The GFS

[isnice]

The 00z model runs seem to have taken a major step towards downplaying the clipper system. The Euro still has the feature and creates an upper-level low that spins over the Canadian border, but the effects are much less exascterbated. Could the GFS actually have been right all along? It's not the first time this year the GFS saw something before all the other models. Maybe this winter will be a GFS victory over the Euro and CMC because it's an American model and is more experienced with La Nina conditions over North America? I'd really like a met to chime in if possible. Thanks.

[Ellinwood]

Let the event happen first, then do a verification on which model was right... no one can know for sure which one was right until it actually happens.

Also, nit-picking certain systems is not a solid basis for saying that one model is better than the other. The majority of the time, the ECM (ensemble) still reigns supreme.

[Ender]

The 00z model runs seem to have taken a major step towards downplaying the clipper system. The Euro still has the feature and creates an upper-level low that spins over the Canadian border, but the effects are much less exascterbated. Could the GFS actually have been right all along? It's not the first time this year the GFS saw something before all the other models. Maybe this winter will be a GFS victory over the Euro and CMC because it's an American model and is more experienced with La Nina conditions over North America? I'd really like a met to chime in if possible. Thanks.

This won't be the first, or second, or third, year where someone look's back at the EC vs. GFS forecasts for a system and says, "Look at that, the EC got it wrong again while the GFS was trying to convince us it had the right idea and low and behold..."

For whatever reason the EC seems to have a "weakness" for over developing systems from the SE to MA coasts from between forecast hour 132 and 180. I seems that it always corrects for this once we're inside 132hr or 120hr at the latest. It did nearly the same thing last December while the GFS insisted not to play along. Then when the two super storms formed it was the GFS that initially latched onto the idea of those storms and the EC that took a little while longer to come around.

[DaculaWeather]

The NWS guys here in Atlanta are leaning toward the GFS:

.LONG TERM /THURSDAY NIGHT THROUGH MONDAY/...SIGNIFICANT DIFFERENCES BETWEEN THE 00Z RUN OF THE GFS AND THE 12ZRUN OF THE ECMWF. WE HAVE BEEN LEANING WITH THE GFS AND WILLCONTINUE TO DO SO. THE H5 TROUGH WITH THE CURRENT SYSTEM WILL MOVEOFF THE ATLANTIC WITH A RIDGE BUILDING OVER THE MIDWEST. THIS WILLPUT THE CWA IN A DRY NORTHWEST FLOW THAT WILL CONTINUE THROUGH THEWEEKEND. THIS WILL ALSO KEEP TEMPS AT OR JUST A FEW DEGREES BELOWNORMAL. A SHORT WAVE MOVING THROUGH NORTHWEST CANADA WILL DIVEINTO THE NORTHERN PLAINS AND JOIN WITH AN UPPER LOW OVERNEWFOUNDLAND CANADA AND THEN DEEPEN OVER THE EASTERN U.S. THISWILL BRING ANOTHER COLD FRONT INTO THE CWA MONDAY NIGHT ANDTUESDAY. ATMOSPHERE WILL REMAIN VERY DRY AND AT THIS TIME THE

PROBABILITIES ARE TOO LOW TO MENTION PRECIP AT THIS TIME.

[jdrenken]

The NWS guys here in Atlanta are leaning toward the GFS:

Interesting especially with how the HPC crew disco goes...

PRELIMINARY EXTENDED FORECAST DISCUSSION

NWS HYDROMETEOROLOGICAL PREDICTION CENTER CAMP SPRINGS MD

339 AM EST TUE NOV 30 2010

VALID 12Z SAT DEC 04 2010 - 12Z TUE DEC 07 2010

THE GLOBAL DETERMINISTIC AND ENSEMBLE GUIDANCE IS IN GOOD

AGREEMENT THAT A DEEP CYCLONE WILL BE SLOWLY PROGRESSING THROUGH

THE NORTHEAST PACIFIC...AND THAT A NEGATIVELY TILTED TROUGH WITH

AN EMBEDDED DEEP CYCLONE WILL LURK NEAR NEW ENGLAND AND ATLANTIC

CANADA THROUGH THE PERIOD /THE 00Z CANADIAN HAS ISSUES WITH THIS

LATTER SYSTEM DUE TO ITS UNIQUE SOLUTION IN CENTRAL CANADA EARLY

TO MID NEXT WEEK/.

THERE ARE PAIR OF FORECAST ISSUES EARLY THIS MORNING. THE FIRST

ISSUE REVOLVES AROUND A PACIFIC SHORTWAVE MOVING OVER THE TOP OF

THE BROAD/INCREASINGLY STRONG RIDGE CENTERED ACROSS THE SOUTHERN

STATES WHICH MOVES NEAR THE US/CANADIAN BORDER BEFORE DIVING

EAST-SOUTHEAST INTO THE MID-ATLANTIC STATES. THE 00Z CANADIAN/00Z

ECMWF MOVE THE SYSTEM THROUGH THE NORTH AND EAST ABOUT A DAY

QUICKER THAN THE 00Z GFS...AS THE GFS USES ADDITIONAL NORTHERN

STREAM ENERGY MOVING THROUGH WESTERN CANADA TO SLOW ITS PACE AS IT

MOVES THROUGH AN OTHERWISE QUICK QUASI-ZONAL/SUPERGEOSTROPHIC

FLOW. ALSO...BY THE TIME IS MOVES NEAR THE MID-ATLANTIC

STATES...THE GFS BRINGS 500 HPA HEIGHTS 4.5-5 STANDARD DEVIATIONS

BELOW NORMAL FOR THIS TIME OF THE YEAR BETWEEN CHARLESTON SC AND

JACKSONVILLE FL...WHICH IS IMPROBABLE. SINCE THE ECMWF ENSEMBLE

MEMBER TIMING HAS BEEN REASONABLY CONSISTENT OVER ITS PAST FEW

DAYS OF RUNS...AND INDICATE A QUICKER PROGRESSION...WILL CONTINUE

TO GO WITH THIS IDEA BY FAVORING A 00Z ECMWF-LIKE SOLUTION...WHICH

LOOKS MUCH MORE PLAUSIBLE THAN ITS PRIOR COUPLE DAYS OF RUNS.

THE SECOND ISSUE INVOLVES A POTENTIALLY CLOSED LOW MOVING THROUGH

THE SOUTHWEST...WHICH BOTH THE 00Z CANADIAN/00Z ECMWF ADVERTISE.

THE 00Z GFS HAS A CLOSED LOW WHICH IT DEVELOPS MUCH FARTHER

OFFSHORE THE SOUTHERN CALIFORNIA COAST...WHICH TAKES AN EXTRA

COUPLE DAYS TO GET INTO THE SOUTHWEST. THE GFS SOLUTION IS SLOWER

THAN THE EITHER CLUSTER OF THE BIMODAL 00Z GEFS MEAN AND MUCH

SLOWER THAN THE 12Z ECMWF ENSEMBLE MEAN SOLUTION. A SLOWLY

PROGRESSIVE NORTHERN STREAM RIDGE IN SOUTHWEST CANADA FAVORS A

PROGRESSIVE CLOSED CYCLONE 20 DEGREES UNDER ITS BASE MOVING

THROUGH THE SOUTHWEST...SO CONFIDENCE IS HIGHER THAN NORMAL IN ITS

EXISTENCE...BUT TIMING IS AN ISSUE...AS IS USUALLY THE CASE LATE

IN THE MEDIUM RANGE PERIOD. SINCE THE CLOSED MEMBERS OF THE 12Z

GEFS MEAN /USING 5520 METER 500 HPA HEIGHT LINE SPAGHETTIS/ WERE

IN LINE WITH A BULK OF THE 12Z ECMWF ENSEMBLE MEMBERS...AND THE

NORTHERN STREAM RIDGE IS ONLY SLOWLY PROGRESSIVE...THINK A QUICKER

PROGRESSION OF THE CLOSED LOW IS ADVISABLE...WHICH IS BEST IN LINE

WITH THE 00Z CANADIAN SOLUTION. THIS FITS INTO THE 00Z GEFS

MEAN/00Z ECMWF COMPROMISE PREFERENCE QUITE WELL.

ROTH

[Road Runner]

The 00z model runs seem to have taken a major step towards downplaying the clipper system. The Euro still has the feature and creates an upper-level low that spins over the Canadian border, but the effects are much less exascterbated. Could the GFS actually have been right all along? It's not the first time this year the GFS saw something before all the other models. Maybe this winter will be a GFS victory over the Euro and CMC because it's an American model and is more experienced with La Nina conditions over North America? I'd really like a met to chime in if possible. Thanks.

No, the GFS is a model and it had many different solutions. The forecaster who picks the correct solution is right.

Look at my first post on November 27th, 2010

My call for December 6th-7th:

Current GFS RUN:

Perfect. Like I said, there is not much potential there.

p.s. The GFS run on November 27th just had a lot of cold air and pushed everything around, refer to my thread.

[SnoSki14]

Well I hope it's wrong because if it's right, then you can forget about anyone in the east getting any snow except around the lakes. The pattern is only going to stay cold enough to support snow for a couple of weeks, then our big La Nina warm up will occur and we'll basically be done for the season. January and February are going to be very warm with lake cutter after lake cutter, that's why I believe many of the seasonal totals in the east, especially in the big cities of NYC, PHL, D.C, and even Boston will only receive between 25-50% of their average snowfall with Boston a little more than that. This is probably going to be like 2007-2008 all over again.

[Analog96]

Well I hope it's wrong because if it's right, then you can forget about anyone in the east getting any snow except around the lakes. The pattern is only going to stay cold enough to support snow for a couple of weeks, then our big La Nina warm up will occur and we'll basically be done for the season. January and February are going to be very warm with lake cutter after lake cutter, that's why I believe many of the seasonal totals in the east, especially in the big cities of NYC, PHL, D.C, and even Boston will only receive between 25-50% of their average snowfall with Boston a little more than that. This is probably going to be like 2007-2008 all over again.

What are you basing that big warm up on? Just ENSO, right?

[isnice]

seriously, I just want comments on model performance

[Lee]

12Z looking much better in the extended again as it, after breaking the block down very temporarily, brings it back with plenty of activity. Of course, just one of it's normal flips @ that range, lol.

[LetItSnowInPhilly]

RoadRunner.....the event is 6 days away and you're already tooting your own horn?

[isnice]

CMC is back on big storm campaign

[Ellinwood]

RoadRunner.....the event is 6 days away and you're already tooting your own horn?

It appears as though he does that a lot... as far as him posting actual verifications with image-to-image comparison, well that's a bit rarer.

[okie333]

Interesting especially with how the HPC crew disco goes...

THE SECOND ISSUE INVOLVES A POTENTIALLY CLOSED LOW MOVING THROUGH

THE SOUTHWEST...WHICH BOTH THE 00Z CANADIAN/00Z ECMWF ADVERTISE.

THE 00Z GFS HAS A CLOSED LOW WHICH IT DEVELOPS MUCH FARTHER

OFFSHORE THE SOUTHERN CALIFORNIA COAST...WHICH TAKES AN EXTRA

COUPLE DAYS TO GET INTO THE SOUTHWEST. THE GFS SOLUTION IS SLOWER

THAN THE EITHER CLUSTER OF THE BIMODAL 00Z GEFS MEAN AND MUCH

SLOWER THAN THE 12Z ECMWF ENSEMBLE MEAN SOLUTION. A SLOWLY

PROGRESSIVE NORTHERN STREAM RIDGE IN SOUTHWEST CANADA FAVORS A

PROGRESSIVE CLOSED CYCLONE 20 DEGREES UNDER ITS BASE MOVING

THROUGH THE SOUTHWEST...SO CONFIDENCE IS HIGHER THAN NORMAL IN ITS

EXISTENCE...BUT TIMING IS AN ISSUE...AS IS USUALLY THE CASE LATE

IN THE MEDIUM RANGE PERIOD. SINCE THE CLOSED MEMBERS OF THE 12Z

GEFS MEAN /USING 5520 METER 500 HPA HEIGHT LINE SPAGHETTIS/ WERE

IN LINE WITH A BULK OF THE 12Z ECMWF ENSEMBLE MEMBERS...AND THE

NORTHERN STREAM RIDGE IS ONLY SLOWLY PROGRESSIVE...THINK A QUICKER

PROGRESSION OF THE CLOSED LOW IS ADVISABLE...WHICH IS BEST IN LINE

WITH THE 00Z CANADIAN SOLUTION. THIS FITS INTO THE 00Z GEFS

MEAN/00Z ECMWF COMPROMISE PREFERENCE QUITE WELL.

I'll take it. Longer-range on the 0z Canadian shows around 1.5" QPF for my area, much of which is snow.

[Dr No]

Everything sets up well for a significant storm around day 6/7. Recent guidance runs that remove the threat are just flavors of the day.

I mentioned that I use Nogaps as a sanity check. It has one of the more surpressed biases that exist. If it continuously surpresses a storm as well as mainstream guidance, that is a likely outcome. 0Z and 12Z Nogaps runs continues to amplify a storm in the SE/TN Valley during that timeframe, so much so, that MA would probably be rain.

That tells me that the mainstream guidance will likely bounce around quite a bit still, as obviously Nogaps is keying on a factor that exists, causing it to supply a more amplified solution. Reality might not give a snow event to anyone, but the threat is there.

[isnice]

Everything sets up well for a significant storm around day 6/7. Recent guidance runs that remove the threat are just flavors of the day.

I mentioned that I use Nogaps as a sanity check. It has one of the more surpressed biases that exist. If it continuously surpresses a storm as well as mainstream guidance, that is a likely outcome. 0Z and 12Z Nogaps runs continues to amplify a storm in the SE/TN Valley during that timeframe, so much so, that MA would probably be rain.

That tells me that the mainstream guidance will likely bounce around quite a bit still, as obviously Nogaps is keying on a factor that exists, causing it to supply a more amplified solution. Reality might not give a snow event to anyone, but the threat is there.

Wise words

[Dr No]

Another note is that the 11/30/2010 18Z GFS ensemble members all have the clipper well-established at hr 84. After that, they show a variety of resulting solutions, some that shear it out entirely, others that give some light snow Sunday AM.

Many members are bullish with a potential significant storm in the east on day 6/7.

http://raleighwx.americanwx.com/models/gfsensemble/members/18zf162.html

[Road Runner]

It appears as though he does that a lot... as far as him posting actual verifications with image-to-image comparison, well that's a bit rarer.

Where did you get this information from? I used my verifications as analogs to forecast a future event that was one month away, and I am doing better than the GFS. I have the right to get excited because of how many things had to happen for this to go as planned. There are forecasters out there who get excited for every potential event on the models, but I have an actual theory behind everything I see. It is too bad you can't understand.

In other words, I not only posted the verification for each call that was one month away, I used those one month verifications as analogs to post another nor'easter call one month away. Because this is my last call and my last attempt for all of you to understand, I not only posted a date but I put up the 500mb analog maps using current observations to come up with a PRECISE forecast for the nor'easter development.

[Cyclonicjunkie]

seriously, I just want comments on model performance

It sucks, it likes making fools out of people.

RoadRunner.....the event is 6 days away and you're already tooting your own horn?

lolz, Roadrunner is a good one.

[Ellinwood]

Moving this into Road Runner's thread to keep from derailing this one...

[isnice]

I think it's pretty safe to say now that the GFS was right. The model barely even budged from its original solution, yet we all doubted it as the Euro and GGEM changed constantly. What do we think? How does the GFS have a better handle on these storms thus far?