Winter 2019-2020 Idea

[donsutherland1]

During September, the ENSO Region 1+2 anomaly fell below -1.00°C for multiple weeks. Since 1950, only a single El Niño winter (2002-03) followed a September with one such week. Therefore, based on a combination of the current guidance and historical ENSO data, the underlying ENSO assumption for winter 2019-2020 is a neutral (warm) ENSO state. There could be periods where the ENSO Region 3.4 anomaly is somewhat above +0.5°C, but the winter average should be neutral. For much of the winter the Region 1+2 anomaly should be negative.

The September-October ENSO Region 1+2 anomaly was -0.82°C. Since 1950, 12/16 (75%) cases when the Region 1+2 anomaly averaged -0.99°C to -0.50°C had a predominantly negative PDO. In fact, the PDO fell to -0.45 in October. The last time the PDO was negative for a month was November 2018 when it was -0.05. A negative PDO winter favors a negative PNA.

Since 1950, there were five years that saw the AO average -1.000 or below in November and -1.500 or below during the second half of November. Out of the total of 15 months that followed, the AO averaged < 0 during 12 (80%) and -0.500 or below during 10 (67%). This data suggests at least somewhat above average frequency of AO blocking for winter 2019-2020. However, winters during which the AO rises to +3.000 or above in December typically have less frequent blocking. Therefore, the assumption is that there will be periods of blocking and periods where blocking is absent.

The NAO appears to have transitioned to a predominantly positive state. A generally positive NAO has also been modeled on the seasonal guidance. The core assumption is a positive NAO.

The large pool of warm SSTAs south of Alaska has been cooling gradually in recent weeks. Warm SSTAs in that region have often been a precursor of a predominantly negative Eastern Pacific Oscillation (EPO). Based on this trend, the first half of winter 2019-2020 may see more frequent EPO blocking than the second half.

Key Assumptions:
1. Neutral-warm ENSO
2. Generally positive NAO
3. Negative to somewhat positive PDO
4. AO variability
5. EPO variability with a tendency toward more positive values later in the winter.

A composite temperature anomaly map is below (DJF 1950-51 through DJF 2018-19):

The multi-model (C3S)outlook is below:

Additional factors to be considered include the observed ongoing warming occurring in the Arctic region.

Based on all of the above factors (including some weight being placed on the seasonal climate models), my estimated December-February temperature thinking is:

Colder than normal (1.0° to 2.0° below normal):

Eastern Canada (eastern Ontario, Quebec, Labrador, New Brunswick, Newfoundland, Nova Scotia, PEI)

Near Normal (0.5° below normal to 0.5° above normal):

Desert Southwest, Great Lakes Region, New England

Somewhat above normal (0.5° to 1.5° above normal):

Middle Atlantic Region, Plains States, West Coast, Canada (except for Eastern Canada and Northwest Canada)

Warmer than Normal (1.0° to 3.0° above normal):

Southeast

Much Warmer than Normal (More than 3.0° above normal):

Alaska, Northwest Canada (Yukon)

Select seasonal snowfall estimates are below:

Albany: 60"-70”
Atlanta: 0.5"-2.5"
Baltimore: 10”-20”
Binghamton: 80”-90”
Boston: 30”-40”
Buffalo: 100”-110”
Burlington: 80”-90”
Chicago: 25”-35”
Detroit: 30”-40”
Nashville: 5”-10”
New York City: 20”-30”
Newark: 20”-30”
Philadelphia: 10”-20”
Providence: 25”-35”
Richmond: 5”-10”
Scranton: 45”-55”
Sterling: 10”-20”
Washington, DC: 7”-17”
What could lead to higher amounts:

1. Prolonged deep blocking (AO of -1.000 or below)
2. The MJO's persistently being in Phases 7, 8, 1, and 2 at a high amplitude
3. The ENSO Region 3.4 anomaly falling to somewhat negative values as the winter progresses
4. A mainly negative NAO

What could lead to lesser amounts:

1. The development of a persistently positive AO coupled with a strongly negative SOI
2. The MJO's persistently being in Phases 4, 5, and 6 at a high amplitude
3. Dramatic warming in ENSO Region 1+2

[michsnowfreak]

Thanks as always for the write-up Don, although I will say hopefully your snowfall estimates are a bit low in the Great Lakes. Chicago is already in 8.3" on the season and you only have them in the 25-35" forecast range (they average approximately 37"), Detroit is already at 9.5" on the season and you have them in the 30-40" range, average approximately 43".

[Hoosier]

Thanks for your thoughts, Don.

 

1 hour ago, michsnowfreak said:

Thanks as always for the write-up Don, although I will say hopefully your snowfall estimates are a bit low in the Great Lakes. Chicago is already in 8.3" on the season and you only have them in the 25-35" forecast range (they average approximately 37"), Detroit is already at 9.5" on the season and you have them in the 30-40" range, average approximately 43".

 

Chicago had its 2nd snowiest start to the season when looking at snow amounts through November 15.  I ran the numbers on the 10 snowiest starts for Chicago (through Nov 15) and guess what, 5 years finished snowier than average and 5 had less snow than average.  Given where ORD is at right now, it will likely be hard to hold them down in the low end of that range but certainly a mediocre season is not only conceivable, but well within the realm of possibility.  

[40/70 Benchmark]

@donsutherland1 I think you are on the right track, Don. Very similar to my thoughts.

[donsutherland1]

7 hours ago, michsnowfreak said:

Thanks as always for the write-up Don, although I will say hopefully your snowfall estimates are a bit low in the Great Lakes. Chicago is already in 8.3" on the season and you only have them in the 25-35" forecast range (they average approximately 37"), Detroit is already at 9.5" on the season and you have them in the 30-40" range, average approximately 43".

I hope that they are low, too. Personally, I found the snowfall estimates that came up quite dissatisfying. I would prefer a notably snowier winter.

[DAFF]

Snow averages are just that, averages. Epic snows are tough to replicate. Even with a locally record start, averages will sort themselves out in time.

[michsnowfreak]

On 11/28/2019 at 8:05 AM, donsutherland1 said:

I hope that they are low, too. Personally, I found the snowfall estimates that came up quite dissatisfying. I would prefer a notably snowier winter.

 Me Too! Lol. I was just curious if you took the heavy early season snow into account with those estimates. 

[donsutherland1]

46 minutes ago, michsnowfreak said:

 Me Too! Lol. I was just curious if you took the heavy early season snow into account with those estimates. 

I took regional snowfall to date as one piece of information for some context, which added a degree of confidence to my thinking. That suggested that the numbers could be plausible. The major assumptions largely drove the forecast.

[raindancewx]

We're at 3.9" officially in Albuquerque for November 2019, and a lot of the front range cities are well above average too.

In years with at least two inches of November snow in Albuquerque, we average another 10 inches of snow from Dec-May. Historically 3:1 to see above average snow here after at least two inches in November. It's a pretty reliable indicator statistically.

If we get to 13 inches or more, that's typically bad for the NE in El Ninos. Nino 3.4 has been over 27.0C since late Sept, with the subsurface warm and the SOI negative. So I consider this an El Nino.

[Roger Smith]

I just hope it doesn't go full 1949-50 on us, extreme cold developed over my part of the world that winter and after a coldish start to winter it turned very mild in the east at least to mid-Feb before becoming more seasonable. I mention that winter because we are in a very unusual dry pattern here where the PNW meets western Canada. It appears to be breaking down a bit to cold and snowy. That was about the same transition that brought in the severe cold later in Dec 1949 and all of Jan 1950 which ran 15-20 (F) deg below normal in many parts of the west while Jan 1950 was a top ten mild month in the east. The storm track was eventually set up near Chicago running SW-NE. That lasted to mid-Feb then more of a cold pattern with coastal lows. I think we'll see at least a modified version of this in 2019-20 and a good start to winter in the east perhaps reversing form some time around or just after New Years. 

[weathafella]

Finally seeing this Don.  Excellent work!  It matches my thoughts as well.  Thank you for taking the time to put this out!

[donsutherland1]

18 hours ago, weathafella said:

Finally seeing this Don.  Excellent work!  It matches my thoughts as well.  Thank you for taking the time to put this out!

Thanks Jerry. I hope all is well with you and also that the snowfall amounts are underdone. 

[donsutherland1]

Winter Forecast Update...

Over the past 6 weeks, the ENSO Region 3.4 anomaly had averaged +0.53°C. Gradual cooling of that region is forecast for the remainder of the winter. As a result, the base case idea of a neutral-warm ENSO remains on track. At the same time, the cool anomaly in ENSO Region 1+2 has largely disappeared (-0.05°C average anomaly over the past 6 weeks). Persistent warmth in this region could have an adverse impact on seasonal snowfall in parts of the eastern United States.

The NAO had been positive for 100% of the first 15 days of December before falling to -0.069 on December 16. The core assumption of a predominantly positive NAO remains on track.

During the December 1-15 period, the Arctic Oscillation (AO) averaged +1.584 with a peak figure of +3.059on December 3. Nevertheless, a portion of the northern Middle Atlantic and southern New England regions had colder than normal temperatures on average.

Since 1950, there were 13 prior cases where the AO averaged +1.000 or above during the first fifteen days of December. Such outcomes have typically been followed by somewhat less than normal snowfall in parts of the Great Lakes Region, Middle Atlantic States, and southern New England.

These figures are modestly below those in the winter idea posted at the beginning of this thread. However, the differences are sufficiently small to suggest that the overall idea remains on track.

What could lead to higher snowfall amounts:

1. Prolonged deep blocking (AO of -1.000 or below)
2. The MJO's persistently being in Phases 7, 8, 1 and 2 at a high amplitude
3. The ENSO Region 3.4 anomaly falling to somewhat negative values as the winter progresses
4. A mainly negative NAO

What could lead to lower snowfall amounts:

1. A persistently positive AO coupled with a strongly negative SOI
2. The MJO's persistently being in Phases 4, 5, and 6 at a high amplitude
3. Dramatic warming in ENSO Region 1+2

 

[donsutherland1]

Seasonal snowfall forecast, season-to-date snowfall and anomalies (through 1/5) for select cities:

Albany: Forecast: 60"-70"; 30.2" (10.6" above normal)
Baltimore: Forecast: 10"-20"; 0.3" (3.8" below normal)
Binghamton: Forecast: 80"-90"; 28.9" (0.9" above normal)
Boston: Forecast: 30"-40"; 11.5" (0.9" below normal)
Burlington: Forecast: 80"-90"; 27.9" (1.2" above normal)
Chicago: Forecast: 25"-35"; 10.8" (0.4" below normal)
Detroit: Forecast: 30"-40"; 12.9" (0.3" below normal)
New York City: Forecast: 20"-30"; 2.5" (3.6" below normal)
Newark: Forecast: 20"-30"; 4.2" (2.8" below normal)
Philadelphia: Forecast: 10"-20"; 0.1" (4.4" below normal)
Providence: Forecast: 25"-35"; 8.0" (3.6" below normal)
Washington, DC: Forecast: 7"-17"; 0.4" (3.0" below normal)

[raindancewx]

For reference -

[LibertyBell]

On 1/5/2020 at 8:35 PM, donsutherland1 said:

Seasonal snowfall forecast, season-to-date snowfall and anomalies (through 1/5) for select cities:

Albany: Forecast: 60"-70"; 30.2" (10.6" above normal)
Baltimore: Forecast: 10"-20"; 0.3" (3.8" below normal)
Binghamton: Forecast: 80"-90"; 28.9" (0.9" above normal)
Boston: Forecast: 30"-40"; 11.5" (0.9" below normal)
Burlington: Forecast: 80"-90"; 27.9" (1.2" above normal)
Chicago: Forecast: 25"-35"; 10.8" (0.4" below normal)
Detroit: Forecast: 30"-40"; 12.9" (0.3" below normal)
New York City: Forecast: 20"-30"; 2.5" (3.6" below normal)
Newark: Forecast: 20"-30"; 4.2" (2.8" below normal)
Philadelphia: Forecast: 10"-20"; 0.1" (4.4" below normal)
Providence: Forecast: 25"-35"; 8.0" (3.6" below normal)
Washington, DC: Forecast: 7"-17"; 0.4" (3.0" below normal)

Seems like the average snowfall line runs from about Binghamton to Boston.  Time will tell if it descends south as the season progresses.

 

[donsutherland1]

Winter forecast update...

To date, the idea of a warmer than normal winter has verified quite well. Overall, in much of the CONUS, the warmth has been greater than suggested. Exceptions are eastern Canada where it has also been warmer than normal and Alaska where it has been colder than normal.

One key scenario for lower seasonal snowfall, primarily for the Middle Atlantic and southern New England areas, was a predominantly positive to strongly positive Arctic Oscillation (AO). Through January 25, the AO had been positive on 75% days and +1.000 or above on 68% days.

A second key scenario of lower seasonal snowfall was the MJO's staying disproportionately in Phases 4, 5, and 6. Of the 30 days this winter during which the MJO had an amplitude of 1.000 or above, the MJO was in Phases 4-6 57% of the time. That is well above the near 38% figure for equal frequency for each of the MJO's eight phases.

Based on the guidance and expected pattern through January 31, New York City will likely have less than 6" seasonal snow by the end of January. Since 1869, there have been 34 such cases. In 24 (71%), seasonal snowfall finished under 20" in New York city. In just 4 (12%) cases, seasonal snowfall finished at or above 30". In 7 (21%) cases, seasonal snowfall finished under 10". The mean figure was 16.7". The mean February-April figure was 13.1".

The January 16-31 AO is likely to finish at an average of +1.000 or above. Since 1950, there have been 10 such cases. Mean February-April snowfall in New York City was 13.0" (Median: 10.4"). Based on the outcomes, there would be an implied 70% probability that winter 2019-2020 would finish with less than 20" snow in Central Park, which is almost identical to the historic frequency for cases when seasonal snowfall through January 31 was less than 6".

In addition, the likely seasonal snowfall distribution through January 31, was overwhelmingly associated with much less than average seasonal snowfall in the Middle Atlantic region. There were 5 prior cases that met the following criteria for seasonal snowfall through January 31: Boston: < 20"; New York City: < 6"; Philadelphia: < 3"; and Washington, DC: < 2": 1895-96, 1918-19, 1931-32, 1949-50, and 1972-73. Mean snowfall amounts for those cases were: Boston: 24.1"; New York City: 14.4"; Philadelphia: 5.9"; and, Washington, DC: 4.2". Winter 1895-96 saw much above normal snowfall in March in Boston leading to 38.7" seasonal snowfall and in New York City resulting in 46.3" seasonal snowfall. Winter 1895-96 was the only winter in the above set to have a colder than normal January in the East.

Overall, the spatial distribution of snowfall to date is consistent with a lower snowfall scenario across much of the Middle Atlantic region.

I have adjusted my snowfall amounts for the following cities:

Atlanta: 0.5" or less (through 1/25: None)
Baltimore: 4"-12" (through 1/25: 1.8")
Binghamton: 65"-80" (through 1/25: 37.3")
New York City: 12"-22" (through 1/25: 4.8")
Newark: 14"-24" (through 1/25: 6.9")
Philadelphia: 4"-12" (through 1/25: 0.3")
Richmond: < 6" (through 1/25: 1.0")
Scranton: 30"-45" (through 1/25: 18.2")
Sterling: 4"-12" (through 1/25: 2.9")
Washington, DC: 2"-8" (through 1/25: 0.6")

What could lead to higher amounts:

1. Prolonged deep blocking (AO of -1.000 or below)
2. The MJO's persistently being in Phases 7, 8, 1, and 2 at a high amplitude
3. The borderline to weak El Niño would strengthen appreciably by the middle of February

What could lead to lesser amounts:

1. A persistently positive AO coupled with a strongly negative SOI
2. The MJO's persistently being in Phases 4, 5, and 6 at a high amplitude
3. The MJO's reaching an amplitude of 2.500 or above for an extended period of time

The amounts for all the other cities remain unchanged from the initial figures.

[Roger Smith]

I may be hopecasting but it feels like this winter is going to ease into some kind of big storm maybe in March. There wasn't much to the winter of 1992-93 before February (one fairly big storm in early December?). Energy seems to be running rather low but the models are starting to amplify. You get the feeling maybe this winter won't go quietly. But it has been one big snoozefest to date, even where I live and we get a lot of snow, it has not been all that active. We have nickel-and-dimed our way to 18" of snow cover but normal here is 24-30". 

[raindancewx]

Weather Trends 360 has a pretty good track record for seasonal forecasts - they have a much less snowy March for the NE corridor than last year. We'll see how that goes. I think the best window for NE snow is Feb 15-Mar 15, coinciding with the return of wet/cold/powerful storms into the SW, and some weak/transient blocking. We'll see. On their map, yellow is less snow than last year, and the blues are more snow than last year. The three red lines are >1", >6", and >12". 

[RodneyS]

16 hours ago, Roger Smith said:

I may be hopecasting but it feels like this winter is going to ease into some kind of big storm maybe in March. There wasn't much to the winter of 1992-93 before February (one fairly big storm in early December?). Energy seems to be running rather low but the models are starting to amplify. You get the feeling maybe this winter won't go quietly. But it has been one big snoozefest to date, even where I live and we get a lot of snow, it has not been all that active. We have nickel-and-dimed our way to 18" of snow cover but normal here is 24-30". 

Here is an interesting historical fact about two prior DC winters.  According to https://www.stormfax.com/elnino.htm, both 1913-14 and 1959-60 were ENSO neutral, just as this winter is. In the winter of 1913-14, no measurable snow fell in DC until February 13, 1914.  Then a Valentine's Eve/Valentine's Day storm dropped 4.5 inches.  That seemed to open the snow gates, as after that snow fell on an additional nine days through March 22, 1914, resulting in 28.6 inches of total snow for the season, of which a record 19.3 inches fell in March.  In the winter of 1959-60,  once again no measurable snow fell in DC until February 13th, and once again a Valentine's Eve/Valentine's Day storm dropped significant snow -- this time, 6.2 inches.  And once again, that seemed to open the snow gates, as after that snow fell on an additional seven days through March 16, 1960, resulting in 24.3 inches of total snow for the season, of which 17.1 inches fell in March -- the second most ever -- second only to March 1914.  

[Roger Smith]

Was there not a very deep low off the NJ coast around March 1-2, 1914? 

[raindancewx]

I actually consider 1913-14 to be a very weak El Nino, like this year. The issue with 1959-60 is that it had near record high solar activity, very unlike this year. For solar, it seems like the equinoxes are the peak for precipitation effects, while the solstices are the peak for temperature effects, when I look at various US sites statistically. In other words...you'd expect different snow outcomes even in a similar temperature pattern to March 1960. I will say - March 1960 was about 27.0C in Nino 3.4. If this event collapses rapidly, that's not impossible. We're only at 27.0C or so in January. So you'd just need the ocean to not warm, or to warm less than usual from Jan-Mar in Nino 3.4.

[RodneyS]

I don' t know about the deep low on March 1-2, 1914, but DC recorded only a trace of snow on each of those days.  However, it recorded 0.3 inches on March 5th, 4.0 on March 6th, 0.7 on March 7th,  5.3 on March 11th, 5.2 on March 20th, and 3.8 on March 22nd.  What is remarkable is how similar that winter was to 1959-60 in terms of no measurable snow falling until February 13th, and then way above average after that. This year, DCA has recorded on a percentage basis infinitely more snow so far than those two seasons: 0.6 inches divided by zero. 

[Great Snow 1717]

13 hours ago, Roger Smith said:

Was there not a very deep low off the NJ coast around March 1-2, 1914? 

Yes there was.  At the time it was referred to as the worst storm since 1888.

[donsutherland1]

On Saturday, January 5, I noted that the idea of a warmer than normal winter has verified quite well in most areas with some exceptions. The main focus of that message concerned the increased likelihood of reduced snowfall amounts in parts of the East relative to what had been forecast.

Now, as January concludes and two-thirds of meteorological winter have concluded, it its useful to look more closely at the details concerning the temperatures. Although the general idea of widespread warmth is on track, the magnitude of the warmth in areas has greatly exceeded the ideas cited. That includes eastern Canada, New England, the Middle Atlantic region, and parts of the Great Lakes region. One key error was the assumption of AO variability. Instead, the AO has been positive on 76% days so far and strongly positive on more than two-thirds of days during meteorological winter to date. Such an outcome in combination with a positive EPO tends to favor excessive warmth in parts of eastern North America, as has occurred.

In contrast parts of western Canada and Alaska were colder to much colder than normal despite my thinking of warmth there. In large part, the early tendency for an EPO+ (which I thought would become more likely as the winter progressed) played a key role in that outcome.

Below are the North American temperature anomalies for December 1, 2019 through January 29, 2020:

Overall, the idea of widespread warmth was a good one. But no meaningful skill was shown when it came to delineating the magnitude of the warmth. Barriers to seasonal forecasting at a level of detail exist, because some of the key variables (particularly select teleconnections) cannot yet be forecast reliably beyond two weeks. Additional variables such as the risk of sudden stratospheric warming (SSW) events (none have occurred so far this winter) that could have a large impact cannot be forecast very well, in part because the dynamics are not well-understood. While a connection between wave-breaking (particularly Wave 2 exists), most such waves neither displace nor split the polar vortex and most don't lead to dramatic warming events that propagate downward. In addition, SSW events can also be driven from bottom-up in response to sustained periods of strong Atlantic blocking.

These barriers likely explain part of the reason that the C3S multi-system seasonal forecast also significantly missed the magnitude of the warmth, even as its outlook was warmer than normal across much of North America. The three-month CFSv2 forecast for December-February from November (November 1-10 initial conditions) was more in line with CONUS temperature anomalies, but showed widespread cold in Canada and warmth in western Canada/Alaska. However, the forecast from the November 21-30 initial conditions had significantly trimmed the CONUS warmth.

 

[Roger Smith]

I think that storm in early March 1914 was more noteworthy for low central pressure and wind speeds than snowfall, must have dumped some snow inland New England but I seem to remember NYC had a record low pressure? Don must know. Another March as cold as 1960 may be a big ask, that one was top ten cold for many locations. It stayed wintry to the end of the month in the GL region.

Will speculate that Feb 19-23 is the sweet spot for any late winter heroics, should be a good set of energy peaks then, hoping the GFS is on the right track with its recent output showing a rather promising pattern developing. Of course the best part of the GFS is usually day 16.

Here's that storm on the wetterzentrale map archive, appears to have been a fast deepening coastal that moved up from east of FL on Feb 28 to east of NJ then stalled close to current JFK-ISP. 

https://www.wetterzentrale.de/reanalysis.php?map=2&model=noaa&var=1&jaar=1914&maand=03&dag=02&uur=0600&h=0&tr=360&nmaps=24#mapref

 

The winter of 1913-14 was a strange one, at Toronto they had a very mild Dec, Jan went from extreme cold mid-month to record warmth end of month, then Feb returned to record cold mid-month, then this coastal bomb. 

 

[donsutherland1]

Final December 2019-February 2020 Temperature Anomalies for North America:

[Roger Smith]

It was the 10th mildest winter in the past 180 at Toronto (unadjusted for urban heat island, if adjusted, would rank 15th). Some fairly similar winters include 1918-19 and 1932-33. So far there hasn't been much sign of a late winter recovery, seems that the most likely outcome will be a gradual transition to a warmer than average spring. I can't imagine a more boring winter than this, weather-wise, on a global basis. Even Moscow is too warm. 

[Great Snow 1717]

5 hours ago, Roger Smith said:

It was the 10th mildest winter in the past 180 at Toronto (unadjusted for urban heat island, if adjusted, would rank 15th). Some fairly similar winters include 1918-19 and 1932-33. So far there hasn't been much sign of a late winter recovery, seems that the most likely outcome will be a gradual transition to a warmer than average spring. I can't imagine a more boring winter than this, weather-wise, on a global basis. Even Moscow is too warm. 

Easily the most boring winter of my life.