February 8-12 model discussion

[FoothillsNC]

00z GFS Ens mean looked close to it's OP run...the 6z Ens mean however looked more south and east with the low track.

The worst part might be when the Euro goes to lunch like it always does and gives us hope.

The Euro might just be amplifying the PNA too quickly/sharply, which is why it digs the s/w so far west. That might be an error or bias of that model, along with the fact it overdoes the eastern Canada vortex frequently. The truth probably will lie in the middle...so the key thing to watch next run is the strength of the Newfoundland vortex and the shape of the central US trough, as it is tied with the flow and PnA out west. The lead s/w Monday coming through the east is the one that could bomb out (or not) which affects the flow which steers the next s/w. Its really another 'thread the needle" tandem play , which is always hard to do. But not impossible...so at this point anything from a Gulf coast to east coast hugger to midwest track is on the table since its far enough out.

[Amos83]

As Robert has mentioned to avoid an apps runner or to far inland track we need the lp in eastern Canada to really wind up. The stronger it is the better because it will not only prevent our storm form cutting but will also help lock in the cold air ahead of it.

Plenty of time for things to trend back in our favor. Of course also plenty of time for more cutting as well. I am happy to see the ensembles further east than the op runs.

[burgertime]

The Euro might just be amplifying the PNA too quickly/sharply, which is why it digs the s/w so far west. That might be an error or bias of that model, along with the fact it overdoes the eastern Canada vortex frequently. The truth probably will lie in the middle...so the key thing to watch next run is the strength of the Newfoundland vortex and the shape of the central US trough, as it is tied with the flow and PnA out west. The lead s/w Monday coming through the east is the one that could bomb out (or not) which affects the flow which steers the next s/w. Its really another 'thread the needle" tandem play , which is always hard to do. But not impossible...so at this point anything from a Gulf coast to east coast hugger to midwest track is on the table since its far enough out.

So if that s/w really bombs out that keeps the system behind it suppressed and forces it to ride up the coast right? Sort of acting like a mini vortex?

[FoothillsNC]

So if that s/w really bombs out that keeps the system behind it suppressed and forces it to ride up the coast right? Sort of acting like a mini vortex?

Yep. The Euro looks like its placing more emphasis on the 2nd wave and doesn't do much with this Monday wave, its been back and forth on its strength. My hunch is its actually overdoing the PNA out west this time, which forces the 2nd s/w too far to the southwest. Could be wrong though.

[valkhorn]

06gfs COLD

Let see if this happens.. lol

It has before. January 1985, January 1988, and February 1996 were very similar setups with heavy snow through TN and extreme cold after it.

[bingcrosbyb]

The Euro might just be amplifying the PNA too quickly/sharply, which is why it digs the s/w so far west. That might be an error or bias of that model, along with the fact it overdoes the eastern Canada vortex frequently. The truth probably will lie in the middle...so the key thing to watch next run is the strength of the Newfoundland vortex and the shape of the central US trough, as it is tied with the flow and PnA out west. The lead s/w Monday coming through the east is the one that could bomb out (or not) which affects the flow which steers the next s/w. Its really another 'thread the needle" tandem play , which is always hard to do. But not impossible...so at this point anything from a Gulf coast to east coast hugger to midwest track is on the table since its far enough out.

Very true. And even just comparing the 12z NAM @ 84 vs the 00z GFS @ 96 with the first s/w, the western ridge is not nearly as amplified on the NAM (granted it's waaay out there). As a result, both models handle the first s/w differently (NAM keeps two pieces separate while GFS merges the two on the coast). Lots to be ironed out. But if the PNA doesn't amplify too quickly......

http://www.nco.ncep....am_500_084l.gif

http://www.nco.ncep.noaa.gov/pmb/nwprod/analysis/namer/gfs/00/images/gfs_500_096l.gif

[griteater]

Robert, Why is the current storm tracking up the east coast without any blocking?

Actually, all of the models show 2 weakish sfc lows with today's system. One going west of the apps and another going up the coast. It makes sense that there would be a sfc low west of the apps given that the upper low vort tracks from W TN to WVA.

[griteater]

That's too far out for the GGEM to be of any use, in my opinion.

GGEM has been pretty good in my view this winter in terms of recognizing the NW trend early on.

[tnweathernut]

GGEM has been pretty good in my view this winter in terms of recognizing the NW trend early on.

From my memory, without checking, the GGEM was the model that first hit on the idea of snow from TX into west TN with this system. It actually seems like it performed very well for what is going on right now.

[valkhorn]

Very true. And even just comparing the 12z NAM @ 84 vs the 00z GFS @ 96 with the first s/w, the western ridge is not nearly as amplified on the NAM (granted it's waaay out there). As a result, both models handle the first s/w differently (NAM keeps two pieces separate while GFS merges the two on the coast). Lots to be ironed out.

http://www.nco.ncep....am_500_084l.gif

http://www.nco.ncep.noaa.gov/pmb/nwprod/analysis/namer/gfs/00/images/gfs_500_096l.gif

Agreed. Although using the NAM@84 is a bit of a stretch, I do think the Euro may be in the wrong camp. The GFS has shown remarkable consistency and the ensembles seem to be siding with the GFS. The mid-range GFS isn't as bad as the long-range Euro in my opinion, and considering this exact setup has happened before in TN, history is on our side.

I don't have access to the pay versions of the Euro, but if it's not dumping a lot of snow then that's why it's not showing it to be as cold I think. Historically cold nights in our area are almost always tied into having decent snowcover of at least 6" or more.

Also, looking back at climatology, East TN is more likely to get heavy snows in the valley through deformation bands or powerful lows like in March 1993, or decent overrunning precip like in Feb. 1996 or Jan. 1988. Jan 1985's snow wasn't by any means a record setting snowfall in Knoxville, but the 850mb isotherms were approaching -25 to -30C over our area _with_ snowpack. Interestingly enough, Knoxville's heaviest 1 day snow was all the way back in the 1950s, and was due to a heavy deformation band on the back side of a low.

[valkhorn]

From my memory, without checking, the GGEM was the model that first hit on the idea of snow from TX into west TN with this system. It actually seems like it performed very well for what is going on right now.

I remember it was all over the map in the extended range. It first had a heavy band of snow for central MS and AL, then switched to heavy IP for the same area. It has given us in East TN decent QPF snows to all rain. It may have accurately got the setup in the mid-range, but by no means did it show it correctly in the long range at lower resolution.

[griteater]

I remember it was all over the map in the extended range. It first had a heavy band of snow for central MS and AL, then switched to heavy IP for the same area. It has given us in East TN decent QPF snows to all rain. It may have accurately got the setup in the mid-range, but by no means did it show it correctly in the long range at lower resolution.

Which one did?

[tnweathernut]

One thing that is really interesting to note is the fact the models did BUST horribly on what is happening now (for northern AL at least, and probably other areas as well). They were largely predicting temperatures for that area of around 40 and acual observed temps were in the upper 20s. Some areas saw as much as 4 inches of snow there. Also interesting was the models were about 6 hours too slow with the precip. This particular airmass was the little brother to what is coming down next week.

If the models busted this poorly with a stale airmass, surely there will be surprises with the next airmass and storm too. I have a gut feeling that someone will be very happy with this next system, just don't know who yet. I can say the southern apps (around KTRI) are due for a pounding. Nickle and dimed all year to this point.................. and really since 1996.

[valkhorn]

Which one did?

Exactly. Long range is long range. This is in the GFS's mid range, the GGEM's long range, out of scope of the NAM, and I'm pretty sure the long range of the Euro.

Models tend to have a better track range in the short-range, then mid-range, then long range. So, I'd take the mid range GFS and ensembles, than the long range Euro or GGEM.

[tnweathernut]

I remember it was all over the map in the extended range. It first had a heavy band of snow for central MS and AL, then switched to heavy IP for the same area. It has given us in East TN decent QPF snows to all rain. It may have accurately got the setup in the mid-range, but by no means did it show it correctly in the long range at lower resolution.

Was speaking to its look from 3-4 days out..........and in, but as for the bolded, you are probably right.

[griteater]

Not only do we have to get the storm suppressed to the south, there are other obstacles to overcome with low level temps given the high sliding off the mid-Atlantic coast with return flow from the SE, and a weak sfc low showing up over the Great Lakes - both no-no's in the SE for snow.

[valkhorn]

One thing that is really interesting to note is the fact the models did BUST horribly on what is happening now (for northern AL at least, and probably other areas as well). They were largely predicting temperatures for that area of around 40 and acual observed temps were in the upper 20s. Some areas saw as much as 4 inches of snow there. Also interesting was the models were about 6 hours too slow with the precip. This particular airmass was the little brother to what is coming down next week.

If the models busted this poorly with a stale airmass, surely there will be surprises with the next airmass and storm too. I have a gut feeling that someone will be very happy with this next system, just don't know who yet. I can say the southern apps (around KTRI) are due for a pounding. Nickle and dimed all year to this point.................. and really since 1996.

Exactly. They didn't account for the temperatures at the onset of the precip mainly because they didn't correctly estimate the evaporative cooling that would occur. The Jan 10 storm for Knoxville had accumulations that were cut somewhat because it took a while for the air to become saturated, and the models didn't handle that well I remember. We also weren't aware of the heavy deformation bands that set up on Jan 10 and Dec 25th until a couple of days before the event.

Yesterday at my parents house in Meridian, MS, they had about a quarter inch of freezing rain with temps in the upper 20s and it caught everyone off guard.

In the long range, I like to look at general patterns and climatology.

In the mid range, I like to look at the general setup with any cold and precip, but it's by no means specific.

in the short range is where I hash out the details.

I can't be very specific, but I do think I40 really has good odds next week since the cold will be there and the moisture doesn't seem to be dying off anytime soon. The odds in my opinion would be better if you go further north at this point than if you would further south - but the southern TN border has done VERY well this winter, so who knows.

[tnweathernut]

Not only do we have to get the storm suppressed to the south, there are other obstacles to overcome with low level temps given the high sliding off the mid-Atlantic coast with return flow from the SE, and a weak sfc low showing up over the Great Lakes - both no-no's in the SE for snow.

Also agree with this.....alot actually. Storms like this usually have a warm nose somewhere that is underestimated - even in the shorter range 2-3 days.

[tazaroo]

Not only do we have to get the storm suppressed to the south, there are other obstacles to overcome with low level temps given the high sliding off the mid-Atlantic coast with return flow from the SE, and a weak sfc low showing up over the Great Lakes - both no-no's in the SE for snow.

Almost every storm we have tracked this year has had that pesky great lakes low. I don't know why that is unless its Nina climatology.

[griteater]

Almost every storm we have tracked this year has had that pesky great lakes low. I don't know why that is unless its Nina climatology.

I think it goes back to not having the good west based -NAO blocking. If we have that, you get confluence at the upper levels over the Great Lakes and solid high pressure. Without it, upper waves are allowed to flow into that area, with associated sfc lows. So, to get anything to work in this pattern, we just have to have perfect timing.

[valkhorn]

Not only do we have to get the storm suppressed to the south, there are other obstacles to overcome with low level temps given the high sliding off the mid-Atlantic coast with return flow from the SE, and a weak sfc low showing up over the Great Lakes - both no-no's in the SE for snow.

Depends on your definition of the southeast. Here in East TN, it can be a completely different situation altogether. Not only can we have apps runners and backend snows, we can have overrunning precip and clippers that can give us a good 1" of snow in the valley and 8-12" in the highest elevations. When I tell people who don't live here how often Mt. Leconte at 6600' gets a foot of snow or more from a clipper, northwest flow, and a bit of lift, they find it hard to believe - and it's almost never within the resolution of most models. The NWS at GSP does a great job of hammering out specifics for that, and MRX helps out, too. Mt. Leconte gets a good 80-90" of snow each season.

I'm sure the situation is quite different in the NC high country with CAD, so until I've lived there it's hard to speculate.

A good example of how odd snow can work around here is that last year NE TN from 1500-2500' got 3 to 4 feet of snow overall. This winter, CHA has had two very large snows and that's about it. We have had several smaller <4" snows and a few half inch and inch clippers in TYS. We've averaged about the same amount of snow that CHA has, and this winter TRI hasn't done well at all.

[valkhorn]

Also agree with this.....alot actually. Storms like this usually have a warm nose somewhere that is underestimated - even in the shorter range 2-3 days.

Exactly. Climatology would pretty much argue against such an unusual event similar to 1988 or 1985. It's very rare - but it has happened. Each time wasn't very well forecasted either. Even March 93 was underplayed significantly until it actually happened.

I still say with the setup we probably have a good 20 to 25% chance of it happening, which is not the best odds in the world but it's the best I can give this far out.

[tazaroo]

I will say that some of the 6z gfs ensemble members look pretty good for a se winter storm. There were only a few real apps runners and also a few suppressed cold front passage scenarios.

I would definitely take my chances with the gfs right now. I hope it is beating the Euro this time for once.

[tnweathernut]

Exactly. Climatology would pretty much argue against such an unusual event similar to 1988 or 1985. It's very rare - but it has happened. Each time wasn't very well forecasted either. Even March 93 was underplayed significantly until it actually happened.

I still say with the setup we probably have a good 20 to 25% chance of it happening, which is not the best odds in the world but it's the best I can give this far out.

Still very far out, but if I were in middle TN and/or west TN and north of 40, I think those odds are about double your 20-25% even this far out. Basically making them a coin flip which from 5-6 days out I'd take every time.

On a side note, it's impressive how a coin can land on tails 6, 7 times in a row when you are looking for heads.

[valkhorn]

I will say that some of the 6z gfs ensemble members look pretty good for a se winter storm. There were only a few real apps runners and also a few suppressed cold front passage scenarios.

I would definitely take my chances with the gfs right now. I hope it is beating the Euro this time for once.

Yah, I agree. Although I'll readily admit I'm well aware of my confirmation bias for East TN. I only wish we had the GFS for the past 100 years, then I could tell how it handled similar situations before in the mid-range. The best I can do at this point is look at weather maps of similar events that I know happened and try to figure out a pattern - but unfortunately these are so rare and the way TYS can get heavy snowfall can be so varied that I can't really put all my eggs in the GFS basket, but if it's consistent for the next 2 or 3 days then I think we'll have something.

12Z GFS will show us what's what within the hour. Even if it drops it completely, it could show up again on the 0Z. Also, at this range, I don't really pay attention to 6Z and 18Z runs. I do when it's in a shorter range though.

[tnweathernut]

In the next day or two, the models should start coming together and giving us a consensus for where this low will form and how it will track. WIll be fun to follow if the GFS holds it's course and/or trends a little colder. Once under 3-4 days I don't think we will see any major shifts. Unfortunately, once we have a good general track at that time, we have a whole other list of things we will have to watch to get a good snow here in the southeast

[valkhorn]

Still very far out, but if I were in middle TN and/or west TN and north of 40, I think those odds are about double your 20-25% even this far out. Basically making them a coin flip which from 5-6 days out I'd take every time.

On a side note, it's impressive how a coin can land on tails 6, 7 times in a row when you are looking for heads.

Good point. BNA historically has not had the heavy types of snow that CHA and TYS and TRI have had. BNA's 24 hour record I believe is under 12", while TYS has had a 24 hr record of 17", and CHA 20". TRI I'm not sure about, but I'd hazard a guess with the 2' storms they've gotten over the years it would probably be around 18-22".

[beneficii]

Exactly. Climatology would pretty much argue against such an unusual event similar to 1988 or 1985. It's very rare - but it has happened. Each time wasn't very well forecasted either. Even March 93 was underplayed significantly until it actually happened.

I still say with the setup we probably have a good 20 to 25% chance of it happening, which is not the best odds in the world but it's the best I can give this far out.

Really? I had heard that the models had locked onto the Superstorm about 5 days out.

[valkhorn]

Really? I had heard that the models had locked onto the Superstorm about 5 days out.

That's not the whole story. They were aware that something was going to happen, but they had no idea of the degree that it happened. Accumulation forecasts about 24-48 hrs before the event were severely underdone.

Case in point: Meridian, MS (where I was living at the time) had only a forecast of a dusting of snow less than 12 hours from the event. I remember the transition from Winter Weather Advisory to Winter Storm Warning to Heavy Snow Warning as the snow was falling over the course of just a few hours.

BHX hadn't forecasted the 17" they got either. There are videos on Youtube where James Spann was very surprised at how much snow was actually going to fall at the last minute. They knew there was going to be some snow, but they certainly didn't forecast over a foot with near blizzard conditions and thunder 24 hours before the event.

[Tacoma]

this storm next week will be gulf coast up eastern seaboard, the strong cold air will keep this storm south is what i'm seeing.