Winter Outlook 2018-2019

[donsutherland1]

On October 29, 2011 an early-season storm brought 2.9” snow to Central Park and more than a foot of snow in parts of the City’s northern suburbs. The remainder of the 2011-12 winter was characterized by exceptional warmth and little snowfall. Winter 2011-12 finished with 7.4” snow.

Yesterday, another historic early-season storm dumped 6.4” snow on Central Park. That is the City’s biggest snowstorm so early in the season. In addition, cities such as Allentown, Baltimore, Harrisburg, Hartford, Islip, Newark, Philadelphia, Providence, Scranton, Washington, DC, and Wilmington, DE set daily snowfall records for November 15.

Is another warm and relatively snowless winter on the way?

This time one can expect a very different evolution than 2011-12. The upcoming winter and 2011-12 have almost nothing in common.

 
Although the predominant state of the Arctic Oscillation (AO) cannot be forecast with great accuracy in advance, extreme events ahead of the forthcoming winter sometimes provide insight. Currently, the AO is forecast to drop to or below -4.000 prior to the end of November.

That is a rare case with only 1959, 1985, and 2010 having the AO reach -4.000 or below in the November 20-30 period. The December-February AO averages for those three winters were -1.510, -1.707, and -0.951 respectively.

Currently, warm SSTAs cover the Gulf of Alaska. Warm SSTAs in that region have often been a precursor of a predominantly negative Eastern Pacific Oscillation (EPO). Since October 1, 80% of days have seen a negative EPO.

El Niño conditions (Central Pacific-centered) already exist and are forecast by the guidance to persist through the winter. Therefore, my confidence concerning the state of ENSO is high. On account of forecasting limitations and sample size limitations, my confidence is moderate, not high with respect to both those major hemispheric teleconnection indices.

El Niño-Negative EPO-Negative AO winters are often cold in the eastern third to half of the United States and southern Canada.

Therefore, I expect winter 2018-19 to be colder than normal in the eastern half of the United States and warmer than normal along the Pacific Coast. December and February had the highest probability of seeing colder than normal conditions in the East. January still favored colder than normal weather in the East, but January 1959 and January 1980 wound up warmer than normal. Probabilities are not the same thing as anomalies. The composite January anomaly was the coldest of the three winter months in the East (mean temperatures). At this point in time, I suspect that Winter 2018-19 will feature a colder than normal December, January, and February in parts of the East.

Considering warm SSTAs in the Gulf of Alaska and the EPO’s recent tendency, I expect a decidedly negative EPO on average. In addition, considering the strength of the Atlantic blocking that followed the extreme late November AO cases above, perhaps the closest matches between past winters in terms of overall outcomes could be 1977-78, 2002-03, and 2009-10. All three winters have an EPO that averaged -0.25 or below and an AO that average -0.25 or below. All three winters featured much above average snowfall in the East.

Composite Maps (with double-weight given to 1977-78, 2002-03, and 2009-10):

Select seasonal snowfall estimates are below:

Albany: 70”-80”
Atlanta: 3”-6”
Baltimore: 30”-40”
Binghamton: 85”-95”
Boston: 55”-65”
Buffalo: 100”-110”
Burlington: 80”-90”
Chicago: 45”-55”
Detroit: 50”-60”
Nashville: 5”-15”
New York City: 45”-55”
Newark: 45”-55”
Philadelphia: 40”-50”
Providence: 50”-60”
Richmond: 20”-30”
Scranton: 50”-60”
Sterling: 28”-38”
Washington, DC: 25”-35”

Corrected for typo (11/26/2018): Richmond should have read 20"-30" not 30"-40". I regret the error.

[EastonSN+]

Great write up. Thanks Don.

[griteater]

Enjoyed it Don, thanks for posting.  Hope you will continue to do these.  One thing - check your surface temp anomaly maps - it looks like you need to change 2009 to 2010 in your Jan and Feb composites

[donsutherland1]

1 hour ago, griteater said:

Enjoyed it Don, thanks for posting.  Hope you will continue to do these.  One thing - check your surface temp anomaly maps - it looks like you need to change 2009 to 2010 in your Jan and Feb composites

Thanks. I fixed it now.

[Roger Smith]

Thanks for sharing that, another cold and snowy outlook for the northeast. With regard to January and my own suspicion that there's a brief relaxation of the cold during early January, I wonder if the pattern may resemble either 1933-34 or 1947-48 to some extent, with some records falling in late December and again in late January or February. It could point to good snowstorm potential on either side of that milder interval. I like the period just before Christmas for a major snowfall event. 

[LibertyBell]

2 hours ago, Roger Smith said:

Thanks for sharing that, another cold and snowy outlook for the northeast. With regard to January and my own suspicion that there's a brief relaxation of the cold during early January, I wonder if the pattern may resemble either 1933-34 or 1947-48 to some extent, with some records falling in late December and again in late January or February. It could point to good snowstorm potential on either side of that milder interval. I like the period just before Christmas for a major snowfall event. 

I was thinking 1933-34 and 1898-99, both had November and February big events (correct me if I'm wrong.)

 

[LibertyBell]

3 hours ago, donsutherland1 said:

Thanks. I fixed it now.

Don great write up, for our area I think the current historic snowstorm bears more of a resemblance to the post Sandy snowstorm which dumped double digit snowfall in Central NJ and close to it on Long Island.  I wonder why that didn't cause such disruption since it was also a big surprise and came during rush hour?  I was hoping someone could post the PNS from both I wanted to see who did better with that one vs this one- that one seemed to be more historic for the coast and occurred earlier than this one did too.

 

[uncle W]

52 minutes ago, LibertyBell said:

I was thinking 1933-34 and 1898-99, both had November and February big events (correct me if I'm wrong.)

 

1933-34 had a December/February big events...

 

51 minutes ago, LibertyBell said:

Don great write up, for our area I think the current historic snowstorm bears more of a resemblance to the post Sandy snowstorm which dumped double digit snowfall in Central NJ and close to it on Long Island.  I wonder why that didn't cause such disruption since it was also a big surprise and came during rush hour?  I was hoping someone could post the PNS from both I wanted to see who did better with that one vs this one- that one seemed to be more historic for the coast and occurred earlier than this one did too.

 

I think the snow came down faster yesterday...most of my snow fell in four hours...

[40/70 Benchmark]

1977-'78 is my main analog....agreed.

I think you are a little on the light side in Boston, though.....especially if el nino remains weak.

[Hailstorm]

Don, that is a very insightful and constructive outlook as usual.

It looks like you are predicting that the core of the cold anomalies will be centered in the OH / TN valleys.

Is there any probability that the core of those negative departures can shift more towards the Northeast if the Aluetian low / -EPO ridge adjusts eastward?

I mention this because those are very significant cold anomalies shown on the maps.

[michsnowfreak]

Good writeup Don! I will take colder than normal with 50-60" of snow at Detroit. It certainly would not be the snowiest Winter of recent years with so many good ones, however it's still solidly above the 43" average and when you throw in extra cold it could be a very white Winter, which is what I like.

[donsutherland1]

16 hours ago, 40/70 Benchmark said:

1977-'78 is my main analog....agreed.

I think you are a little on the light side in Boston, though.....especially if el nino remains weak.

I might be. I gave a some weight to 2009-10, which featured suppression for part of the winter. I don't think the blocking will be as strong as it was then. My thinking for New England is probably on the conservative side.

[donsutherland1]

10 hours ago, michsnowfreak said:

Good writeup Don! I will take colder than normal with 50-60" of snow at Detroit. It certainly would not be the snowiest Winter of recent years with so many good ones, however it's still solidly above the 43" average and when you throw in extra cold it could be a very white Winter, which is what I like.

Thanks Josh. So far, Detroit is off to a pretty good start. It has also had among its 15 wettest years on record. If it remains wet, there will be an opportunity for more snowfall than what I am suggesting.

[donsutherland1]

16 hours ago, Hailstorm said:

Don, that is a very insightful and constructive outlook as usual.

It looks like you are predicting that the core of the cold anomalies will be centered in the OH / TN valleys.

Is there any probability that the core of those negative departures can shift more towards the Northeast if the Aluetian low / -EPO ridge adjusts eastward?

I mention this because those are very significant cold anomalies shown on the maps.

Thanks Hailstorm. Exact placement of the features is subject to error. We'll have to see how things play out, but there is some possibility that the core of the cold might be somewhat farther to the East. 

[donsutherland1]

19 hours ago, uncle W said:

1933-34 had a December/February big events...

 

I think the snow came down faster yesterday...most of my snow fell in four hours...

The rate of snowfall was quite impressive. I believe there was a period where it came down at around 2" per hour. 

[AfewUniversesBelowNormal]

Nice job, Don. 

So far in the first half of this November, the Scandinavian block has been a dominant 500mb feature. 

This continues on models through November 26, making it a November leading anomaly. Since 1948, using both the + there, and - signal surrounding it, these are the top matches. 

Rolled forward to December, it's a really nice -AAM signal, with the +/- balancing act in the Pacific vs Atlantic 

January is blocking 

Dec-Feb Temps

[40/70 Benchmark]

8 hours ago, donsutherland1 said:

I might be. I gave a some weight to 2009-10, which featured suppression for part of the winter. I don't think the blocking will be as strong as it was then. My thinking for New England is probably on the conservative side.

2009-10 was a much stronger el nino....maybe more n steam this season. 1977-78 is a better ENSO and QBO analog. We will see, though...thanks for the well thought out presentation.

[michsnowfreak]

15 hours ago, donsutherland1 said:

Thanks Josh. So far, Detroit is off to a pretty good start. It has also had among its 15 wettest years on record. If it remains wet, there will be an opportunity for more snowfall than what I am suggesting.

Sounds great! Trends this decade, even this century, say unless there's high confidence in a stinker, snowy is the way to go. Detroits off to a good start with a parade of nickel and dime snowfalls since November 9th, but much more in the northwest suburbs than the southeast burbs where i live. Absolutely no complaints in mid November though. Its snowing again today.

[weathafella]

Don, thanks so much for your work with this write up!  It would seem you have the right idea here.   To me the autumn particularly October onward resembles 2002 a lot.  Good luck with the forecast and thanks again!

[donsutherland1]

1 hour ago, weathafella said:

Don, thanks so much for your work with this write up!  It would seem you have the right idea here.   To me the autumn particularly October onward resembles 2002 a lot.  Good luck with the forecast and thanks again!

Thank you, Jerry. It will be interesting to see how things proceed going forward.

[Isotherm]

@donsutherland1 - great work, and good luck with verification. There are plenty of commonalities between our forecasts, which is nice to see. I'm most interested to see the inter monthly progression, as I have thought and still do that a rather non-canonical progression is possible.

[donsutherland1]

4 hours ago, Isotherm said:

@donsutherland1 - great work, and good luck with verification. There are plenty of commonalities between our forecasts, which is nice to see. I'm most interested to see the inter monthly progression, as I have thought and still do that a rather non-canonical progression is possible.

Thank you, Isotherm. I'm really impressed with the level of detail you put forth in your outlook and how well you have done with your seasonal forecasts.

[stadiumwave]

Great job @donsutherland1

[donsutherland1]

To date, Chicago has received 12.7" snow and Detroit has picked up 6.7".

At present, everything appears to be on track for a normal to snowier than normal winter. The odds probably lean toward the latter.

To examine where things stand and where they might wind up, I constructed a snowfall profile using the following criteria for total snowfall through November 30:

Chicago (12.7" so far): 4" or more
Cleveland (4.4" so far): 1" - 10"
Detroit (6.7" so far): 2" or more
Milwaukee (6.6" so far): 2" - 12"

The following years fit that criteria (for which a common record exists for the above cities):

1893, 1947, 1953, 1955, 1959, 1974, 1975, 1978, and 1980

Mean seasonal snowfall for Chicago and Detroit was as follows:

Chicago: 47.2" (standard deviation: 17.8")
Detroit: 44.2" (standard deviation: 10.7")

I also constructed a sum of the squares errors analysis. The 5 years with a sum of the squares error < 60" were:

1893, 1959, 1975, 1978, and 2015. 

Mean seasonal snowfall for Chicago and Detroit was as follows:

Chicago: 51.9" (standard deviation: 22.3")
Detroit: 44.0" (standard deviation: 8.7")

The majority of winters cited above did not feature an El Niño. Further, 2015-16 featured a super El Niño event.

Thus, at least at this point, my thinking is that the snowfall figures will likely exceed what is described above. Therefore, I continue to have confidence that my thinking for Chicago (45"-55") and Detroit (50"-60") remains reasonable.

[michsnowfreak]

12 hours ago, donsutherland1 said:

To date, Chicago has received 12.7" snow and Detroit has picked up 6.7".

At present, everything appears to be on track for a normal to snowier than normal winter. The odds probably lean toward the latter.

To examine where things stand and where they might wind up, I constructed a snowfall profile using the following criteria for total snowfall through November 30:

Chicago (12.7" so far): 4" or more
Cleveland (4.4" so far): 1" - 10"
Detroit (6.7" so far): 2" or more
Milwaukee (6.6" so far): 2" - 12"

The following years fit that criteria (for which a common record exists for the above cities):

1893, 1947, 1953, 1955, 1959, 1974, 1975, 1978, and 1980

Mean seasonal snowfall for Chicago and Detroit was as follows:

Chicago: 47.2" (standard deviation: 17.8")
Detroit: 44.2" (standard deviation: 10.7")

I also constructed a sum of the squares errors analysis. The 5 years with a sum of the squares error < 60" were:

1893, 1959, 1975, 1978, and 2015. 

Mean seasonal snowfall for Chicago and Detroit was as follows:

Chicago: 51.9" (standard deviation: 22.3")
Detroit: 44.0" (standard deviation: 8.7")

The majority of winters cited above did not feature an El Niño. Further, 2015-16 featured a super El Niño event.

Thus, at least at this point, my thinking is that the snowfall figures will likely exceed what is described above. Therefore, I continue to have confidence that my thinking for Chicago (45"-55") and Detroit (50"-60") remains reasonable.

Thanks for the analysis as always don. Detroit averages about 6" more than Chicago on average, and recent winters Detroit has far been out snowing Chicago. With all the talk of an East Coast Winter this year I was thinking it would be another year where Detroit far out snowed Chicago, however Chicago is certainly off and running. November was quite an odd month in Detroit in that it snowed so often but there was nothing of consequence. Snow fell on 14 days during November including 9 days which had measurable snow. It was cloudy or mostly cloudy 27 of the 30 days. Brushing the car off was common as was waking up to light but picturesque snowfalls. Yet shoveling has not been necessary yet, and most of the snows were low ratio. It has been nice to see the flakes fly so frequently but i am definitely ready for more! Climo thru Nov 30 is only 1.6", so it really picks up steam in Dec.  I am sure plow drivers have been salivating seeing so many flakes fall but not having to plow anything yet but climo is a reminder that we are just getting started.

[donsutherland1]

8 hours ago, michsnowfreak said:

Thanks for the analysis as always don. Detroit averages about 6" more than Chicago on average, and recent winters Detroit has far been out snowing Chicago. With all the talk of an East Coast Winter this year I was thinking it would be another year where Detroit far out snowed Chicago, however Chicago is certainly off and running. November was quite an odd month in Detroit in that it snowed so often but there was nothing of consequence. Snow fell on 14 days during November including 9 days which had measurable snow. It was cloudy or mostly cloudy 27 of the 30 days. Brushing the car off was common as was waking up to light but picturesque snowfalls. Yet shoveling has not been necessary yet, and most of the snows were low ratio. It has been nice to see the flakes fly so frequently but i am definitely ready for more! Climo thru Nov 30 is only 1.6", so it really picks up steam in Dec.  I am sure plow drivers have been salivating seeing so many flakes fall but not having to plow anything yet but climo is a reminder that we are just getting started.

Thanks Josh.

I agree with you. Hopefully, December will wind up becoming a big month en route to a fantastic winter for the Detroit area.

[Roger Smith]

Just a subjective comment but I recall the autumn of 1967 producing weather like this November, quite cold with frequent light snowfalls (in southern Ontario), and the weather following that was very mild for a part of mid-December, think I remember something like high 50s just before Christmas then much colder again, and otherwise a rather cold winter (in the east anyway) with a memorable snow to ice storm in Toronto in mid-January (1968).  Not much recall of what may have happened elsewhere in that winter, there were long dry intervals in February where I was, almost daily sunshine, and another large snowfall around 12th of March, then a warm early spring developed. 

[donsutherland1]

Some updates (largely a recapitulation of thoughts expressed in regional forums but relevant to my overall winter thoughts):

Main takeaway: Winter 2018-19 remains on track to see above to much above normal snowfall in the Great Lakes Region, Middle Atlantic Region, and southern New England.

From December 17:

Today marked the 6th consecutive day that the Southern Oscillation Index (SOI) was +10.00 or above. That sets a new record for most consecutive days during which a winter month (December, January, February) had an ENSO Region 3.4 anomaly of +0.5°C or above.  The prior meteorological winter record was 5 consecutive days, which was set in February 2015. The prior December record was 4 consecutive days, which was set in December 2015.

The MJO has also been in Phase 4 with an amplitude of 2.000 or above for 4 consecutive days. No December during which the ENSO Region 3.4 anomaly was +0.50°C or above had more than 2 consecutive such days. 3 prior Decembers had 2 consecutive days meeting such criteria: December 2002, December 2006, and December 2015. December 2006 was the only December case that saw such an MJO coincide with a positive SOI.

In short, the current global synoptic setup is a highly anomalous event.

If some of the ensembles are right, this setup should breakdown over the next 7-10 days. The SOI could go negative toward the latter part of that timeframe. The MJO will likely move into Phase 5 at a fairly high amplitude toward the latter part of that timeframe. While some previous guidance had suggested a low amplitude outcome, a low amplitude appears very unlikely to verify over the next 10 days and probably longer.

The last time the MJO had an amplitude below 1.000 was October 30. To date, the MJO has had an amplitude of 1.000 or above for 46 consecutive days. During winter 2007-08, the MJO had a high amplitude for 69 consecutive days. The meteorological winter record is 78 consecutive days set during the December 13-Feburary 28, 2018 period (with the MJO finally moving into low amplitude on March 13, 2018 after 90 consecutive days at high amplitude).

The ongoing rare MJO-SOI events will lead to a lot of uncertainty for perhaps the next several weeks. Nevertheless, even as the last week of December now looks below average in terms of prospects for accumulating snow in the northern Middle Atlantic region (some but not a high possibility exists), the base case going forward appears to be the following:

1. The ocean and atmosphere will likely couple (probably in January) leading to the atmosphere's becoming consistent with the ongoing El Niño.

2. The MJO will remain at high amplitude through the remainder of December as it progresses into Phase 5 and then Phase 6. It will likely move into the colder Phases in January.

3. The warm SSTAs in the Gulf of Alaska will lead to the decay and breakdown of the ongoing EPO+, but that will probably happen during the first half of January, consistent with a similar evolution of events in winter 2002-03.

4. The PDO has gone positive in December after a somewhat negative November (based on daily data). Some caution is advised, as daily data was used. The JSAO.Washington.edu PDO page has not been updated recently (its last monthly value was +0.09 for September). A positive PDO has typically been present during snowy El Niño winters.

Note: December 18 marked the 7th consecutive day with an SOI value of +10.00 or above. December 18 was also the 5th consecutive day during which the MJO was in Phase 4 with an amplitude of +2.000 or above.

From December 18:

Classic El Niño/EPO-/AO- 500 mb pattern:

The EPS weeklies released yesterday are showing an evolution toward that pattern in coming weeks.

 

[griteater]

Enjoyed your thoughts Don, keep em coming.  I agree with your thoughts regarding the MJO moving on into the colder phases by January - https://www.americanwx.com/bb/topic/50522-mid-to-long-term-discussion-2018/?do=findComment&amp;comment=5072289

Anthony M made mention of the analog yesterday, but this winter seems to be following 2003-2004.  The MJO timing is a very close comparison, and with the SOI and MJO behavior, this winter is kind of acting like a neutral year as +ENSO isn't locked in (2003-2004 was +ENSO Neutral).  I'm curious as to why you think the ocean and atmosphere will begin to couple in January?  I could envision a similar scenario as Feb 2004 with the MJO progression, but only time will tell. 

In addition to the tropical forcing comparison, there was a SSW on Jan 5, 2004 (which looks to be similar timing as this year).

Here are charts from 2004: MJO / January 500mb and Sfc Temps / February 500mb and Sfc Temps

 

 

 

[donsutherland1]

5 hours ago, griteater said:

Enjoyed your thoughts Don, keep em coming.  I agree with your thoughts regarding the MJO moving on into the colder phases by January - https://www.americanwx.com/bb/topic/50522-mid-to-long-term-discussion-2018/?do=findComment&amp;comment=5072289

Anthony M made mention of the analog yesterday, but this winter seems to be following 2003-2004.  The MJO timing is a very close comparison, and with the SOI and MJO behavior, this winter is kind of acting like a neutral year as +ENSO isn't locked in (2003-2004 was +ENSO Neutral).  I'm curious as to why you think the ocean and atmosphere will begin to couple in January?  I could envision a similar scenario as Feb 2004 with the MJO progression, but only time will tell. 

In addition to the tropical forcing comparison, there was a SSW on Jan 5, 2004 (which looks to be similar timing as this year).

Here are charts from 2004: MJO / January 500mb and Sfc Temps / February 500mb and Sfc Temps

 

 

 

 

 

 

Thanks Griteater.

Two factors give me cautious optimism about an ocean-atmosphere coupling at some point in January:

1. The forecast for a decline in the SOI to negative levels

2. The development of a PDO+ in December (daily values)

I don't believe the coupling will be immediate, and it could be preceded by dual Atlantic and Pacific blocking.

Finally, I don't think the Southeast has seen its last snowfall for this winter. Raleigh, Charlotte, Asheville, Greensboro, etc., should see more accumulations. Indeed, I believe Atlanta will experience accumulating snow at some point. 2004, if it provides insight, would be good for eastern North Carolina. Hopefully, places like Cape Hatteras and Wilmington will get a decent snowfall at some point.