Thanksgiving Eve SECS/MECS Discussion II

[gkrangers]

Have at it.

[Sickman]

Let's keep this one neat and tidy.

 

Storm Mode is probably forthcoming.

[Isotherm]

If some of this afternoon's model trends are to believed, one could be led to the conclusion that the city / coast are heading for an unprecedented event, one which hasn't been seen since 1950. When I say unprecedented, I mean the following: snowfall accumulation of 1.5" or greater at Central Park.

 

Upon examination of previous events since 1950, we see that 100% of cases in which 1.5"+ totals occurred, coincided with a significant negative dip in either the AO or NAO modality. Those periods are as follows:

 

November of 1952, produced 1.7" for the month at NYC. The AO monthly value was around -1.8, with a mid-month dip to -3.0 standard deviation.

 

November of 1953, 2.2" at NYC. The AO daily reached a minimum of -1.2 around November 21st.

 

November of 1967, 3.2" at NYC. The AO daily reached a minimum of -2.1 on November 6th, and a significant snowfall ensued around that date (+/- a couple days).

 

November of 1978, 2.2" at NYC. The AO daily reached a minimum of -2.1 on November 26th, and the significant snowfall occurred around then.

 

November of 1989, 4.7" at NYC. The AO daily reached a minimum of -2.2 on November 22nd, and a significant snow ensued.

 

November of 1995, 2.9" at NYC. The AO daily reached a minimum of -1.2 on November 27th, and the snowstorm followed a day later.

 

November of 2012, 4.7" at NYC. The AO daily dipped to -2.5 in the beginning of November, the NAO to -1.8, and a snowstorm followed a few days later.

 

 

Current and forecast: AO and NAO values are slightly positive (AO near +2 SD right now), and should both be around +0.9 SD for the onset of the progged event.

 

 

No two patterns are alike -- this is a fact. However, it is also a fact that history is important in determining the future. The preset patterns for 1.5" + snowfalls at NYC are such that a negative AO dip of significance has always occurred around / just prior to the snowfall event. That common denominator is not present for the progged pattern.

 

If NYC, LI and coastal NJ end up with 1.5"+ or 3-6" out of this pattern, then congrats to all of us -- we have just made history folks (at least since 1950). But it's impressive how overwhelming the historical evidence is given the progged results on some model data. At this juncture, given the above and other factors regarding the pattern that I have mentioned, I would be very, very reluctant to forecast more than an inch or so for LI, coastal NJ, and NYC.

 

We'll see where the overnight guidance and tomorrow takes us.

[IsentropicLift]

Good agreement on the 18z GEFS mean regarding track with the OP. The mean is actually a hair dryer than the OP.

[SnoSki14]

Given Isotherm's above post it just proves that we still have a lot to learn and plenty more research will have to be made if we do in fact make history. However we always have to consider the sample size and sometimes there just isn't enough data out there and enough precedence to draw a conclusion. As rare as this may be going off a small sample, in reality it may not be as rare as we think. There are also other factors to think about beyond the AO. We saw last winter how seemingly unfavorable indices do not necessarily generate the historical outcome. 

[danstorm]

If some of this afternoon's model trends are to believed, one could be led to the conclusion that the city / coast are heading for an unprecedented event, one which hasn't been seen since 1950. When I say unprecedented, I mean the following: snowfall accumulation of 1.5" or greater at Central Park.

 

Upon examination of previous events since 1950, we see that 100% of cases in which 1.5"+ totals occurred, coincided with a significant negative dip in either the AO or NAO modality. Those periods are as follows:

 

November of 1952, produced 1.7" for the month at NYC. The AO monthly value was around -1.8, with a mid-month dip to -3.0 standard deviation.

 

November of 1953, 2.2" at NYC. The AO daily reached a minimum of -1.2 around November 21st.

 

November of 1967, 3.2" at NYC. The AO daily reached a minimum of -2.1 on November 6th, and a significant snowfall ensued around that date (+/- a couple days).

 

November of 1978, 2.2" at NYC. The AO daily reached a minimum of -2.1 on November 26th, and the significant snowfall occurred around then.

 

November of 1989, 4.7" at NYC. The AO daily reached a minimum of -2.2 on November 22nd, and a significant snow ensued.

 

November of 1995, 2.9" at NYC. The AO daily reached a minimum of -1.2 on November 27th, and the snowstorm followed a day later.

 

November of 2012, 4.7" at NYC. The AO daily dipped to -2.5 in the beginning of November, the NAO to -1.8, and a snowstorm followed a few days later.

 

 

Current and forecast: AO and NAO values are slightly positive (AO near +2 SD right now), and should both be around +0.9 SD for the onset of the progged event.

 

 

No two patterns are alike -- this is a fact. However, it is also a fact that history is important in determining the future. The preset patterns for 1.5" + snowfalls at NYC are such that a negative AO dip of significance has always occurred around / just prior to the snowfall event. That common denominator is not present for the progged pattern.

 

If NYC, LI and coastal NJ end up with 1.5"+ or 3-6" out of this pattern, then congrats to all of us -- we have just made history folks (at least since 1950). But it's impressive how overwhelming the historical evidence is given the progged results on some model data. At this juncture, given the above and other factors regarding the pattern that I have mentioned, I would be very, very reluctant to forecast more than an inch or so for LI, coastal NJ, and NYC.

 

We'll see where the overnight guidance and tomorrow takes us.

 

Great write-up and research.

 

But, absolutely no disrespect intended, I am not sure why some cling to indices and their historical implications as much as they do.  The small sample size of the historical record makes definitive - and statistically significant - statements practically worthless.  

 

E.g., snow in itself is rare in November.  1"+ happens in far fewer years than not in NYC.  To drill down further against that already reduced sample using indices is pretty bad test/control methodology.  If we had a historical record of about 1000 years, maybe, but the data is pretty thin.

 

And this is not to say that I think it will snow a lot in NYC... I'm not entirely convinced, though I would put the odds of 2"+ at >50% for KNYC ... but I would never base a forecast on what history has to say about the interaction between an index and sensible weather.

[SnowGoose69]

I think the 1938 storm had a +NAO and AO

[wthrmn654]

http://www.weather.gov/okx/winter?selected=Suffolk,%20NY#table

Amazing little thing Upton has posted.....go to Suffolk....plum island ny will likely see more then Islip...being further north, with amaximum possibly of 6. If chose this is all experimental but interesting not the less!

[MJO812]

It's good that the models are converging on the track. Now we just have to figure out the intensity.

[Heisy]

Question for mets;

 

How accurate have you found the snow probabilities on the SREFs to be at this range or even closer? 

[Isotherm]

Great write-up and research.

 

But, absolutely no disrespect intended, I am not sure why some cling to indices and their historical implications as much as they do.  The small sample size of the historical record makes definitive - and statistically significant - statements practically worthless.  

 

E.g., snow in itself is rare in November.  1"+ happens in far fewer years than not in NYC.  To drill down further against that already reduced sample using indices is pretty bad test/control methodology.  If we had a historical record of about 1000 years, maybe, but the data is pretty thin.

 

And this is not to say that I think it will snow a lot in NYC... I'm not entirely convinced, though I would put the odds of 2"+ at >50% for KNYC ... but I would never base a forecast on what history has to say about the interaction between an index and sensible weather.

 

 

Thanks. I agree that the sample size is already rather small given the rarity of November. History alone shouldn't be the sole basis of a forecast, but it can lay the foundation. The progged downstream north atlantic pattern is also far from ideal for a coastal significant snowfall. We would like to see some semblance of higher heights near Greenland or at least a 50/50 type blocking feature in SE Canada to hold the cold source in place for the I-95 corridor. This seems like a very thread the needle type set-up to me. The Western ridge axis is decaying and a bit west of where we'd like, which also argues for more resistance near the east coast. So there are other factors in addition to historical precedent that I think support than notion of little accumulation for coastal areas. Inland looks like a virtual lock for a paste bomb.

[TeleConnectSnow]

QUESTION:

Why is it that the models have a dry NW side of this system? Wouldn't the intensity throwback moisture/with the WARM SST off the coast help the LOW deepen more rapidly?

[gkrangers]

Thanks. I agree that the sample size is already rather small given the rarity of November. History alone shouldn't be the sole basis of a forecast, but it can lay the foundation. The progged downstream north atlantic pattern is also far from idea for a coastal significant snowfall. We would like to see some semblance of higher heights near Greenland or at least a 50/50 type blocking feature in SE Canada to hold the cold source in place for the I-95 corridor. This seems like a very thread the needle type set-up to me. The Western ridge axis is decaying and a bit west of where we'd like, which also argues for more resistance near the east coast. So there are other factors in addition to historical precedent that I think support than notion of little accumulation for coastal areas. Inland looks like a virtual lock for a paste bomb.

MoCo will jackpot at some point this winter. Hardly ever fails.

[jm1220]

Thanks. I agree that the sample size is already rather small given the rarity of November. History alone shouldn't be the sole basis of a forecast, but it can lay the foundation. The progged downstream north atlantic pattern is also far from idea for a coastal significant snowfall. We would like to see some semblance of higher heights near Greenland or at least a 50/50 type blocking feature in SE Canada to hold the cold source in place for the I-95 corridor. This seems like a very thread the needle type set-up to me. The Western ridge axis is decaying and a bit west of where we'd like, which also argues for more resistance near the east coast. So there are other factors in addition to historical precedent that I think support than notion of little accumulation for coastal areas. Inland looks like a virtual lock for a paste bomb.

I agree with what you're saying but I think it's a tougher call for places like here. If the precip comes in like a wall and it's 37-38 or so on Weds morning with a north wind, it should cool down here quickly and it becomes 33 degree heavy snow. If the precip is spotty and temps warm up to the low 40s before the precip starts, we likely waste a lot as rain or nonaccumulating snow. We've had surprises here before especially with the Oct 2011 and Nov 2012 events which were supposed to be rain or slop at the shore. Of course this also means that we need the 850mb/925mb lows southeast of here or it's rain anyway. I will say that if it snows a lot around here it could cause tree damage since some of our trees still have leaves. 

 

The safe bet here is maybe a couple of inches of slop and a solid chunk of the precip gone to rain, but I can see how it turns into more if things work in our favor. And of course where I am can pretty much have its own climate, which is what happened on Feb 13 last winter.

 

If I was making a call now I'd say 8-12" for Orange, Putnam, W Passaic and down into Morris, Sussex, Warren Counties-basically anywhere west of I-287. 4-8" for the closer western and northern suburbs, west of I-95. 2-4" for most around the city and northern Long Island and down through Monmouth County away from the shore, and 1-3" for my area, coastal Monmouth and east through Long Island. 

[TeleConnectSnow]

The only thing concerning is all the models really have been coming in drier and drier each run. Even the Euro.

[IsentropicLift]

The only thing concerning is all the models really have been coming in drier and drier each run. Even the Euro.

Sounds like you want to join Tony

[snywx]

The only thing concerning is all the models really have been coming in drier and drier each run. Even the Euro.

 

The Euro, Uk, RGEM, NAM are all coming in wetter. The GFS is the only model that came in drier. 

 

Can you please stop posting and just observe

[SnowGoose69]

The only thing concerning is all the models really have been coming in drier and drier each run. Even the Euro.

That's very common with fast moving systems in the absence of blocking, models start cutting QPF back inside 48 hours. DT frequently mentions this

[TeleConnectSnow]

Sounds like you want to join Tony

It's part of the discussion, I've seen several members talking about this subject on the various sub forums. It's an honest question as to why they are trending drier.

[TeleConnectSnow]

That's very common with fast moving systems in the absence of blocking, models start cutting QPF back inside 48 hours. DT frequently mentions this

Now this is a mature comment to my question instead of just hating. Thank you VERY MUCH SnowGoose for that answer. Explains it for me.

[TeleConnectSnow]

The Euro, Uk, RGEM, NAM are all coming in wetter. The GFS is the only model that came in drier. 

 

Can you please stop posting and just observe

Also, The EURO came in drier by some from 00z-12z? I did observe that, just looked at both runs on WxBell, and the QPF decreased.

[swamplover56]

The models are really not coming in drier it's just model noise at this point the differences in qpf between 1.1 to 1.3 is really nothing just variance

[Rjay]

If I was making a call now I'd say 8-12" for Orange, Putnam, W Passaic and down into Morris, Sussex, Warren Counties-basically anywhere west of I-287. 4-8" for the closer western and northern suburbs, west of I-95. 2-4" for most around the city and northern Long Island and down through Monmouth County away from the shore, and 1-3" for my area, coastal Monmouth and east through Long Island.

100% agree with this. I actually just put out a prediction on a another board that's quite similar...almost exact.

[Isotherm]

I agree with what you're saying but I think it's a tougher call for places like here. If the precip comes in like a wall and it's 37-38 or so on Weds morning with a north wind, it should cool down here quickly and it becomes 33 degree heavy snow. If the precip is spotty and temps warm up to the low 40s before the precip starts, we likely waste a lot as rain or nonaccumulating snow. We've had surprises here before especially with the Oct 2011 and Nov 2012 events which were supposed to be rain or slop at the shore. Of course this also means that we need the 850mb/925mb lows southeast of here or it's rain anyway. I will say that if it snows a lot around here it could cause tree damage since some of our trees still have leaves. 

 

The safe bet here is maybe a couple of inches of slop and a solid chunk of the precip gone to rain, but I can see how it turns into more if things work in our favor. And of course where I am can pretty much have its own climate, which is what happened on Feb 13 last winter.

 

If I was making a call now I'd say 8-12" for Orange, Putnam, W Passaic and down into Morris, Sussex, Warren Counties-basically anywhere west of I-287. 4-8" for the closer western and northern suburbs, west of I-95. 2-4" for most around the city and northern Long Island and down through Monmouth County away from the shore, and 1-3" for my area, coastal Monmouth and east through Long Island. 

 

 

 

Yeah, stranger things have happened. Should be interesting to see how this transpires. Regarding October 2011, I would also argue that was a more favorable pattern than the present, w/ an apparent 50/50 vortex and east based Scand. -NAO block. Nov 2012 had strong high latitude blocking as noted before.

 

 

I would probably go < 1.5" for NYC/LI/coastal NJ and 3-6"+ NW of I-95, 4-8"+ far NW NJ/HV right now.

 

[LnbWx]

http://www.weather.gov/okx/winter?selected=Suffolk,%20NY#table

Amazing little thing Upton has posted.....go to Suffolk....plum island ny will likely see more then Islip...being further north, with amaximum possibly of 6. If chose this is all experimental but interesting not the less!

 

Like this experimental page. Lot of information, concise and easy to read. Any questions on precipitation totals on the backyards can be redirected to this page.

[RU848789]

As often occurs in these situations, collaboration between the NYC and Philly NWS offices could be better. No way that Staten Island, to my east and Union County, to my north, can have 2-4" and 4-6", respectively, while adjacent northern Middlesex County has 1-2" for most and 2-4" for the far NW part of the county.

It's also rare that Trenton and Philly do better than Edison, when it's not a storm that's suppressed to our south, i.e., when the total precip is about the same, implying that there will be more rain in Edison.

http://www.erh.noaa.gov/okx/stormtotalprecip.php

http://www.erh.noaa.gov/phi/StormTotalSnow/stormtotal.php

[uncle W]

If some of this afternoon's model trends are to believed, one could be led to the conclusion that the city / coast are heading for an unprecedented event, one which hasn't been seen since 1950. When I say unprecedented, I mean the following: snowfall accumulation of 1.5" or greater at Central Park.

 

Upon examination of previous events since 1950, we see that 100% of cases in which 1.5"+ totals occurred, coincided with a significant negative dip in either the AO or NAO modality. Those periods are as follows:

 

November of 1952, produced 1.7" for the month at NYC. The AO monthly value was around -1.8, with a mid-month dip to -3.0 standard deviation.

 

November of 1953, 2.2" at NYC. The AO daily reached a minimum of -1.2 around November 21st.

 

November of 1967, 3.2" at NYC. The AO daily reached a minimum of -2.1 on November 6th, and a significant snowfall ensued around that date (+/- a couple days).

 

November of 1978, 2.2" at NYC. The AO daily reached a minimum of -2.1 on November 26th, and the significant snowfall occurred around then.

 

November of 1989, 4.7" at NYC. The AO daily reached a minimum of -2.2 on November 22nd, and a significant snow ensued.

 

November of 1995, 2.9" at NYC. The AO daily reached a minimum of -1.2 on November 27th, and the snowstorm followed a day later.

 

November of 2012, 4.7" at NYC. The AO daily dipped to -2.5 in the beginning of November, the NAO to -1.8, and a snowstorm followed a few days later.

 

 

Current and forecast: AO and NAO values are slightly positive (AO near +2 SD right now), and should both be around +0.9 SD for the onset of the progged event.

 

 

No two patterns are alike -- this is a fact. However, it is also a fact that history is important in determining the future. The preset patterns for 1.5" + snowfalls at NYC are such that a negative AO dip of significance has always occurred around / just prior to the snowfall event. That common denominator is not present for the progged pattern.

 

If NYC, LI and coastal NJ end up with 1.5"+ or 3-6" out of this pattern, then congrats to all of us -- we have just made history folks (at least since 1950). But it's impressive how overwhelming the historical evidence is given the progged results on some model data. At this juncture, given the above and other factors regarding the pattern that I have mentioned, I would be very, very reluctant to forecast more than an inch or so for LI, coastal NJ, and NYC.

 

We'll see where the overnight guidance and tomorrow takes us.

the AO was -2.5 about a week ago and sometimes storms come around the ao minimum...it could be another year where the ao isn't the driver...temperatures are starting out pretty high for the beginning of the storm...Nov. 1955 is the first snowfall I remember a little bit about...what I remember about it was it was a wet snow...0.38" of water equivalent and 1" of snow...temperatures were in the 30's...it could be a quicker, drier, colder event...if so 1-4" in the city can still happen....Time will tell...

[Stormlover74]

As often occurs in these situations, collaboration between the NYC and Philly NWS offices could be better. No way that Staten Island, to my east and Union County, to my north, can have 2-4" and 4-6", respectively, while adjacent northern Middlesex County has 1-2" for most and 2-4" for the far NW part of the county.

It's also rare that Trenton and Philly do better than Edison, when it's not a storm that's suppressed to our south, i.e., when the total precip is about the same, implying that there will be more rain in Edison.

http://www.erh.noaa.gov/okx/stormtotalprecip.php

http://www.erh.noaa.gov/phi/StormTotalSnow/stormtotal.php

Good point..the cutoff usually isn't that severe in these situations. Chances are the city and western/sw burbs will all be within a couple inches of each other with areas S&E getting much less and well N&W getting much more

[LongBeachSurfFreak]

Rates rates rates. 35 DBZ and it doesn't matter if it's right to the beach on Long Island.

I could see somewhere on the island ending with a larger total then somewhere NW of 287. Based off where banding sets up.

We actually saw a preview of this during the last event with areas in south central jersey that saw 35 plus get a nice coating while areas 75 miles further north in say CT saw zero due to lesser rates.

There was a quick hitter here in 07 where south of the southern state has 3" of paste stuck to everything while north of there were rates were much less it never accumulated. (I forget the exact dates)

I agree with a general 1-3" right at the coast but there could easily be 6" areas in the best banding and someone who is gonna have to cry they can't accumulate (I hope that's not me)

[jm1220]

Rates rates rates. 35 DBZ and it doesn't matter if it's right to the beach on Long Island.

I could see somewhere on the island ending with a larger total then somewhere NW of 287. Based off where banding sets up.

We actually saw a preview of this during the last event with areas in south central jersey that saw 35 plus get a nice coating while areas 75 miles further north in say CT saw zero due to lesser rates.

There was a quick hitter here in 07 where south of the southern state has 3" of paste stuck to everything while north of there were rates were much less it never accumulated. (I forget the exact dates)

I agree with a general 1-3" right at the coast but there could easily be 6" areas in the best banding and someone who is gonna have to cry they can't accumulate (I hope that's not me)

I think the earlier in the day it starts the better too. If it's already coming down by dawn temps won't have a chance to rise and if precip is heavy enough it should be snow. Keeping the wind direction northerly from 850mb down is another factor.