Winter 2020-21 Discussion

[snowman19]

29 minutes ago, WinterWxLuvr said:

It seems that the gfs modeling is right in line with what can be typical in Ninas. At range models will send cold fronts south and east bringing shots of cold but as you get closer the cold tends to hit a wall and gets shunted east and has a difficult time making southeastward progress. The gfs had already done this once, this week, and not sure that it won’t next week too.

The models are back to showing their phantom negative AO and negative NAO blocks in the long range, only to completely lose them as you move forward in time. They’ve been doing this for years now. BAMWX pointed it out yesterday 

[poolz1]

The AO and NAO have been negative for roughly the past two weeks and this was telegraphed nicely by the ens, imo.  Neither the GEFS nor the EPS are forecasting any HLB for the foreseeable future.  Not arguing the fact that phantom blocking periods have not been an issue the past year+ but I dont think we can say the same yet for this cold season. Head fakes and magnitude issues are always going to happen with LR ens guidance.

[snowman19]

@frd Looks like griteater came up with that composite you had posted about last month for October’s relationship with the wintertime North Pacific High (poleward or flat) and +QBO/basin-wide La Niña: 

 

[WinterWxLuvr]

8 hours ago, WinterWxLuvr said:

It seems that the gfs modeling is right in line with what can be typical in Ninas. At range models will send cold fronts south and east bringing shots of cold but as you get closer the cold tends to hit a wall and gets shunted east and has a difficult time making southeastward progress. The gfs had already done this once, this week, and not sure that it won’t next week too.

Here it is again. Check out the gfs forecast for next Thursday vs it’s forecast from two days ago.

This is a common occurrence in Nina years. Cold can, at times really struggle to progress southeastward.

[TSG]

10 hours ago, Eskimo Joe said:

 

[frd]

16 hours ago, snowman19 said:

@frd Looks like griteater came up with that composite you had posted about last month for October’s relationship with the wintertime North Pacific High (poleward or flat) and +QBO/basin-wide La Niña: 

 Never stated the winter will be awesome for cold and snow. I would at least entertain the possibility of a cold period in December,  as well as some significant temp swings. I still feel that things are progressing for some cold in November as well. The Ural High looks to set up and may stay for a while. Contrasting forecasts regarding the pv from the GEFS and the CFSv2 continue.  Also,  the Nina is changing up a bit in terms of expanse. That is important to consider for a seasonal forecast despite what the models are showing in the high latitudes with a lack of blocking presently.

 

.    

 

 

[CAPE]

31 minutes ago, frd said:

 Never stated the winter will be awesome for cold and snow. I would at least entertain the possibility of a cold period in December,  as well as some significant temp swings. I still feel that things are progressing for some cold in November as well. The Ural High looks to set up and may stay for a while. Contrasting forecasts regarding the pv from the GEFS and the CFSv2 continue.  Also,  the Nina is changing up a bit in terms of expanse. That is important to consider for a seasonal forecast despite what the models are showing in the high latitudes with a lack of blocking presently.

 

.  

It will be interesting to see if the extended range GEFS is on to something, as it continues to advertise increasing h5 heights up top, and especially in the NAO domain towards mid November. Also keeps the Pacific (puke) in check. Some of the MJO forecasts have the tropical convection moving into the W Pac, and then possibly into Phase 8, so that may give some credibility to the idea of increasing high latitude h5 heights down the line.

[snowman19]

1 hour ago, frd said:

 Never stated the winter will be awesome for cold and snow. I would at least entertain the possibility of a cold period in December,  as well as some significant temp swings. I still feel that things are progressing for some cold in November as well. The Ural High looks to set up and may stay for a while. Contrasting forecasts regarding the pv from the GEFS and the CFSv2 continue.  Also,  the Nina is changing up a bit in terms of expanse. That is important to consider for a seasonal forecast despite what the models are showing in the high latitudes with a lack of blocking presently.

 

.    

 

 

As far as the Niña, I mentioned over in the New England forum yesterday, it has become a basin-wide event. It is no longer “east-based”. All ENSO regions are solidly into a La Niña now. ENSO region 4 is the coldest it has been in over a decade....approaching -1.0C  

[ldub23]

This UKW is the strongest yet of this La Niña event due to a persistent series of strong EWBs that has been observed during the last few weeks. The arrival of this UKW to the east eq. Pacific should push ENSO solidly into moderate-strong La Niña territory through winter 2020-21. pic.twitter.com/1tVRZdk0DL
-- Tyler Stanfield (@TylerJStanfield) October 22, 2020

[40/70 Benchmark]

21 hours ago, snowman19 said:

@frd Looks like griteater came up with that composite you had posted about last month for October’s relationship with the wintertime North Pacific High (poleward or flat) and +QBO/basin-wide La Niña: 

 

I don't understand just selecting four random seasons because they are +NAO/AO. Not to say that this winter can not be very mild with a +NAO and flat Aleutian ridge, but that approach just seems lackadaisical and unscientific to me. Then again, I guess the analogs used to to derive the composite do not matter if accurately reflects how you feel the season will play out.

[snowman19]

4 hours ago, ldub23 said:

This UKW is the strongest yet of this La Niña event due to a persistent series of strong EWBs that has been observed during the last few weeks. The arrival of this UKW to the east eq. Pacific should push ENSO solidly into moderate-strong La Niña territory through winter 2020-21. pic.twitter.com/1tVRZdk0DL
-- Tyler Stanfield (@TylerJStanfield) October 22, 2020

This La Niña is not playing around. I think strong is definitely in the cards now....  

 

[psuhoffman]

2 hours ago, 40/70 Benchmark said:

I don't understand just selecting four random seasons because they are +NAO/AO. Not to say that this winter can not be very mild with a +NAO and flat Aleutian ridge, but that approach just seems lackadaisical and unscientific to me. Then again, I guess the analogs used to to derive the composite do not matter if accurately reflects how you feel the season will play out.

I don't agree with what he did.  But I also don't agree with some of the analog composites being throw around (not by you) that throw together an amalgemom of vastly different Nina years to give an average that is skewed or some mix of opposite patterns that is equally unlikely to occur.   With the exception of 1996 which was an extreme version of type 2, there seems to be 2 very clear distinct class of nina winters.  When there is a flat central pac ridge we tend to get much more southeast ridge and +AO and those years are some of the worst complete crap non winters in the mid atlantic.  In nina years with a more poleward pacific ridge we tend to get more trough into the northeast and also a higher probability of some blocking and while those years are still usually dry due to a suppressed STJ we at least get some respectable cold and closer to median snowfall.  The problem I see with some of the analog sets being throw around is they use way too many factors and end up taking a small sample that includes years from both of those types and the average H5 pattern you get is simply an average of two opposite patterns and unlikely to be what the dominant winter pattern really looks like.  One example I saw thrown around last month was particularly egregious.  They were filtering for QBO and ENSO matches.   Problem is there were only 3 years in that dataset.  And one was 1996.  1996 was skewing the mean of the whole set to look like a -NAO was likely when both of the other two years had a positive NAO.  That analog composite was completely misleading and useless.  I think some people that don't really understand how correlations and analogs work are starting to have too much fun with maps.  

[snowman19]

7 minutes ago, psuhoffman said:

I don't agree with what he did.  But I also don't agree with some of the analog composites being throw around (not by you) that throw together an amalgemom of vastly different Nina years to give an average that is skewed or some mix of opposite patterns that is equally unlikely to occur.   With the exception of 1996 which was an extreme version of type 2, there seems to be 2 very clear distinct class of nina winters.  When there is a flat central pac ridge we tend to get much more southeast ridge and +AO and those years are some of the worst complete crap non winters in the mid atlantic.  In nina years with a more poleward pacific ridge we tend to get more trough into the northeast and also a higher probability of some blocking and while those years are still usually dry due to a suppressed STJ we at least get some respectable cold and closer to median snowfall.  The problem I see with some of the analog sets being throw around is they use way too many factors and end up taking a small sample that includes years from both of those types and the average H5 pattern you get is simply an average of two opposite patterns and unlikely to be what the dominant winter pattern really looks like.  One example I saw thrown around last month was particularly egregious.  They were filtering for QBO and ENSO matches.   Problem is there were only 3 years in that dataset.  And one was 1996.  1996 was skewing the mean of the whole set to look like a -NAO was likely when both of the other two years had a positive NAO.  That analog composite was completely misleading and useless.  I think some people that don't really understand how correlations and analogs work are starting to have too much fun with maps.  

I don’t understand the people who are using complete ratters (i.e. 88-89, etc.) along with 95-96 and 10-11 as their analogs. It makes no sense

[psuhoffman]

31 minutes ago, snowman19 said:

I don’t understand the people who are using complete ratters (i.e. 88-89, etc.) along with 95-96 and 10-11 as their analogs. It makes no sense

It could make sense if there was a particular detail that caused the divergence in those years...and you know how and intend to tease out and compensate for that factor...to come up with a workable forecast.  But putting together a pattern composite of those years as a forecast would be bad forecasting imo.  

[snowman19]

33 minutes ago, psuhoffman said:

It could make sense if there was a particular detail that caused the divergence in those years...and you know how and intend to tease out and compensate for that factor...to come up with a workable forecast.  But putting together a pattern composite of those years as a forecast would be bad forecasting imo.  

Agreed. However, if it’s being done as a way to not be wrong no matter what happens this winter, that’s a pathetic way to ‘forecast’ IMO. If it’s a ratter, they can say “See, I used 88-89 as one of my analogs!” Or if it’s a very cold and snowy winter they can say “See, I used 95-96 and 10-11 as my analogs!”  Not saying that’s their intent, but if it is, that’s pretty sad

[raindancewx]

This is what I had for snow nationally in my winter forecast. I went with 2007 as the main analog - it's a pretty strong La Nina, has low solar, with slow cooling in Nino 4, following an El Nino, and it has low sea ice like this year. Atlantic looks fairly similar. The other years are in there to "fix it" in terms of some of the issues I had with it. 1995 is in there because it has similar tropical pacific SSTs so far to 2020, and it offsets the super low ACE of 2007 to a large degree, with its super high ACE. It also "fixes" some of the unusually low snow totals you'd have in the NE without it included. 2003/2012 have been decent non-ENSO match for a while (super weak monsoon, very hot in the West at times, landfalling NE tropical systems, etc), and last year is in there because it's recent and because I'm still somewhat skeptical on how cold Nino 4 will get. Generally, the areas expected to do well for snow (NW, Rockies, New England, Midwest) already are.

 

[PrinceFrederickWx]

15 hours ago, snowman19 said:

I don’t understand the people who are using complete ratters (i.e. 88-89, etc.) along with 95-96 and 10-11 as their analogs. It makes no sense

I understand it, as it's been the main problem I'm having when thinking about this winter. For the most part I'm leaning towards snowless, but there's some similarities I see with 95/96. My question with 95/96 though has always been if it was a 1/1,000 outlier event or something that's more likely given the right circumstances. It's never made sense to me for Mid-Atlantic snowfall vs. ENSO to have a bimodal distribution in weak La Nina and moderate El NIno- seems like one of those would flatten out over time as the sample grows larger (and the last 25 years seems to indicate the weak La Nina will be the one that flattens).

...anyway, for those that have messaged me, the snowfall contest will be posted early November. Should be interesting...

[RIC_WX]

8 minutes ago, PrinceFrederickWx said:

My question with 95/96 though has always been if it was a 1/1,000 outlier event or something that's more likely given the right circumstances.

The difference is entirely in the NAO.  In 95/96, we had alternating periods of cold/snow and otherwise mild and sunny wx, it was the winter that had something for everyone.  I believe it's remembered as cold because winter showed up early and hung around into April in addition to all the snow - but in the means it wasn't really cold, and plenty of very mild days.  The blocking slowed the overall pattern progression enough to extend the cold intrusions and pull the storm track into the ideal location for us repeatedly - shift the SE ridge 500 miles southeast and watch what happens.

Hard to argue we are seeing a much different setup this year, but absent any sustained blocking the cold is in and out in 36 hours, the SE ridge is back over our heads instead of off the coast, the storm track is west of the mountains.  Unless you can explain what will break the persistence of the past several winters to deliver more than transient periods of bootleg blocking we are getting a very similar pattern we would have seen in 95/96 if the block hadn't been present.

[CAPE]

14 minutes ago, RIC_WX said:

The difference is entirely in the NAO.  In 95/96, we had alternating periods of cold/snow and otherwise mild and sunny wx, it was the winter that had something for everyone.  I believe it's remembered as cold because winter showed up early and hung around into April in addition to all the snow - but in the means it wasn't really cold, and plenty of very mild days.  The blocking slowed the overall pattern progression enough to extend the cold intrusions and pull the storm track into the ideal location for us repeatedly - shift the SE ridge 500 miles southeast and watch what happens.

Hard to argue we are seeing a much different setup this year, but absent any sustained blocking the cold is in and out in 36 hours, the SE ridge is back over our heads instead of off the coast, the storm track is west of the mountains.  Unless you can explain what will break the persistence of the past several winters to deliver more than transient periods of bootleg blocking we are getting a very similar pattern we would have seen in 95/96 if the block hadn't been present.

There is a correlation between the AO phase and the ENSO state, and the AO is the index most highly correlated to snowfall for DC-BWI. The NAO is almost always in the negative phase when there is a negative AO, so the NAO is a factor. I don't have the numbers, but I would bet a predominately -AO for DJF is more likely during a Nino than a Nina. Not sure if there is a large enough sample size, but a Moderate, CP-based Nino might have the highest correlation to a -AO. @psuhoffman can probably give the numbers off the top of his head.

[snowman19]

30 minutes ago, RIC_WX said:

The difference is entirely in the NAO.  In 95/96, we had alternating periods of cold/snow and otherwise mild and sunny wx, it was the winter that had something for everyone.  I believe it's remembered as cold because winter showed up early and hung around into April in addition to all the snow - but in the means it wasn't really cold, and plenty of very mild days.  The blocking slowed the overall pattern progression enough to extend the cold intrusions and pull the storm track into the ideal location for us repeatedly - shift the SE ridge 500 miles southeast and watch what happens.

Hard to argue we are seeing a much different setup this year, but absent any sustained blocking the cold is in and out in 36 hours, the SE ridge is back over our heads instead of off the coast, the storm track is west of the mountains.  Unless you can explain what will break the persistence of the past several winters to deliver more than transient periods of bootleg blocking we are getting a very similar pattern we would have seen in 95/96 if the block hadn't been present.

IMO this winter is going to come down to the AO/NAO, if they are predominately positive this winter, everyone south of New England (really central New England) is screwed. Just -NAO without -AO isn’t going to help IMO, they both need to be predominantly negative together to get a good winter this time.  This La Niña is basin-wide and is very likely peaking strong, it’s also very strongly coupled. I don’t think we are going to see a good PAC side, it’s probably going to be garbage. Griteater pointed this out, but the pressure anomalies in the North Pacific, GOA and Alaska so far this month and projected going into November are now suggesting a flat Aleutian ridge as opposed to the +QBO Ninas that had poleward ones. If the AO/NAO cooperate, then it’s game on for cold and snow, see 73-74 super Niña, 10-11, 95-96 

 

[clskinsfan]

54 minutes ago, snowman19 said:

IMO this winter is going to come down to the AO/NAO,

 

This is basically every winter for us regardless of Enso state though. Our area simply does not do well without some blocking. 

[CAPE]

2 minutes ago, clskinsfan said:

This is basically every winter for us regardless of Enso state though. Our area simply does not do well without some blocking. 

We generally need some HL 'help' to get average or above average snow. In the absence of a -AO/NAO, a -EPO can get it done.

[clskinsfan]

3 minutes ago, CAPE said:

We generally need some HL 'help' to get average or above average snow. In the absence of a -AO/NAO, a -EPO can get it done.

Someone last year wrote a post (Maybe PSU but I cant remember) about the statistics and importance of a -NAO for our snows. The overwhelming majority of our snows seemed to occur with a -NAO even if it is only transient. 

[psuhoffman]

26 minutes ago, clskinsfan said:

Someone last year wrote a post (Maybe PSU but I cant remember) about the statistics and importance of a -NAO for our snows. The overwhelming majority of our snows seemed to occur with a -NAO even if it is only transient. 

I posted a chart of the state of the major indices before all of the Baltimore/Washington areas warning snowfalls going back to 1948. Its in the snow climo study thread. As @CAPE said we need some HL help. An east based poleward EPO ridge can compensate for a bad NAO. The AO is tricky because there is overlap from both the EPO and NAO so you get some cross contamination when you simply use that index. But taken holistically it’s evident we need some help up top somewhere.   We’re simply too far south if you get a contracted polar jet and zonal pattern. We need buckling somewhere. 
 

Specific to a La Niña the NAO becomes even more important. Baltimore rarely gets any snow of significance in a Nina without NAO help. I’ve not run the numbers but it’s likely similar for most of the region. It’s makes sense..in a Nina the pacific ridge is likely to be too far west to get er done all by itself. Add in a lack of STJ and fast NS jet and it makes sense that without the NAO buckling the flow in our favor we would be SOL. 

[clskinsfan]

3 minutes ago, psuhoffman said:

I posted a chart of the state of the major indices before all of the Baltimore/Washington areas warning snowfalls going back to 1948. Its in the snow climo study thread. As @CAPE said we need some HL help. An east based poleward EPO ridge can compensate for a bad NAO. The AO is tricky because there is overlap from both the EPO and NAO so you get some cross contamination when you simply use that index. But taken holistically it’s evident we need some help up top somewhere.   We’re simply too far south if you get a contracted polar jet and zonal pattern. We need buckling somewhere. 
 

Specific to a La Niña the NAO becomes even more important. Baltimore rarely gets any snow of significance in a Nina without NAO help. I’ve not run the numbers but it’s likely similar for most of the region. It’s makes sense..in a Nina the pacific ridge is likely to be too far west to get er done all by itself. Add in a lack of STJ and fast NS jet and it makes sense that without the NAO buckling the flow in our favor we would be SOL. 

I thought it was you. Thanks PSU. And I am glad to be back for the winter. Is Bob back yet?

[snowman19]

5 minutes ago, psuhoffman said:

I posted a chart of the state of the major indices before all of the Baltimore/Washington areas warning snowfalls going back to 1948. Its in the snow climo study thread. As @CAPE said we need some HL help. An east based poleward EPO ridge can compensate for a bad NAO. The AO is tricky because there is overlap from both the EPO and NAO so you get some cross contamination when you simply use that index. But taken holistically it’s evident we need some help up top somewhere.   We’re simply too far south if you get a contracted polar jet and zonal pattern. We need buckling somewhere. 
 

Specific to a La Niña the NAO becomes even more important. Baltimore rarely gets any snow of significance in a Nina without NAO help. I’ve not run the numbers but it’s likely similar for most of the region. It’s makes sense..in a Nina the pacific ridge is likely to be too far west to get er done all by itself. Add in a lack of STJ and fast NS jet and it makes sense that without the NAO buckling the flow in our favor we would be SOL. 

If you look back at La Nina winters in the NYC metro area, they are also screwed without AO/NAO assist. All the good La Niña winters were predominantly negative AO/NAO, the bad ones were predominantly positive AO/NAO

[Winter Wizard]

With regard to the AO and Nina winters, there have been 41 -AO winters since 1950-51. Among those 41, 11 occurred during La Nina years, 14 occurred during El Nino, and 16 in ENSO-neutral. By percentage, this comes out to, 52.3% of La Nina years (11/21), 53.8% of El Nino years (14/26), and 69.5% (16/23) ENSO-neutral winters. 

Among those 21 La Nina years at DCA, the average snowfall is 12.0", below the 1981-2010 average of 15.4". In +AO years, the average is just 8.8" with only one season (1971-1972) finishing above average (1999-2000 was exactly average). In -AO Nina years, the average is 14.4", however only 3 out of 11 featured above average snowfall. If you remove the massive outlier of 1995-96, the average falls to just 11.2".

For BWI, those numbers don't get much better. In fact, they arguably paint a bleaker picture. Among those 21 La Nina years at BWI, the average snowfall is 16.0", below the 1981-2010 average of 20.1". In +AO years, the average is just 12.1" with only one season (1999-2000) finishing above average. In -AO Nina years, the average is 18.9", however only 1 out of 11 featured above average snowfall and it's, you guessed it, 1995-96. If you remove the massive outlier of 1995-96, the average falls to just 14.5". 

If you break it down by month, in Nina winters at DCA, 4/21 Decembers have seen above average snowfall (3/4 in -AO years), 7/21 Januarys (4/7 in -AO years), and 4/21 Februarys (2/4 in -AO years). At BWI, it's 7/21 Decembers (5/7 in -AO years), 8/21 Januarys (5/8 in -AO years), and 5/21 Februarys (4/5 in -AO years).

Long story short, historically speaking, La Ninas are no bueno for the Mid-Atlantic, which should not come as a surprise. Clearly Atlantic blocking helps, as PSU outlined, but still an uphill battle from there. I'd be interested to see how the Pacific numbers shake up too - I imagine the +AO years had at least a semblance of +PNA/-EPO/-WPO. 

[psuhoffman]

@Winter Wizard

what sticks out for our area is when you look at each snowfall event you find that even the ones in a +AO/NAO Nina came during a period of blocking within the +AO. 2000 was a good example. All the snow came during a 10 day period with blocking. The rest of the winter was virtually snowless. 

[CAPE]

Capital Weather Gang article with a potpourri of winter outlooks. DT is thinking back-loaded and above normal snowfall for DC. 

https://www.washingtonpost.com/weather/2020/10/21/washington-dc-winter-forecasts-2020/?fbclid=IwAR0KbqBko3L1eyFPIZbWE5vr9l-LDSN2NUloXhSlZxFWXhsDGyG7lWz-j4w

[Scraff]

1 hour ago, CAPE said:

Capital Weather Gang article with a potpourri of winter outlooks. DT is thinking back-loaded and above normal snowfall for DC. 

https://www.washingtonpost.com/weather/2020/10/21/washington-dc-winter-forecasts-2020/?fbclid=IwAR0KbqBko3L1eyFPIZbWE5vr9l-LDSN2NUloXhSlZxFWXhsDGyG7lWz-j4w

In DT we trust.