Tuesday, March 5, 2013

NO major changes- 1-81 special, north of Lexington!

Last nights model data went with no surprises..

GFS moved towards the EURO

EURO moved towards the GFS.

2-4 inches LYH and ROA, more just north. Parts of Amherst may see close to 10 inches.

460 is the southern extent of "decent" snows--

We get a TON of rain before hand.

FWIW, Short term model NAM is a tad warm, we don't get much accumulation from that one.

Monday, March 4, 2013

Some thoughts before the evening model chaos

Hard to blog when I, well-- work a real job. :)

This morning I hinted that the 6-14 forecasted by the NWS for Lynchburg was likely too high and they've adjusted everyone region wide down, with highest amounts of 6-10 for Amherst  and Rockbridge county.

1. My statement of the ULL needing to touch NE South Carolina is huge-- and this won't happen. HUGE snow option pretty much off the table Lynchburg. Track now is MAYBE just north of 1-40 in North Carolina. If this had very strong cold air, we could deal with this, but in March with no parent high is a no go. 

2. The trend has been more rain and less snow. This was obvious before the 6-14 call this AM from the NWS, but it was "watch" not "warning". My guesses have been a bit high as well. One issue is assuming a trend stops. As this as trending more north and warmer, assuming 36 hours this is done is foolish. So, my current thoughts of 2-5 may be high. 

3. I need to follow my own rules of things ALWAYS trend north. This was NOT a confluence instance (Like Feb 19th last year) that kept it south, but vort interactions. That's a different breed. Some of the data now makes this a full blow Noreaster that runs up the east coast bring big snows along or maybe just west of   of I-95. 

4. The 18z GFS showed another way to get close to 6 inches in Lynchburg. The low takes the further north route and we do get decent rains. However, it REALLY cranks up around SE VA and sends warm air WELL to the North of the low. this pushes the snow bands well to the west and even south west where the decent, accumulating snows make it back to close to 460 in our region, especially east of Roanoke. 


So, with that-- my current thoughts which I will update later tonight or early tomorrow.



For  ROA- 

Mainly rain, ending as a snow, under 2 inches 60%
Ends as snow 2-5 inches 40%
Over 5 inches 10%

For LYH

Mainly rain, ending as snow-- under 2 inches 40%
Ending as snow 2-5 inches 40%
over 5 inches- 20$

Danville--
Anything above 2 inches would shock me. 

NRV- gets a little crazy with elevation. maybe 2-4 inches.

** Edited to say-- you can see how precarious this is** I can totally see Amherst getting 10 inches in the city and by the time you hit Fosters Fuel on 29 it's 2 inches. 


Winter Storm Watches HOISTED..

I'm impressed with the National Weather Service from Roanoke- they just went BIG!

Winter Storm Watch western areas 4-8 inches possible (Including Roanoke, Blacksburg)

Winter Storm Watch eastern areas 6-14 inches north of the Staunton River. (Campbell County, Lynchburg north and east)


And, they may go down in flames. :( My confidence is not that high.


It's a battle of the models.

-

1.GFS, NAM, and SREFS: The crank up the low a JUST smidge north and east and we get mainly rain as the incredible snow bands sit somewhere closer to Charlottesville.

2,GGEM, Euro and UKMET are more south with the low and keep a heavy band, especially over northern and eastern areas much of Wednesday.

What do these have in common? The First set upload data in something called 3dvar (3d variational data assimilation) and the second in 4dvar (4d variational data assimilation)

4dvar is an upgraded model and has tested to be superior in the big picture, but has lost some individual events so to speak. The NAM and GFS has both showed the big event at times, but the overall trends have been a NEAR miss. The Euro has been consistent and just varied a little the strength.

What is the differences on a functional level?

The biggest differences is how quickly this storm "winds up" and exactly where. The 4DVAR models are convinced it will be more to the south and west and hang that thing back longer to get us under a good snow band for hours. There are a ton of players this system.

What do you think, Keith?

I'm so glad you asked. :)

Confidence is low. The consistency of the Euro I like, despite being a little south of it's ensembles. (Ensembles are the same model run 50 extra times with minor changes to weed out flaws in modeling)
However, at this time I would have liked to see the 3DVAR group move towards the 4DVAR group. They flirt, but don't commit.

Some of the models literally every 5 miles you head North east out of Lynchburg ads an inch to your total.

My current guess-
Blacksburg and Roanoke 3-6
Lynchburg  4-8
Amherst Appomattox- 6-10 maybe more in the higher elevations.

If the banding sets up a tad south, these could be higher, especially Lynchburg region.

Per the norm, will update this later-- not a lock at all. When banding sets up, it gets crazy-- like trying to guess exactly where a thunderstorm will form.

Sunday, March 3, 2013

This is a madhouse-- Ash Wednesday Snowstorm

Feels like being cloned.

More detailed update. 

The devil is in the details. Models that are weaker on the block over New England are further north, while those that are stronger are more south. It keeps the energy suppressed to the south.

I felt great after the 12z runs until the 18z GFS which was more north. The other models are a tad north as well, but the EC is about the same for our region and bit south overall. It has about 2 inches of liquid that falls mainly as snow.

Here's the blocking map I made earlier-- the block and 500mb low (vort) if as modeled on the EC will keep this storm in a great position for us.





Any changes to your thoughts??

1. Despite the EC and UK met staying south, the north trend concerns me. The blocking is strong--a few mets who I REALLY respect and trust are harping how strong the block is. If the block over Canada is as strong as modeled on the ECMWF, we should get a major snow event.

2. It's a lock we get SOME snow. Maybe 2 inches, maybe 12. I'll break down the chances later.

3. Banding will exist-- some winners and some losers. We've had too many events WAYY to high on total precip on the models.

4. Temps- BARELY cold enough in the Lynchburg to Roanoke area. Danville seee's a TON of rain before snow. We won't get the usual 10-11 -1 ratios. It will be more like 7-8 to 1.


Vegas odds-- Snowstorm total. 

1. For Lynchburg and Roanoke
2 inches 80%
6 inches 50%
10 inches 25%
15 inches 10%

2. Martinsville-Danville.

 2 inches 60%
4 inches 40%
8 inches 20%

As we get close, we will massage these numbers. 

If you wanted exact ranges now-- my guess would be

Higher elevations 8-12
Roanoke to Lynchburg 6-10
Southside 2-5. 


Saturday, March 2, 2013

Quick thoughts before the evening models

Just wanted to post a quick map before the evening data. Each model run brings us closer to a REAL solution. A big event is happening as the main shortwave is getting closer and closer and should be sampled better.

Most of the data was good if like big snows. However, the 18z GFS took the more north track and was a TON of rain ending as some snow.  More like 2-6 inches rather than the BIG event I want.



I tried fooling with GIMP software and it was not cooperating with me. :)

So, this was done ghetto style on one note.

1. I drew two red lines-- One goes down into SC and NE towards Hampton Roads.
the second line goes across Kentucky, along the VA NC line and near Hampton roads.

These are the two presented tracks for the upper air low. For Lynchburg and Roanoke to get a big snow, we REALLY need the southern track. If it touches down in South Carolina, we are golden. If it takes that more northern track we will mix with rain for a while and it gets more complicated.


2. I drew those blue lines up east of Maine, near Nova Scotia, That is tied into point 1. The stronger, more west that it, the better chance we have of driving that energy down into South Carolina. When the 18z GFS had the more north solution, that blocking was weaker.(There is an upper low tied to that blocking as well called a 50/50 low)


While I liked the 12z runs and think it's plausible, as stated-- the north trend ALWAYS scares me.

Help, I'm stepping into the twighlight zone...

Well, most of the evening model data was promising if you are a snow lover. I tried to stay awake for the ECMWF or Euro and nodded off around 60 hours. Waking up and seeing it was UNREAL.

The Euro was a madhouse-- with literally 15-25 inches of snow from Roanoke to Lynchburg with maybe close to 30 inches in places NW of Richmond.

Don't take this literal-- at all, yet. Potential has certainly grown and we could be staring down a potentially historic weather event. We've moved away from a full phase event where two vort maxes merge near our region and more towards just a strong closed upper air low being pushed south of our region and slowing down due to the stronger blocking. The low gets vertically stacked which means the surface low and upper air low are on top of each other and it's a beautiful event.

Here's a link to an article about the Ash Wednesday storm. This guy was up towards Massanutten, but it conveys how intense that storm was. This storm, 51 years later, would fall on Ash Wednesday. Lynchburg had almost 18 inches of snow from this event that stalled off the VA coast for day and was one of the costliest storms ever in terms of coastal beach erosion.

http://www.glenallenweather.com/links/digest/greatestsnow.htm



What could go wrong??

We are still 100 hours away. There will be changes in the final solutions.. Here's a quick guide to ways to sabotage our winter event.

1. Temp profiles are not horrible, but March with sun and no cold air advecting. In our region of the state, with a late afternoon Tuesday start, it may start as rain. Models are tad warm at the surface and this is always a concern. Taken literal, the EURO has rain in Danville for HOURS after it's snow in Lynchburg and Roanoke. Models TEND to be a bit cold at this time frame in general. So, we could get the BIG event, but too much rain that limits it.

Further, if we don't get quite as heavy stuff--we could get what many refer to as "white rain" where it snows for hours but doesn't lay. The

Once we get under the "deformation zone" we should get plenty cold-- but deformation = banding. I've mentioned before how banding can be hit or miss. This could start and end as rain-- once the storm occludes, it quits making cold air.

2. The upper air low doesn't dig as much. This would cause both temperature issues AND less precipitation. From my experience, you want the upper air low to travel about along 1-40 in North Carolina to max things out here. This is a very strong low and moves from North Central South Carolina NE towards let's say Kill Devils Hill on the Outer Banks. That's an incredible track. Until we hit the 72 hour mark, I'm not sure this willing to bet on a track like that. If it slides NORTH,we are warmer and a less hours under the "heavy" bands. This issue comes in if it DOESN'T dig down into South Carolina- what if it's only North East Tennessee. This can't be ruled out yet.

3. It doesn't slow as much-- Very possible. Spring is notorious for slow moving upper air lows-- and I'd love to see it but this looks extreme on the EC. A faster solution could yield a nice event, let's say 6-12 inches, but not historic.

4. Snow ratio's are horrible. On the January 17th event this year. Near my place on Timberlake rd, we had close to 4 inches from about .7 liquid. So, it was close to a 5-1 ration. Normal ratio is 10-1 or so. Between warmer temps, slower spells that could mix with rain, even if we get the 2.5 liquid as modeled. Ratio's could be as low as 6-1. That would be 14-15 inches of the heaviest concrete snow you'd ever seen in your life.


Vegas odds-

1. Over 2 inches of snow 75%
2. Over 6 inches of snow 50%
3. Over 12 inches of snow 25%
4. Over 18 inches of snow 5%.

I'd love to show the vorticity  maps I get from a vendor, but I'm not willing to risk that. However, here is a qpf map from another site I found on FB. It shows one of the 6 hour snap shots of how much precipitation falls in 6 hours. MOST of VA has at least .5, places in the east exceed an inch. The model has 2 time frames like this plus another not far off.


Friday, March 1, 2013

The March 6 hype--

Social Media has made weather forecasting next to impossible. Just about everyone has an idea some of the data shows a potential big snow next week. Someone posts a map, people take it as gospel and chaos ensues.

Here is the skinny..

1. Pattern is ripe for a big storm, but has some flaws and issues.
2. We have decently cold air aloft, but the lack of High Pressure to our north to push in LOW LEVEL cold is an issue.
3. Model data spread is from suppressed in South Carolina to blowing up late and missing our region, but bringing a good event from DC north.
4. Timing is everything-- the exact strength of the blocking to our northeast and timing of the shortwaves will determine this event.

My guess at this point-


33% chance we see an accumulating snow. From that, 10% shot it's over 6 inches and 3% shot it's over a foot.

Key issue is how hard it snows. If we get 12-18 hours of light to moderate snow, we could get ..50 of liquid that only accumulates a slushy inch and is basically melted before the last flurry ends.

Trends last night-

If there were any trends last night, it was for a later phase which leaves us drier as the storm is disorganized before this point.

The key shortwave will be coming ashore later tonight over Western Canada-- that will help. I think by mid day tomorrow we should have a general idea of if this is a legit threat.

FWIW--


The HPC HAD an area of liquid in our region of between .5 and .75. After the model data last night they shifted that east and we are in a .25 to .5 for the period. Not that this is gospel, but it was a step back for our region and a step forward for places to our north and east.