) East Pacific Oscillation (EPO)	+.30 to +.60	-.43	Biased .73 positive
Arctic Oscillation (AO)	-.15 to -.45	-.38	Verified
North Atlantic Oscillation (NAO)	+.15 to +.45	+.23	Verified

 The PDO was forecast to be strongly negative, however, it still ended up verifying as significantly more so than forecast (.30 below range) given the historic nature of the amplitude of western CONUS troughing. This may seem curious given the rather paltry intensity of the -1.0 ONI, however, the fact that this particular cool ENSO event had become so strongly coupled with the atmosphere over a 2-3 year period, as evidenced by the protracted maintenance of moderate to strong MEI values, undoubtedly played a role in the resultant amplitude of said pattern. This rendered the stellar forecast of the polar domain largely moot. This is a change from previous forecast efforts, that excelled with respect to ENSO and struggled mightily with the polar domain.

December-March 2022-2023 Snowfall Outlook Verification

 

City	Predicted Snowfall	Actual	Forecast Error
Boston, MA	40-50"	12.4"	222.6%
NewYork, NY(Central Park)	25-35"	2.3"	987%
Philadelphia, PA	15-25"	0.3"	4,900%
Baltimore, MD	10-20"	0.2"	4,900%
Washington, DC	5-15"	0.4"	1,150%
Albany, NY	75-85"	55"	36.4%
Hartford, CT	50-60"	24.2"	106.6%
Providence, RI	30-40"	11.6"	158.6%
Worcester, MA	75-85"	47.2"	58.9%
Tolland, CT	55-65"	32.1"	71.3%
Methuen, MA	60-70"	40.5"	48.1%
Hyannis, MA	20-30"	2.3"	770%
Burlington, VT	90-100"	72.8"	23.6%
Portland, ME	85-95"	56.5"	50.4%
Concord, NH	80-90"	63.8"	25.4%

 

In summary, the two most prominent errors with respect to the 2022-2023 seasonal forecast were underestimating the historically anomalous nature of the Pacific cold phase and the misdiagnosis of the modoki nature of la nina as basin-wide instead of modoki.

 

 

 

 Theoretically speaking, this would have had the most profound impact on snowfall along the coast, as well as the Mid Atlantic region and this is precisely where the errors were largest, as reflected by the seasonal snowfall verification table above. In fact, snowfall was over forecast by anywhere from 158.6% in Providence, RI to as much as a ghastly 4,900% in Philadelphia, PA and Baltimore, MD. Across the higher terrain, interior and NNE, the error was not quite as drastic, ranging from a mere 23.6% in Burlington, VT up to as much as 106.6% in Hartford, CT. The margin of error over interior SNE ranged from 48.1% in Methuen, MA up to 106.6% in Hartford, CT.

The mean snowfall forecasting error for the 2022-2023 winter season throughout the 15 select Mid Atlantic and northeast locations was a horrendous 900.6%. This was the worst snowfall forecast amongst the nine forecast seasons in terms of mean forecasting error by a considerable margin over the 2019-2020 effort (678.3%  mean error). The seasonal forecast ranges verified in none of the cities and was closest in Burlington, VT and Concord, NH, with errors of 23.6% and 25.4%, respectively. The horrific results in the mid Atlantic region are the primary reason why this forecast struggled, where snowfall was over forecast by an unprecedented 4,900% in Philadelphia, MA and Baltimore, MD.

The mean forecasting error with respect to the nine seasonal snowfall outlooks are as follows:

2022-2023: 900.6%

2021-2022: 36.2%

2020-2021: 14.1%

2019-2020: 678.3%

2018-2019: 90.3%

2017-2018: 13.2%

2016-2017: 35.3%

2015-2016: 71.4% 

2014-2015: 10.9%

Nine season running mean error: 205.6%

29/132 (22%) of snowfall outlooks have been under forecast.

76/132 (57.6%) of snowfall outlooks have been over forecast.

26/132 (20%) of snowfall outlooks have verified within the forecast range.

Below is the verification for ENSO, as well as the various teleconnections over the past several years:

 

Index Value	Seasons Verified	Mean Forecast Error	Forecast Bias
Pacific Decadal Oscillation (PDO)	0/7 Seasons	.56	Biased Positive 4/7 Seasons
Perennial North American Pattern (PNA)	1/7 Seasons	.23	Biased Negative 3/7 Seasons
ENSO	8/9	.001 too cold .27 EMI	Biased cold 1/9 Seasons Biased west 2/5 Seasons
(J-M) East Pacific Oscillation (EPO)	3/7 Seasons	.36	Biased Positive 3/7 Seasons
Arctic Oscillation (AO)	1/7 Seasons	1.11	Biased Negative 5/7 Seasons
North Atlantic Oscillation (NAO)	1/7 Seasons	.79	Biased Negative 5/7 Seasons

 

 

The reason that the forecast is not considered a complete and utter failure is that the general conceptualization of the work offered plenty of value and insight relative to the winter that ensued. This is reflected both by the relative accuracy of the teleconnection forecast, as well as the seasonal forecast composite, which was generally accurate. But simply failed to capture the highly anomalous magnitude of the historic Pacific cold phase that defined the 2022-2023 winter season.

 

Final Grade for 2022-2023 Season: D

[Terpeast]

Nice work, @40/70 Benchmark I enjoy reading your posts. 

This was a tough one to forecast because it falls on the extreme end of the climate spectrum. Who would have guessed that JF would see +8 temps while California gets buried as record snows effectively erased their decades-long drought? 

Raindance did a good job, but even he didn’t forecast these extremes. 

Doug K down here in DCA only gave us 2-6” of snow in his outlook, and even he was too generous with that. 

Sure, the nina and +qbo, IOD, maritime forcing, etc. Were all unfavorable, but if you told me last October how this winter would transpire, I would have thought you were nuts.