February Med/Long Range Discussion Part 2

[HighStakes]

I think that storm may have been the impetus for the 18z gfs "happy hour" slogan. There was an 18z run of the gfs that gave us a blizzard-type storm 120 hours out (as any good weenie, I still vividly remember those maps and the classic cbb.) Of course, the gfs lost it but brought it back and low and behold, it came through. You're right though, the 13" at BWI was gone fast, in about 3 days if memory serves me. The icing on a crapastic winter.

That storm did disappear fast. I got 16 in Reisterstown and it sublimated to around 10-12 by early afternoon. I don't think it is possible for a storm to produce that much snow to have such a limited impact. We did have a radio show before that storm and nobody was super excited. I think HM was the only one who picked up on the possibility of the crazy deform band that exploded in the early morning hours.

[mitchnick]

Although the storm we're talking about is not only a thread the needle event but also a flawed event cured by cold air advection, the fact that the Euro and both gfs models show it with a decent ending is striking. This is not the kind of storm scenario one would expect to be modeled by both the gfs and euro, but there it is. I can only take that as a positive at this point.

[mattie g]

12z GEFS analogs centered on Feb 29 have a solid signal.. prob one of the best of winter. For DCA, most are cold around period. 7 of 10 have accumulating snow within 5 days either side. 4 have 3"+ events, one just missed. Feb 1 2010 is on there (2010 did show up at times prior to the blizzard).

This is a very Wes-y post.

[Ian]

This is a very Wes-y post.

I try to learn from the best.

[cae]

I couldn't care less what the GGEM shows and anyone who does, especially at this range, might want to seek counseling.  The GGEM ensembles are a decent tool but the op is pretty bad. 

 

Heh.  Sign me up for counseling.  Seriously, none of the op runs mean much 8 days out.  It's best to take them together as an ensemble, and the GGEM is useful as a member of that ensemble.  The fact that we've recently had Euro, GFS, and GGEM op runs show hits is a positive signal.  The ops seem to be a little more bullish about this storm than their corresponding ensembles.

 

FWIW, the 8-day verification scores for the GGEM aren't much below that of the GFS, and by 6 days the gap between the two is pretty small.  I mentioned earlier that this appears to be a seasonal pattern - in the summer, the gap widens.

[psuhoffman]

Heh. Sign me up for counseling. Seriously, none of the op runs mean much 8 days out. It's best to take them together as an ensemble, and the GGEM is useful as a member of that ensemble. The fact that we've recently had Euro, GFS, and GGEM op runs show hits is a positive signal. The ops seem to be a little more bullish about this storm than their corresponding ensembles.

FWIW, the 8-day verification scores for the GGEM aren't much below that of the GFS, and by 6 days the gap between the two is pretty small. I mentioned earlier that this appears to be a seasonal pattern - in the summer, the gap widens.

I've been through this debate too many times. Those scores are based on h5 verification over the whole northern hemisphere. I don't care what they say as my focus is much more specific. I base my analysis on years of observing the models abd their tendencies. The ggem is just awful compared to the gfs and euro with winter east coast storms.

[gymengineer]

Pretty much my sentiments on that winter and the storm.  Don't remember anything about December, January was a train wreck of warmth (similar to, but not quite as warm as, this past December's torch).  February was meh and warm, other than the one event.  Talk about a "one and done" season, for the most part!  

December '05 for the region was a 5 day period that produced with a fluffy clipper followed by a snow-to-heavy sleet-to light rain storm on the 9th. In MCPS terms-- it was a 2-hr delay followed by a closing  

[cae]

I've been through this debate too many times. Those scores are based on h5 verification over the whole northern hemisphere. I don't care what they say as my focus is much more specific. I base my analysis on years of observing the models abd their tendencies. The ggem is just awful compared to the gfs and euro with winter east coast storms.

 

Yeah, I'd read that here, but when I went back and compared old model runs to the actual storms I didn't see it.  For example, these were the GFS and GGEM runs a week before this most recent storm:

 

 

 

Clearly neither was right (they never are), but to my eyes the GGEM was closer.  There were other runs where the GFS was closer to the final solution, and other runs where the GGEM was closer.  You look at enough of those, and then you try to figure out how each model is doing on average... and that's what led me to verification scores (and not just H5).  You're right in that they cover the whole NH, but they seem to match local reality pretty well.  A couple of years ago, when the GGEM was doing well, it was reflected in the verification scores.  And last year, when it was struggling in the beginning of the season, that showed up as well.

 

I respect your knowledge and your opinion, and I think you make some great posts.  But I still think the ensemble of ops, even the lesser ones, is a better forecasting tool than any individual op run.  We might just have to disagree on this one. 

[Amped]

00z GFS is rain. Gotta get that central Canada vortex out of the way like the 12z Euro, or we're toast.

[AlaskaETC]

What does the CMC say?

 

ETA: Big whiff

[cae]

What does the CMC say?

 

ETA: Big whiff

 

Out to sea.

[Deck Pic]

The Winchester crew will like the 0z Euro

[Heisy]

Lol I just got back from work no posts from the EURO? It has an insane 500mb signature. Huge hit for NW burbs. 

[Heisy]

Also the 00z UKMET @ 144 looks really good with a 50/50 low and HP building in. 

[AlaskaETC]

The Winchester crew will like the 0z Euro

The interior is due for a big event that screws the cities

[Blizz]

Lol I just got back from work no posts from the EURO? It has an insane 500mb signature. Huge hit for NW burbs. 

 

Clown map

[showmethesnow]

The 12Z Euro ensembles hinted at this and I was pleasantly surprised to see the 00Z op run actually throw out this solution. While I normally don't get too worked up over an op run 7-8 days out the 00Z Euro was a fun one to look at and is best case/best upside for DC/BALT region with such marginal temps to begin with. Besides the obvious, the nice positioning of a 1040 banana high to the north and west as well as a decent track for a low coming out of the gulf it had one other feature that stood out. It develops a vertically tilted trough while the low is still around the outer banks and then pulls that trough up placing the DC/Balt corridor in the prime region/time to be with such a feature before the low vertically stacks and occludes. I know the models show the temps to be marginal for the 95 corridor but with the setup being shown I think you could expect to see the 700's on down tucked much closer into the low on an axis from the surface low to the 700 low. So though they may start out as rain with that setup I would expect they would see a fairly quick switch over to snow and probably lots of it. 

[CAPE]

Looking at the EPS indv members, there isnt great support for the op. Plenty of misses(or rain), a few snowy solutions in the metros, a few more for the mountains. Long way to go, but not a strong signal as of now.

[WinterWxLuvr]

Looking at the EPS indv members, there isnt great support for the op. Plenty of misses(or rain), a few snowy solutions in the metros, a few more for the mountains. Long way to go, but not a strong signal as of now.

Not much support, I agree. Long way out too. It could be just about anything by the time we get there. Latest gfs isn't all the way out but it still has the storm so who knows. I have to travel next Wed so there is that.....isn't that the way luck goes.

[showmethesnow]

Looking at the EPS indv members, there isnt great support for the op. Plenty of misses(or rain), a few snowy solutions in the metros, a few more for the mountains. Long way to go, but not a strong signal as of now.

The started looking over the various ensemble runs so haven't looked into the individual members. So are you saying they don't support the op with the evolution of the low or are you saying the snowfall maps don't show a similar solution? 

[showmethesnow]

After looking over the 00Z ensemble runs all I can say is that the Euro is the one we should probably be concentrating on for the coastal, at least for now. CMC has been bouncing around like a kangaroo hopped up on candy, well maybe not that bad. And the GFS hasn't been much better. Euro has been, for the most part, rock solid with only minor adjustments from run to run and if the solution that it has been showing for awhile now comes close to verifying it will have had an excellent two week stretch running back to the previous storm. The GFS and the CMC though are another story. They were horrible leading up to the last storm and for this coastal about the only thing that can be determined from them is that they favor a coastal low. Maybe. 

 

Anyway, the Euro for the most part has stayed very similar to previous runs  with a slight adjustment of the coastal east and has dampened the higher pressures to the north a touch. All and all a decent look if we can get the temps to cooperate.

 

Edit: Wanted to add that the 00Z ensembles mean had the same slowing down, semi stalling of the low seen with the 12Z ops that suggest the possibility of a capture of the low ongoing.

[Heisy]

6z GFS has a titanic storm. Surface is too warm for cities, but just from a meteorological standpoint it is an absolute monster of a storm. 

[33andrain]

06z GFS 192-195 is beautiful

[CAPE]

The started looking over the various ensemble runs so haven't looked into the individual members. So are you saying they don't support the op with the evolution of the low or are you saying the snowfall maps don't show a similar solution? 

There is the usual spread on low locations. A decent cluster off the NC coast, but some would be misses, and some others rainy. Storm signal is there, just not sure how snowy. There are some key pieces showing up in the guidance, but exact placement and timing are unknowns. Plenty of time, and we at least have something to track at this point.

[showmethesnow]

There is the usual spread on low locations. A decent cluster off the NC coast, but some would be misses, and some others rainy. Storm signal is there, just not sure how snowy. There are some key pieces showing up in the guidance, but exact placement and timing are unknowns. Plenty of time, and we at least have something to track at this point.

Actually did a quick glance of the individuals. There are quite a few members that show the signature stall of capture or partial capture with many of those being in a somewhat favorable position for the MD/NVA/SEPA region. Most of the big hits/near misses on the snowfall maps are captures or what look to be captures. Not saying that this is what will verify come game time nor that a capture will occur in a favorable location but this would be the best solution to expect with marginal temps as well as it having the biggest upside on snowfall potential. 

[psuhoffman]

Yeah, I'd read that here, but when I went back and compared old model runs to the actual storms I didn't see it. For example, these were the GFS and GGEM runs a week before this most recent storm:

Clearly neither was right (they never are), but to my eyes the GGEM was closer. There were other runs where the GFS was closer to the final solution, and other runs where the GGEM was closer. You look at enough of those, and then you try to figure out how each model is doing on average... and that's what led me to verification scores (and not just H5). You're right in that they cover the whole NH, but they seem to match local reality pretty well. A couple of years ago, when the GGEM was doing well, it was reflected in the verification scores. And last year, when it was struggling in the beginning of the season, that showed up as well.

I respect your knowledge and your opinion, and I think you make some great posts. But I still think the ensemble of ops, even the lesser ones, is a better forecasting tool than any individual op run. We might just have to disagree on this one.

That's because your only comparing it to the gfs which was equally awful. The U.K. And euro had the right idea. The gfs added confusion. Looking at the ggem just added more confusion not more clarity. More wrong answers never get you a right answer. The gfs is right more often so it makes the cut, barely, but there is a threshold where your just increasing noise with too many permutations. Plus the gfs has some very clear biases that I am familiar with so perhaps I will admit this is personal preference but the ggem just isn't right enough with the track of east coast systems for me to give it any weight. It's like the jma navgem or any other obscure model. It's there. It's better if it looks good then bad but it's only worth a glance nothing more.

[kdskidoo]

The interior is due for a big event that screws the cities

LOL I've been saying this for a few years now, but it hasn't happened yet.

[psuhoffman]

LOL I've been saying this for a few years now, but it hasn't happened yet.

I wonder if that changes when the amo flips. Warm water is likely to increase the baroclinic zone off the coast and draw storms off the coast.

[showmethesnow]

Nice to see that the 06Z GFS ensembles have once again found the coastal low that the 00Z run had seemed to misplace.

[psuhoffman]

While everyone squabbles over every 50 mile wobble of the day 7 storm the real news is the coming window looks better and better each run. Eps by day 15 looks nice. High heights all across Canada from the epo to the nao domain and a strong trough signal in the east for day 15 centered to our south. Gefs has the high latitude blocking look but then somehow has higher heights across the conus also. Kinda weird given the look to the north but the gefs have been trying to do that all winter so I've started to ignore them when we're headed into -epo periods. Bottom line we have an extended window coming with a favorable pattern for a late season snow.