Review of 2020-2021 Winter Outlook

[40/70 Benchmark]

https://easternmassweather.blogspot.com/2021/05/winter-2020-2021-seasonal-recap.html

B+

Mid winter SSW only real wrinkle.

Winter 2020-2021 Seasonal Recap & Forecast Verification

December 2020 Verification

Well Behaved Arctic; Problematic Pacific

The main premise of the December portion of the 2020-2021 Winter Outlook was the re emergence of high latitude blocking as a prominent winter feature.

 

The following is an excerpt from the actual Winter Outlook released in early November.

 

"There has only been one winter (DM) month, March 2018, that has averaged a negative NAO since February and March 2013. It is postulated that there is a high likelihood that we will add to that list this season. Most likely in December, but possibly March or even both. "

-Eastern Mass Weather 11-5-20

 

This was an especially bold call considering the state of long range guidance at the time, which was suggesting the development of a rather formidable polar vortex once again. 

However it worked out reasonably well, as the NAO averaged at least modestly negative.

 

 

And the polar vortex underwent consistent assaults, as evidenced by the extremely negative AO.

 

 

Guidance ultimately adjusted, and the consensus then scrambled to follow suite.

Although the forecast worked out fairly well locally, there was some limitations with respect to the Pacific pattern that were of a much larger detriment to the national forecast.

 

Here if the forecast temperature anomaly map along the with attendant H5 chart from early November:

 

 

 

 

 

"The month of December should average 1-2 degrees below normal for northern New England, near normal to a degree above normal for central southern New England, and 1-2 degrees above normal from north to south near New York City points southward into the mid atlantic. Cold will dump over the high plains and bleed eastward".

-Eastern Mass Weather 11-5-20

 

The temperature forecast verified locally across southern New England and the Boston area.

 

 

(Courtesy of Eric Fisher via Twitter)

 

But there was a significant RNA pattern forecast that is quite typical of significant la nina, but what actually verified was a formidable PNA that is more representative of an el nino. This in conjunction with lower than expected heights near Alaska wreaked havoc with the national temperature pattern. 

 

 

 

 

 

Note that while the forecast was for cold centered in the northern plains and warmth to the south, which is a thermal gradient that is typical of la nina, the opposite actually verified. Due to the PNA pattern that materialized, cooler anomalies verified to the south with warmer anomalies to the north, which is observed more frequently during el nino events. Moreover, the warmth over the northern plains was augmented by lower than forecast heights near AK, which is indicative of the positive phase of the EPO. This results in a modifying Pacific jet that eradicates the cold reservoir across Canada and the northern plains.

 

 

Actual December H5 Pattern

vs Forecast December 2020 H5 Pattern

 

 

 This variation in the pattern ultimately had important ramifications on the anticipated forms of cyclogenesis, as implied in the above annotation.

 

December Snowfall

The forecast was for month was for the development of storms of fairly moderate intensity, since the forecast higher heights in the southeast would work in concert with the expected negative NAO shear amplifying shortwaves. 

 

"It should be an active period marked by above average precipitation, especially for points north of Washington, DC. Primary modes of cyclogenesis include Alberta clippers, southwest flow events, and Miller B redevelopers, the latter of which especially prevalent during periods of negative NAO. New England should be favored for snowfall, especially central and northern parts of the region. The mid atlantic will need breaks in the RNA to coincide with negative NAO flex for significant snowfall, which is not exceedingly likely, but possible. A White Christmas is more probably than climatology would suggest this season. While the month is likely to average a negative NAO in the mean, this is not meant to imply that excessive blocking will reside for the entirety of the month".

-Eastern Mass Weather 11-5-20

 

Forecast Precipitation Anomalies for December 2020

 

vs Actual Precipitation Anomalies for December 2020

 

 

The southeast ridge was not the significant factor that it was forecast to be, and one major event did in fact impact the northeast US near mid month, which is when the pattern was expected to become more active.

 

 

                                    (Courtesy Eric Fisher via twitter)
 

 

Although this system deposited greater than 12" of snowfall over a relatively large area just one week prior to the holiday, much of the area was denied a white Christmas, nonetheless. This was due in large part to some idiosyncrasies regarding the positioning of the higher heights in the NAO domain.

 

Much like ENSO Events, Location Crucial for Blocks

The month of December was not quite as active in terms of winter weather across the area as implied, in large part because the large Christmas event this past week was entirely rainfall across the region. The easy first guess would be because the NAO was not predominately negative, as expected, but this is not true as previously discussed.

 

 

A more exhaustive review of the data reveals why.

Latter December has been characterized by a "thumb ridge" of higher heights across the north atlantic to the south of Greenland, as opposed to a true block over Greenland.

 

Although technically registering as a negative NAO due to higher heights in the southern periphery of the NAO domain, this is not a block because it does nothing to "block" a storm from cutting inland. This was evidenced by the inland track of the Christmas low, however, long range guidance suggests that this will change heading into the month of January as NAO becomes more ideal. 

January 2020 Recap

What Went Wrong

There were two major flaws with respect to the January forecast.

1) The Pacific pattern was again rather poorly forecast, as the residual + PNA/+EPO pattern of December carried over through roughly the first 2/3 of the month of January.

 

 

 

This obviously wreaked havoc with the national temperature composite in much the same manner as it in December in that the gradient was inverse of what was forecast. While the anomalies in New England were fairly accurate across at leas the souther portion of the region, it was cooler to the south, and warmer the north, which was opposite of the forecast.

 

"Temperatures should be near normal to 1 degree above across northern New England, 1-3 degrees above normal in central and southern New England, and 2-3 degrees above in the mid atlantic. Any storm activity will entail large precipitation type issues near the coast, with the most snow reserved for northern New England. The RNA pattern should only strengthen, as overall storm activity wanes".

 

-Eastern Mass Weather 11-5-20

 

Here is the forecast temperature anomaly composite:

vs the the actual:

 

 

Note that while locally throughout southern New England the forecast managed to succeed, it suffered nationally as it did in December, due to the poorly forecasted December PNA that lingered throughout the first 2/3 of January. 

 

2) The second main issue with the January portion of the outlook was that early season high latitude blocking regrouped after a brief relaxation around the holidays,  instead of relenting entirely, as forecast. The blocking was likely reenforced and protracted by the early January Sudden Stratospheric Warming event, which was not forecast given the westerly QBO and significant cold ENSO event in place.

 

 

 

 

 

 

 

It is evident in the forecast composite that while some blocking was forecast to persist into the new year:

 

It clearly remained more prevalent than forecast, as evidenced by the actual monthly composite:

 

While this portion of the season outlook was certainly flawed, it was not without its strengths.

 

What Went Right

The month of January was forecast to see lower heights build into the Alaskan region as part of a positive East Pacific Oscillation, and as illustrated in the verification composite above, the positive heights that persisted throughout January remains in place. The Aleutian ridge relocated to the southeast, away from Alaska, as forecast. This led to an active Pacific jet, which made it very difficult to time phasing well enough for any major northeast storms, as is often the same in moderate to strong la nina events. In fact, thus mid winter lull was very well forecast.

"A lull in winter for the northeast, especially below the latitude of northern New England and east of New York state, should develop during the month, as storms becomes less frequent and the predominate track shifts inland. Some seasonal cold bouts should be sufficient for some mountain snow, and perhaps some "front end" wintery precipitation further to the south".

 

-Eastern Mass Weather 11-5-20

 

Here was the forecast temperature composite, which depicted a dearth of precipitation along the east coast:

vs reality:

 

 

Note that both the active Pacific jet across the Pacific northwest, as well as the fairly quiet conditions along the east coast were both well forecast. The storm track did indeed feature multiple mixed precipitation events and rain events in the area during what was a well below normal month snowfall wise with most location measuring single digit monthly totals. However, the tid quickly turned to begin what was expected to be a very mild month of February.

 

February Recap

Arctic & Atlantic

The supposition going into the season was that any major disruption of the polar vortex would occur in latter February or March, after a very mild February. However, said disruption occurred in early January, which significantly altered the February landscape from what was forecast. What was forecast to be a rather benign and very mild month

 

Eastern Mass Weather February 2021 Forecast Temp Departures

 

 began with the capstone event of the winter on the first of the month, which set the tone for what was an average month temperature wise with above average snowfall. Thus the residual impact of the January Sudden Stratospheric Warming (SSW) continued the theme of both delaying and muting the anticipated warm up during the second half of the winter season. The anticipated positive AO during the month of February obviously did not work out, although it has finally since recovered. The steep decline centered around mid month coincides with the arctic plunge into the deep south.

 

Accordingly, the NAO was also slower to recover than anticipated, but did manage to average slightly positive for the month. In fact, the NAO was actually neutral at the time of the mid month arctic outbreak, which combined with the arctic and pacific pattern to focus the severe mid month arctic outbreak in the nation's mid section. While the impact of the polar vortex disruption significantly impacted the forecast for the arctic region, the Pacific was much better forecast, which is a reversal from the first half of the season. 

Pacific

The month overall in the mean featured a modest positive PNA, which was actually well forecast. And as can be noted in the annotation below, the peak of the monthly positive PNA spike coincided with the arctic plunge mid month.

 

 

Thus the higher heights over the western US worked in conjunction with the relative dearth of a vehicle to drive the cold east, given the neutral NAO, to focus the wrath of this outbreak over the nation's mid section. This is very evident in the monthly temperature departure chart, which is obviously quite a bit different from the forecast from last fall, posted above. This is due to the misdiagnosis of the arctic region attributable to the major mid season SSW.

 

 

Actual February 2021 Monthly Temp Departures

There was also an active Pacific forecast during the month of February, as evidenced by the positive precipitation anomaly over the pacific northwest in the forecast anomaly chart from last fall, pictured below.

 

This active Pacific jet was well forecast:

 

 

Eastern Mass Weather Forecast February 2021 Monthly Precipitation Anomaly Chart 

However, the intense period of high latitude blocking owed to the SSW, in conjunction with the well forecast modest + PNA resulted in a storm track that was more suppressed than anticipated. This focused snowfall more across southern New England and the northern mid atlantic, rather than northern New England, as expected. 

 

 

 

Actual February 2021 Monthly Precipitation Departure

 

March 2021 Review

Teleconnections

Following the deviation from the forecast in February, owed to the mid winter Sudden Stratospheric Warming protracting the blocking from the first half of the season, March went essentially according to plan. Any residual high latitude blocking relinquished its grip by mid February, paving the way for a very benign and unremarkable weather pattern across the region. The supposition from the winter outlook was that the reemergence of blocking during the month of March, which was implied by the forecast composite, would materialize more slowly than forecast and likely be delayed until April, much like last season. This assertion was based on both the +QBO, as well as the proclivity for moderate or stronger la nina events to truncate the winter season.

This forecast assertion as proved correct, as the polar domain remain relatively hostile to sustained winter weather for the vast majority of the month. 

 

 

As evidenced in the annotation above, some degree of blocking does look to reestablish itself during the month of April. This likely would have happened sooner had the la nina remained weaker, and/or the QBO were less resistant to high latitude blocking. Another implication of la nina remaining at moderate intensity into March is that the RNA pattern did not relinquish its grip as much as it could have if the cool ENSO event had either remained weak, or weakened at a faster rate.

 

Although the largely hostile polar domain and the lingering la nina did lead to relatively mild and quiet month of March with a great dearth of snowfall, it was not exotically warm in the mean, given that Aleutian ridge did indeed extend more poleward this month, as forecast.

 

Composite Verification

Here is the H5 composite for the month of March 2021:

 

 

Actual March 2021 H5 Composite

 

Note that while lower heights were prevalent both across the higher latitudes, as well as the western CONUS,  the Aleutian ridge did extend more to the NW and poleward, as forecast.

 

 

Eastern Mass Weather March 2021 Forecast H5 Composite

 

This acted to partially negate the modifying influence of the lower heights across higher latitudes and the the western US, which resulted in approximately a +1F monthly departure from the mean across the region.

 

 

Actual March 2021 Temp Anomalies

 

This was near the higher end of the forecast range for near normal monthly departures.

 

 

 

Eastern Mass Weather Forecast March 2021 Temp Departures

 

Finally, the month features below average precipitation, which worked in concert with the slightly above average temps to leave the region essentially void of monthly snowfall.

 

 

Actual March 2021 Precip Anomalies

 

These dry conditions were also congruent with the forecast expectations from last November for month of March.

 

Eastern Mass Weather March 2021 Forecast Precip Anomalies

December-March Season

ENSO: The forecast issued on November 5, 2020 was for a mixed-type, moderate strength la nina to peak anywhere between -1.3 to -1.5c during the NDJ tri monthly period. However, while this particular vent did peak with the forecast range (-1.3c), it peaked during the OND period, which is somewhat atypical for mixed-type events. This event did indeed verify as a mixed-type event, per both the location of maximum SST anomalies:

 

And the resultant forcing scheme, as evidenced by aggregate DM Vertical velocity potential:

 

Here is the mixed-type SST, with the max anomaly centered between 135-140 degrees longitude, and forcing composite as a point of reference.

 

 

The attendant forcing regime also very similar to this past season:

 

 

Here is the forecast DM H5 composite issued on November 5, 2020:

 

 

vs DM Reality:

 

 DM ForecastTemp Anomaly composite issued on November 5, 2020:

 

vs DM Temp Anomaly Composite in Reality:

 

The primary difference is that a moderately robust +AO regime was forecast (+.55 to +.85) for the 2021 winter season, however, it verified as a moderately negative AO (-.83) in large part due to the unexpected SSW event in early January. This mid season SSW was an anomaly considering that it occurred during a relatively robust la nina during a westerly QBO, and was thus not accounted for in the forecast. Said SSW extended and reinvigorated the high latitude blocking that plunged record cold and snows into the deep south across the center of the country in February. The implication of this was obviously to center the warmth across the north, rather than the south, as expected. But it also weighted down the NAO slightly lower than expected, since there is some shared domain space between the NAO and AO, thus what was forecast to be a slightly biased +NAO (+.25 to +.55) verified modestly negative at -.14. While not a major difference, this may have aided in the shift of the seasonal snowfall maxima from the forecast point in northern New England, to southern most New England and the northern mid atlantic. This is especially true in the second half of the season, when the compression between the depressed heights in the mid latitudes due to the exotically negative AO, and the higher heights to the south (RNA) sheared out storm systems during their journey through the northeast. Another contributing factor to the displacement of the snowfall maxima to the southwest of forecast was the unanticipated early season PNA joining with the NAO blocking to prevent the major coastal storm from riding inland. Although the PNA was well forecast in the aggregate, it was anticipated to be biased negative early in the season, and positive during the second half. The opposite was the case. The early season PNA coupled with the expected high latitude blocking to suppress the storm track, and then as the RNA became established during the mid winter period it created a shearing influence. The lower heights that prevailed in the vicinity of Alaska, which left the country void of cold in the absence of major high  latitude blocking, was well forecast.

The outlook overall was one of the strongest, and perhaps the most resounding success of all seven Eastern Mass Weather efforts. The one major weakness in accordance with the rest of the industry continues to be the polar domain. While an exotically strong polar vortex demolished the 2019-2020 outlook, this past season saw an unanticipated mid winter SSW weaken the polar vortex enough to significantly mar the forecast especially from a national standpoint. This despite a fairly strong effort with respect to prognosticating the NAO itself, as the prediction for the first negative NAO since March 2013 to occur this past December proved correct.

Finally, here is the forecast precipitation anomaly composite from this past November.

 

Versus the actual DM 2020-2021 precipitation composite:

 

 

 

 

It was a significantly drier season than expected, which is probably at least partially attributable to the aforementioned -AO/-PNA coupling, as this had both a tendency to shear and compress storm activity. The max snowfall area across from northern PA and southern NYS, into the northern mid atlantic and southern most New England was in a belt of latitude both availed of the higher latitude blocking, yet was still far enough north that the active Pacific jet was not prohibitive to significant snowfall, as it was across the lower mid atlantic.

December-March Snowfall Outlook

 

City	Predicted Snowfall	Actual	Forecast Error
Boston, MA	40-50"	38.6"	Snowfall Overforecast by 3.5%
NewYork, NY(Central Park)	11-21"	38.6"	Snowfall Underforecast by 45.6%
Philadelphia, PA	11-21"	23.9"	Snowfall Underforecast by 12.1%
Baltimore, MD	10-20"	10.9"	Verified
Washington, DC	5-15"	5.4"	Verified
Albany, NY	52-62"	54.6"	Verified
Hartford, CT	40-50"	41.2"	Verified
Providence, RI	25-35"	31.2"	Verified
Worcester, MA	50-60"	77.7"	Snowfall Underforecast by 22.4%
Tolland, CT	40-50"	62.7"	Snowfall Underforecast by 20.3%
Methuen, MA	50-60"	51"	Verified
Hyannis, MA	15-25"	15.6"	Verified
Burlington, VT	80-90"	63.1"	Snowfall Overforecast by 21.1%
Portland, ME	70-80"	41.3"	Snowfall Overforecast by 41%
Concord, NH	65-75"	59"	Snowfall Overforecast by 9.2%

 

Note that the only significant deviations from the forecast were across the south western portion of New England and the northern mid atlantic, where snowfall was under forecast by between 20.3 (Tolland, CT) and 45.6 percent (Central Park, NY), respectively. And northern and especially northeastern New England, where it was over forecast anywhere from 9.2 (Concord, NH) to 41 percent (Portland, ME). 

The mean snowfall forecast error of 11.7% is the lowest of the 7 annual forecasts issued. 7/15 forecasts issued verified this past season. This makes the snowfall forecast for the 2020-2021 winter season the most successful yet, narrowly edging out the 13.7% mean error of the 2014-2015 season.

 

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

 

2020-2021: 11.7%

2019-2020: 678.3%

2018-2019: 90.4%

2017-2018: 15.6%

2016-2017: 36%

2015-2016: 71.4% 

2014-2015: 13.7%

 

Six season running mean error: 131%

28/102 (27.5%) of snowfall outlooks have been under forecast.

50/102 (49.0%) of snowfall outlooks have been over forecast.

23/102 (22.5%) of snowfall outlooks have verified within the forecast range.

 

December-March Teleconnection Forecast

 

 

Index Value	Predicted '20-'21 DM   Value Range	Actual  '20-'21 DM Value	Forecast Error
Pacific Decadal Oscillation (PDO)	-1.15 to -1.45	-1.08	Biased .7 negative
Perennial North American Pattern (PNA)	-.15 to +.15	+.12	Verified
ENSO	NDJ -1.3 to -1.5 -2 to -.4 EMI	OND -1.3	Verified: One month too late
(J-M) East Pacific Oscillation (EPO)	+.65 to +.95	+.93	Verified
Arctic Oscillation (AO)	+.55 to +.85	-.83	Biased Positive 1.38
North Atlantic Oscillation (NAO)	+.25 to +.55	-.14	Biased Positive .39

 

 

 

Final Grade: B+