12Z Guidance Discussion 12/23/2010

[Midlo Snow Maker]

Holy mama the GGEM is way south and terrible.

look out ket west

[PoconoD]

The GEM has the northern stream so fast that the Gulf Low might become its own storm!!!!

What could be the implications of that?

[rygar]

This is one of the most garbage threads I've read here in a while. I am amazed at how much stock people are putting both in the NAM past 48 hrs and the GFS also past 48 hours when both of these models up until now have been 1) extremely inconsistent with this storm and 2) always really sucky in general in the time range we are looking at.

What also makes this thread suck so much is that people are jumping ship before today's 12z Euro even came out, calling it a piece of s***, simply because the american models aren't looking good. Seriously? Let the damn Euro run first! It very well may be just as infallible as the other models with this storm, but no one has the right to say that yet!

In summation: this thread=

I don't think you have been paying attention.

[IsentropicLift]

Let's not forget their will still be another 10-12+ runs of the american models and 6-8 of the others. Just as people were saying a few days ago one solution is not likely to hold serve over a several day period the same could possibly be said now. I understand that the players on board are now being sampeled better and that the trends today have not been favorable but its still wayyyyy to early to be giving up on this...Especially if you live from NYC on northeastward.

[Snowmizer]

GGEM

Is that even looking at ALL like what we've been seeing the past few days?

WOW!

[LakeEffectKing]

Normally yes. This is a great point.

But the only thing that I can think of is the sw is being fully sampled for the 1st time in these 12z runs.

If euro still shows a hit with this sw sampling ingested, then this point is very valid.

I think it has been more of a northern stream issue (the handling of it, than the S. Stream.....

[Midlo Snow Maker]

WOW LOL

[KEITH L.I]

GGEM

Is that even looking at ALL like what we've been seeing the past few days?

WOW!

1015 low in the middle of the gulf?..lol..maybe it will become tropical..insane stuff

[PhineasC]

The GGEM looks terrible. No one even gets a whiff of a snowflake.

[SnowGoose69]

Wouldn't it be hilarious if the Gulf low tracks up the coast and buries us Miller A style?

I know it's not happening, but in this kind of pattern, I need a little chuckle.

12/30/2000 sort of occurred due to a failed system that should have happened on the 27th, the kicker of that storm eventually caused 12/30 itself...in this case though there is just not enough amplification for anything to come up the coast.

[WxClimate]

Two ideas:

- It doesn't matter who has a better model when the overall pattern is unpredictable. If the GFS and its ensembles turn out better than the Euro in this case, it's a bit extreme to call it the end of an era for the Euro. With the data assimilation process, it can take many cycles to flush out errors from a bad simulation - and this would be across all ensemble members.

- La Nina is "A" mode of variability, not "THE ONLY" mode of variability.

[stormtracker]

What a major debacle. GGEM gets no significant precip to anybody. I mean, I guess it's good that it's happening 3 days before and not 24 hours before, but wow.

[PhineasC]

What a major debacle. GGEM gets no significant precip to anybody. I mean, I guess it's good that it's happening 3 days before and not 24 hours before, but wow.

Clearly, some new information was ingested by all models because they have significantly changed the evolution of this storm in the past two cycles. These runs are getting uglier and uglier.

[donsutherland1]

The whole no KU in a strong Nina has pretty little merit, the sample size of strong Nina winter's isn't exactly relatively large...only dealing with about 7-9 years of data to work with.

To add clarity to the use of historical data:

1) 1871-2009 saw 25 cases with one or more winter months with an ENSO Region anomaly of -1 or lower.

2) Two KU-type storms (but no HECS as modeled on earlier runs of the ECMWF) -- December 1909 and January 2000 -- occurred during those 25 seasons.

3) Numerous near misses (inland or too far offshore) took place.

History + synoptic realities of moderate/strong La Niña events argued for caution. History suggested that something less than the HECS was far more likely than a HECS. Understanding that such Niñas often experience fast flow, absence of a strong subtropical stream, and must rely much more on phased solutions (inherent higher risk due to timing/location) and the recent synoptic maps also suggested that perhaps the "dream solution" might not be viable. If history is used properly, if can offer useful insight. It does not provide the entire answer and one must look beyond history to get a complete picture, especially as there is always the possibility of an exception. But this time around, the combination of history, general synoptic situation related to La Niña events, and complexity of the present synoptic situation all built a compelling case for caution even as the Euro was making the case for a HECS-type event.

[needbiggerboat]

fyi

[Organizing Low]

Clearly, some new information was ingested by all models because they have significantly changed the evolution of this storm in the past two cycles. These runs are getting uglier and uglier.

looks like a classic northern stream sampling issue

its textbook really.....

[Maximum lawman]

I think it has been more of a northern stream issue (the handling of it, than the S. Stream.....

I'm not quite ready to give up yet. The Euro is still a higher resolution model, so if there's only one outlier model, there's a legitmate reason why it might be an outlier. The northern stream energy is pretty subtle.

[Quasievil]

Clearly, some new information was ingested by all models because they have significantly changed the evolution of this storm in the past two cycles. These runs are getting uglier and uglier.

The wheels came off the cart when the s/w came on shore in CA. Overall we've got ourselves a real train wreck here.

[dtk]

Two ideas:

- It doesn't matter who has a better model when the overall pattern is unpredictable. If the GFS and its ensembles turn out better than the Euro in this case, it's a bit extreme to call it the end of an era for the Euro. With the data assimilation process, it can take many cycles to flush out errors from a bad simulation - and this would be across all ensemble members.

- La Nina is "A" mode of variability, not "THE ONLY" mode of variability.

Huh? One of the main points of having an ensemble forecast is to SAMPLE/MODEL/SIMULATE the uncertainly in the initial conditions (and secondarily in the model itself).

I don't think anyone is calling this an "end of an era for the Euro"...and if they are, they should never post again. One (potentially) poor/overdone set of forecasts does not mean it's going to suck in the future. Did not the EC model nail the recent midwest storm (at least relative to the other models)? We need to learn from this SPECIFIC type of situation, instead of making grandiose statements about overall usefulness of certain models.

I also need to say that I appreciate that most of the mets were fairly conservative throughout. The op EC was the extreme outlier (even relative to its own ensemble)....there was always a red flag there.....

[weatherwiz]

To add clarity to the use of historical data:

1) 1871-2009 saw 25 cases with one or more winter months with an ENSO Region anomaly of -1 or lower.

2) Two KU-type storms (but no HECS as modeled on earlier runs of the ECMWF) -- December 1909 and January 2000 -- occurred during those 25 seasons.

3) Numerous near misses (inland or too far offshore) took place.

History + synoptic realities of moderate/strong La Niña events argued for caution. History suggested that something less than the HECS was far more likely than a HECS. Understanding that such Niñas often experience fast flow, absence of a strong subtropical stream, and must rely much more on phased solutions (inherent higher risk due to timing/location) and the recent synoptic maps also suggested that perhaps the "dream solution" might not be viable. If history is used properly, if can offer useful insight. It does not provide the entire answer and one must look beyond history to get a complete picture, especially as there is always the possibility of an exception. But this time around, the combination of history, general synoptic situation related to La Niña events, and complexity of the present synoptic situation all built a compelling case for caution even as the Euro was making the case for a HECS-type event.

Great post, Don and appreciate the feedback. Just out of curiosity do you happen to have any links of ENSO events pre-1950? The data I was using was from post 1950.

Despite the fact that the Euro had been signaling an HECS-type of event I know I have been taking this with caution given the timerange we were in, while there were some signals that were suggesting the potential there were/are certainly some signals that shy away from this. Unless there is absolute 100% agreement and the number of issues that would prevent such an extreme solution from occurring are relatively minor than I would definitely always take the more conservative approach to things.

[SnowGoose69]

Huh? One of the main points of having an ensemble forecast is to SAMPLE/MODEL/SIMULATE the uncertainly in the initial conditions (and secondarily in the model itself).

I don't think anyone is calling this an "end of an era for the Euro"...and if they are, they should never post again. One (potentially) poor/overdone set of forecasts does not mean it's going to suck in the future. Did not the EC model nail the recent midwest storm (at least relative to the other models)? We need to learn from this SPECIFIC type of situation, instead of making grandiose statements about overall usefulness of certain models.

I also need to say that I appreciate that most of the mets were fairly conservative throughout. The op EC was the extreme outlier (even relative to its own ensemble)....there was always a red flag there.....

Obviously we don't know yet although it seems inevitable...I'd be wary of trusting the Euro during La Ninas in any instance where its markedly slower than the other models and that slow relative speed is a major factor in its end result as it was here.

[cp63]

I think the Euro's bias in the SW bit it big time because of this raging La Nina we're in...I mentioned how it can usually still come out to the right end solution but given the flow to this pattern maybe it was a bit too much for it to be able to do so.

Good point,and I would imagine,something that may hamper it all season...........

[bmorelights]

Look everyone, as a certified internet snowologist I can tell you with certainty that the Euro was right a day ago. It's all going to happen just like that. So get out your snow blowers, portable weenie-roasters, and sixty-inch snow tires, and batten down the hatches!

Really though, I'm not any sort of meteorologist, and I can tell you that an outlier is rarely right. In a data set where one point is way off in left field, you should just quietly scratch out the left fielder and publish with the coherent set. That's what the TV mets did, and it's looking like they are right on. If the Euro is not yelping back towards the GFS with its tail between its legs on the 12z run I'll eat my hat.

Weenie storm cancel...

EDIT: Also, keep in mind that the euro didn't start going sour until the shortwave was firmly over land. I think that was the main problem with the euro's solution... Not correctly handling that shortwave. No model can make really accurate predictions without enough data input, no matter how high the resolution.

[donsutherland1]

Great post, Don and appreciate the feedback. Just out of curiosity do you happen to have any links of ENSO events pre-1950? The data I was using was from post 1950.

Despite the fact that the Euro had been signaling an HECS-type of event I know I have been taking this with caution given the timerange we were in, while there were some signals that were suggesting the potential there were/are certainly some signals that shy away from this. Unless there is absolute 100% agreement and the number of issues that would prevent such an extreme solution from occurring are relatively minor than I would definitely always take the more conservative approach to things.

Thanks Weatherwiz.

You can find the earlier SSTs here: http://www.esrl.noaa.gov/psd/gcos_wgsp/Timeseries/Data/nino34.long.data

[eduggs]

The 12z GFS ensemble mean isn't dreadful, especially for EMa. The mean low is progged close enough, and the time period far enough, to still keep an eye on it. Unfortunately we are converging on a solution just as the timeframe for big model changes comes to an end.

[SnowGoose69]

Good point,and I would imagine,something that may hamper it all season...........

I'm beginning to think its the same reason the GEM has been so good and the NOGAPS very very very progressive for the most part....La Nina plays against the GEM's bias and right into the bias of the Euro and NOGAPS...hence one would believe the GEM would be poor during Ninos and the NOGAPS good during Ninos.

[Ian]

Clearly, some new information was ingested by all models because they have significantly changed the evolution of this storm in the past two cycles. These runs are getting uglier and uglier.

as soon as the southern shortwave was onshore it all fell apart

[weatherwiz]

Thanks Weatherwiz.

You can find the earlier SSTs here: http://www.esrl.noaa...ino34.long.data

Beautiful! Thanks

[dtk]

The 12z GFS ensemble mean isn't dreadful, especially for EMa. The mean low is progged close enough, and the time period far enough, to still keep an eye on it. Unfortunately we are converging on a solution just as the timeframe for big model changes comes to an end.

And with the ensemble mean surface track being seemingly west of the op....still evidence that there is plenty of uncertainty this close to the event! (I didn't mean to imply in my previous post anyone knows for sure what is going to happen.....)

[LakeEffectKing]

I'm not quite ready to give up yet. The Euro is still a higher resolution model, so if there's only one outlier model, there's a legitmate reason why it might be an outlier. The northern stream energy is pretty subtle.

Not so subtle....look at how the Euro has been handling the energy north of N. Dakota/MN...it dives it SSSE' ward....other models have handled it in such a way that it actually buckles the trough over the northern plains....enough so that it broadens the base of the trough, and "deflects" the s/w energy to the north of the southern stream energy. There is essentially a plateau of h500 heights (ie little slope) with pockets of energy "hanging" out, until the plateau gets tipped southward (heights start to fall). The timing and "subtle" motions that ensue, are extremely sensitive and I'd assume, inherently hard to model.

Take a couple of marbles and roll them down a ramp.....you probably can predict farily closely the paths. Now take a level table with a couple marbles on it, and slowly start to tip it....the smallest of "features" in the table will drive the initial motions, until the gravitational force becomes dominant...but those tiny features set the initial course, which means a very small pertabation in the marbles' starting locations could mean large outcomes of their paths, given the SAME tipping of the table.