Network News

X My Profile
View More Activity
The new Washington
Post Weather website
Jump to CWG's
Latest Full Forecast
Outside now? Radar, temps
and more: Weather Wall
Follow us on Twitter (@capitalweather) and become a fan on Facebook
Posted at 8:00 PM ET, 12/16/2010

Sunday snow threat is back, sort of...

By Wes Junker

The Snow Lover's Crystal Ball has been murky the last couple of days concerning the possible snowstorm this weekend. The threat seemed to be waning yesterday as computer models seemed to be approaching a no-storm consensus except for several pesky ensemble members (subsets of these models with slightly varied data inputs) that kept screaming "not so fast."

This morning's model runs trended back toward a storm suggesting that there is more potential than I may have thought when writing yesterday's post. And so, the CWG has dusted off the old crystal ball which still is a little foggy but providing glimpses of a potential snowstorm that might be bigger than today's if everything comes together perfectly. A very big IF.

Next accumulating snow chance: Saturday night into Sunday
Probability of accumulating snow (1" or more): 30%
Maximum accumulation potential: 6"+

Keep reading for technical discussion...


Let me preface this by stating this storm still will most likely skirt us to the east. However, a trend back toward a stormy solution began with last night's run of both the GFS and European (Euro) high resolution models. Then this morning's (12z) Euro model showed a major storm for the East Coast. The GFS run from this morning also had the makings of a major storm but it developed the storm a little later, only simulating a dusting for the D.C. metro region but a bigger storm for the Eastern shore of Md/De/Va.

Twenty-two runs of this same GFS model with only the initial condition changes yielded a wide range of solutions attesting to the fact that the final answers about this storm may not come for another couple of runs. Nine of the 22 members gave DCA some snow. For those interested, you can look at the members yourself. The precipitation is depicted in mm and for those who have forgotten the conversion about 25 mm equal an inch.

Then along comes this afternoon's (18z) model runs (of the GFS and the NAM, another model) and they have shifted back east enough to keep the snow out of DC continuing the yo-yoing of behavior of the models trying to forecast this potential storm. The SREF ensemble (another model with "members" that vary based on initial input) storm track from late this morning (15z) also strongly favored the out-to-sea solution. Welcome to roll-the-dice meteorology.

The maps below can give you a feeling of how small differences in the pattern can produce significant difference in track and development of a storm for the Washington and Baltimore area. These maps were taken from the Penn State e-wall website.

72-hour forecasts of the flow and vorticity (spin in the atmosphere) at 18,000 feet (500 mb) on the left and surface pressures on the right, valid at 7:00 a.m. Sunday morning. The top panel is the Euro model and the bottom panel is the GFS model. Both models are from this morning.

Let's look first at the differences between the Euro (top left) and GFS (bottom left) models at 500 mb (around 18000 ft). Note there are two dashed lines on the bottom 500 mb forecast. On the GFS these two shortwaves in atmosphere are still separate and haven't yet phased. However, on the European model they do phase, and instead of a weaker surface low a little farther off the coast (GFS bottom right), the European model has stronger low right on the coast (top right). The tilt of the trough on both models favors strong upper level divergence (good for precipitation) and is very favorable for rapid development of the surface low.

Both models develop a major storm/blizzard for parts of New England during the next 24 hours (as shown below) but exactly where differs (see top left Euro and bottom left GFS). Yet another, the Canadian, looks fairly similar to the GFS. The UKMET, another good medium range model, is a little farther out to sea and is slower with the development. Bottom line: the models still have not come to a consensus about the storm.

wes-1219-2.jpg72-hour forecasts of the flow and vorticity (spin in the atmosphere) at 18,000 feet (500 mb) on the left and surface pressures valid at 7:00 a.m. Monday morning on the right. The top panel is the Euro model and the bottom panel is the GFS model. Both models are from this morning.

From a forecasters perspective this storm is a huge headache as relatively small changes in the timing, sharpness, and phasing of the troughs (dips in the jet stream or flow 500 mb, 18,000 feet) can lead to anything from no snow in Washington to a six-inch plus storm.

Overall, the threat of snow has risen a little since last night's runs and the probability of getting over an inch is probably around 30 percent. It's hard to gauge the probability of significant snowstorm (greater than 4 inches) comparable to that shown by the Euro this morning since this type of system should have a pretty strong precipitation gradient (i.e. a small distance between a little and a lot of snow).

Right now, I'll put the odds of getting a "big one" somewhere around 10 percent, a fairly conservative call until we get more confirmation that the models are honing in on the true solution. While low, those numbers are still way above the probabilities based strictly on climatology. Such a range also seems consistent with the ensembles.

D.C. still needs the two jet streams to mesh and phase correctly for us to get a major snowstorm, otherwise we receive a glancing blow or a miss. It's the old thread the needle analogy, but the eyehole may have gotten a little bigger since yesterday's discussion.

For southeast Virginia, the eastern shore of Maryland and Delaware, Long Island and New England, the threat is higher for a significant, possibly crippling storm but even there, lots could go wrong. Luckily for me, I don't have to make a guess at the probabilities that far from home.

By Wes Junker  | December 16, 2010; 8:00 PM ET
Categories:  Latest, Snow Lover's Crystal Ball, Winter Storms  
Save & Share:  Send E-mail   Facebook   Twitter   Digg   Yahoo Buzz   StumbleUpon   Technorati   Google Buzz   Previous: D.C. gets first inch; school tomorrow?
Next: Forecast: Continued cold. Any more snow?


Dancing, for what its worth...

Posted by: JDK4 | December 16, 2010 8:10 PM | Report abuse

Oh no, say it ain't snow!!!

Posted by: antztaylor | December 16, 2010 8:14 PM | Report abuse

You rock, Wes. I'm still sort of rooting against you though....

Posted by: Ian-CapitalWeatherGang | December 16, 2010 8:18 PM | Report abuse

Wes is going to be on an internet radio program discussing the storm tonight with other talented meteorologists at 9:30 p.m. Listen in at this link:

Posted by: Jason-CapitalWeatherGang | December 16, 2010 8:20 PM | Report abuse

Is Wes new? The name does not look familiar. Either way, sweet write up!

Posted by: PoorTeacher | December 16, 2010 8:31 PM | Report abuse


Yes-- Wes is our new winter weather expert. He's been posting since November. Read about him here:

Posted by: Jason-CapitalWeatherGang | December 16, 2010 8:32 PM | Report abuse

dancing here in baltimore!

Posted by: bachaney | December 16, 2010 8:37 PM | Report abuse

I'm glued to my seat with all this fluctuation....Thanks Wes! :)

Posted by: Rcmorgan | December 16, 2010 8:46 PM | Report abuse

I'm dancing for NO snow...

Posted by: Murre | December 16, 2010 8:47 PM | Report abuse

Trying to fly out to Hartford on Sunday, should I be looking to make other plans? Even if it misses us, it looks like a sure shot to hit New England hard?

Posted by: ianswank | December 16, 2010 8:48 PM | Report abuse

Great post. I love that Wes is now on CWG.

Posted by: IanK | December 16, 2010 8:54 PM | Report abuse

i'm trying to follow along with all your drawings and so forth, but... i can't see those red and purple contour-type maps. i can't read the text and i can't tell where my front yard (MFY) is.... can you link to larger drawings somehow?

a poster mentioned how the "greenland block" is in place again and how that's good for the long-term trend. true? i sure love having all this cold air around. all moisture events are snow "threats".

1.29" here. i imagine how these totals will drop quickly from compaction and transpiration. very dry snow.

Posted by: walter-in-fallschurch | December 16, 2010 8:55 PM | Report abuse

Wes, what's that Euro saying now? I promise I'll still listen at 9:30.

Posted by: dustinmfox1 | December 16, 2010 9:02 PM | Report abuse

Down here in Williamsburg, a little disappointed with our measly inch or so. Never really got going before changing to sleet and rain. But...pretty excited about the chance for a much bigger storm this weekend.

A good (but not CWG good) geeky Hampton Roads weather site worth looking at is which is the local NBC station's weather blog page.

Posted by: orduckfan | December 16, 2010 9:06 PM | Report abuse

2" in NW Spotsy. 2day, would like 2 c at least double that on Sun. Agree with a 10% chance of a big storm at this point. Difference between no snow or very little & 6+ snow can b 75 miles. 2morrow evenings runs should give us a much better idea.
Does any1 reading this site cross the 301 Bridge, would like 2 know if the MD side is iced over. Would like 2 fish the power plant discharge 4 stripers but afraid it might now b iced over at the Aqua land ramp. Any info would b appreciated.

Posted by: VaTechBob | December 16, 2010 9:23 PM | Report abuse

The problem with Wes' explanations being so well done, is they pop my snow bubbles....

I'm also turning into a model junkie. I thought I had this problem licked last year.

Posted by: Snowlover2 | December 16, 2010 9:28 PM | Report abuse

It looks as if my forecast in a comment Monday afternoon will prove to be the only one to hit the mark: namely,

It's unlikely that models will begin to converge more than a couple days in advance, i.e., not likely too be one of the more predictable storms ( )

In fact, I'd bet we can't say with high confidence that there will or will not be a significant storm for the DC Metro area before tomorrow (Fri) afternoon or Saturday AM (and perhaps maybe not even then).

The questions Wes discussed, especially the phasing and timing issues, are the most difficult to get right in models - extremely susceptible to the chaos introduced by those pesky butterflies.

Posted by: ensemblemean | December 16, 2010 9:35 PM | Report abuse

For you model followers out there... The 0Z NAM is in and shows the weekend storm center west of its previous run. Which is good for the pro snow camp... need that westward trend to continue though to get the precip shield that Wes talked about to make it into DC.

Posted by: pseaby | December 16, 2010 9:47 PM | Report abuse

Suddenly being cold is so much more interesting.

Posted by: FIREDRAGON47 | December 16, 2010 9:56 PM | Report abuse

I'm not impressed. All this supposed technological information, and yet you guys don't know zip; I'm old enough to remember Louie Allen on the old Channel 7 News program, who used to do his weather analysis on a "woodle"board (newspaper tablets), yet seemed to know at least as much as you guys do with all your computer models. You guys are techo-addicts who can't improve on old Louie's "woodle" prognostications.

Sorry for the rant.

Posted by: MillPond2 | December 16, 2010 10:10 PM | Report abuse

Sarcasm Millpond? These are the best forecasters around.

We need sometime up visual for an indicator... something like a fuel gauge. But instead of "full" and "empty" it should say "Western trend" and "out to sea." The line could move after each model comes in.

Posted by: PoorTeacher | December 16, 2010 10:14 PM | Report abuse

CWG said 1-2" for this area. i got 1 1/4". they were right on.

Posted by: walter-in-fallschurch | December 16, 2010 10:16 PM | Report abuse

I'm doing the "no snow" dance. Our work is entirely dependent on the ability to work outside, and we have a big, new job kicking off Monday. Snow would mean no-go until the beginning of the year. :\ I'm hoping that all the snow falls somewhere it can't accumulate, namely, the ocean.

Posted by: marklandterrapins | December 16, 2010 10:22 PM | Report abuse

Thank you CWG for today's accurate forecast.
It's nice not to be surprised by these sort of weather events. I was glad I brought my snow boots to work with me today.

Posted by: FIREDRAGON47 | December 16, 2010 10:23 PM | Report abuse

This is exactaticaly (props to Sonny Eliot) what I was hoping to see with the evolution of CWG. I am being educated as well as entertained and informed. Good work Wes and CWG.

Now, where is that snow shovel...

Posted by: goldiesc | December 16, 2010 10:28 PM | Report abuse

I'm sorry post prognosticators, but I am referring to Sunday's forecast. By the way, Louie Allen was as accurate as the CWG.

Posted by: MillPond2 | December 16, 2010 10:33 PM | Report abuse

Thanks for the link, Jason. I've enjoyed the commentary by Wes and the others...

Posted by: Rcmorgan | December 16, 2010 10:35 PM | Report abuse

I remember Louie Allen, & 2 call his 4casts accurate is like calling water dry. Still remember his snow flurry 4cast that turned into 8-10". Louie was no better or any worse than any of the weathermen of that era. He was entertaining 2 watch, but far from prefect in his predictions.

Posted by: VaTechBob | December 16, 2010 10:37 PM | Report abuse

ianswank, it's not a sure thing that New England will get snow but their odds are much higher than ours and the potential for crippling snowstorm is higher.

Millpond, I grew up with Louis Allen, he was my idol. That said, forecasts are much better now. verification scores for precipitation have improved steadily over the past 30 years.

Posted by: wjunker | December 16, 2010 10:44 PM | Report abuse

MillPond: The beauty of CWG, and that webcast was no exception, is that they explain the whys and wherefores of forecasting. Not only do CWG readers come away from each blog post with a forecast, we also come away with a dose of science education. And when the forecast is hard to call, we get a good explanation of what the variables and uncertainties are to understand just why it's hard to call.

If you just want a quick forecast that might or might not pan out, you're probably in the wrong place. The point of the webcast tonight wasn't about what they KNOW in terms of giving us a firm prediction; it was about why it's being so hard to do that. If you're not interested in that kind of discussion, CWG may not be for you.

Posted by: mhardy1 | December 16, 2010 10:49 PM | Report abuse

I also grew up watching Louie Allen. And he was my idol too. Perhaps my biggest idols are the meteorologists Col Krick and LTC Holtzman who convinced Eisenhower that the weather for june 6, 1944 was benevolent enough, and had a large enough window, to land the biggest invasion force in history on the coast of France.

Posted by: MillPond2 | December 16, 2010 10:58 PM | Report abuse

The 00Z GFS came in with a slightly moree amplified solution putting dc in an inch or two of snow. Still lots of uncertainty. I really struggled with the 1" or greater probabilities in the article and at one time had them as high as 50% which I guess how uch uncertainty I feel about the pattern. Stay tuned tomorrow for another assessment.

Posted by: wjunker | December 16, 2010 11:11 PM | Report abuse

Latest GFS run looks to be trending more to the west to that correct?

Posted by: oriolesfan2323 | December 16, 2010 11:11 PM | Report abuse

Thanks for all the detail, CWG has become the goto site for following weather around the office, especially among the former aerographer's mates and other meteorology nerds. Thanks for leaving us in the know!

Posted by: miglewis | December 16, 2010 11:20 PM | Report abuse

I have to respond to mhardy1: Criticism of CWG is not a reason to assume that I'm not interested in being "educated". I have lived my entire life in the DC area. I am well aware of how difficult it is to accurately predict snowfall totals, as well as snowfall possibility. However, sometimesI feel that computer modeling is a crutch, as well as a fallback, for local meteorologists when predictions don't match reality.

That doesn't mean that I don't appreciate the conundrum they face in trying to accurately predict the probability of snowfall in an area that lies along the Atlantic coastline, where the difference in accumulation or simple rainfall can be measured in the range of ten or fifteen miles.

However, I still don't see a great improvement in what Louie Allen was doing fifty years ago versus what occurs now with computer modeling. Ten years ago, i remember Sue Palka admitting on the ten o'clock Channel Five news that local meteorologists had "blown it" and that a major snow event was headed for the DC area.

That's the price meteorologists pay for trying to predict DC weather.

This little storm was fairly easy to predict because it was not a coastal storm. it was more of the Alberta Clipper variety, it ran into a stable, cold mass of air, which dried it out, and it did not have the gulf coast track which would have provided a lot more moisture. The prediction of 1-2 inches was, relatively, a no brainer.

Posted by: MillPond2 | December 16, 2010 11:45 PM | Report abuse

VaTechBob: Agree with you about Louis Allen. He was a fine artist and very charismatic and frequently used the term, "the ducks are on the pond." However, I too, remember some of his big forecast "turkeys," such as the one MillPond2 refers to.

Posted by: Don-Capital Weather Gang | December 16, 2010 11:48 PM | Report abuse

MillPond: Thanks for the response; I understand better where you were coming from now, and my criticism was off-base, so sorry about that.

Your earlier comment sounded to me like a complaint that you were listening to the webcast and not getting a clear and simple prediction. I assumed you were hoping for a confident forecast rather than a discussion of the reasons for the ambiguous and uncertain forecast, and so I misconstrued you.

Posted by: mhardy1 | December 17, 2010 12:16 AM | Report abuse

I also fondly remember Louie Allen. I remember the time he forecast a mammoth snowstorm and it was sunny the next day! His excuse was that he saw a bright light coming at us and thought it was a giant truck but it turned out to be two motorcycles!

As for this weekend's storm, it seems to me that during strong LaNina's, with the relatively weak southern stream, snowstorms don't amplify enough until they pass north and east of DC, giving us a dusting and then clobbering New England!

Posted by: buzzburek | December 17, 2010 1:19 AM | Report abuse

Buzz, I remember that incident too. It's funny that you Andy and I were talking about Louis Allen at the 19th hole. I was always impressed with his artistic ability.

Last night/s Euro completely caved the only high resolution models that gives us any snow is now the last two GFS runs which give us about an inch. The probabilities in the piece still look OK though I think the probability of big storm may have ebbed a little.

Posted by: wjunker | December 17, 2010 8:23 AM | Report abuse

Uh. . . .we're going to be driving back from Williamsburg to the DC area on Sunday. Are we hosed?

Posted by: BadMommy1 | December 17, 2010 9:20 AM | Report abuse

Posted by: wjunker | December 17, 2010 10:13 AM | Report abuse

Wes: a truly great write-up. It took me an hour to arrive at the same general conclusions you articuled so well, here. Snowlover: I too, promised myself I'd be less of a model freak this year, but clearly I'm headed toward rehab again. I love it when the weather comments get more technical like this, and steer away from "what's wrong with DC drivers/local school systems/politicos" etc... Today starts the really good stuff - and however this one turns out, I do enjoy a good weather soap opera.

Posted by: curtmccormick | December 17, 2010 10:36 AM | Report abuse

Being a SERIOUS snow-lover, I always hold on to the last shred of hope until the end. What happened in January of 2000? I had started a new job and was extremely busy and mentally exhausted in the evenings all week and went to bed without ever watching the new or listening to the radio. I remember waking up and thinking that I saw slivers of light when it should have still been pitch dark. I opened the window and SHOCKED when I saw several inches of snow.

So maybe, just maybe we could still get some on Sunday ... ?

By the way, does anyone remember when that snowstorm was forecast? I'm really thinking I didn't hear anyone mention it at work that day, that's why I was totally shocked.

Posted by: luvhh1 | December 17, 2010 10:57 AM | Report abuse

Posted by: wjunker | December 17, 2010 11:03 AM | Report abuse

looks like wes is trying to tell us something (@10:13 and 11:03). perhaps the 12" predicted by the latest model run has left him speechless?

Posted by: walter-in-fallschurch | December 17, 2010 11:13 AM | Report abuse

Walter, I was trying to answer badmom to say that even in Williamsburg, it looked like it might OK for her. She needs to still keep abreast of the forecast but the models have shifted east.

Posted by: wjunker | December 17, 2010 11:42 AM | Report abuse

Thank you, wjunker! I appreciate the input. Of course, I'd risk permanently alienating my family if I suggested bagging the weekend. . . .but the words "crippling snow" had me there for a minute.

Here's hoping that the whole thing goes off smoothly!

Posted by: BadMommy1 | December 17, 2010 12:06 PM | Report abuse

I remember "Krick Weather" growing up in the Upper Midwest...

It seems that Col. Krick relied rather heavily on the use of analog weather patterns from previous years when issuing his forecasts rather than relying on the use of models and model ensembles as we do nowadays. During the 1960's computerized weather forecasting was in its infancy, although Cray, Inc., headquartered in Chippewa Falls, WI, was beginning to assemble the powerful supercomputer servers capable of performing forecast calculations applicable to modeling atmospheric conditions. This has been the prime reason why meteorology students, have had to be proficient in math "a year beyond calculus". Atmospheric modeling demands considerable knowledge of mathematics, including advanced trigonometry, three-dimensional vector algebra, and an exhaustive knowledge of differential-equation calculus. Before the advent of Cray supercomputers, it would simply take too long to perform manually the mathematical calculations necessary to compose an accurate atmospheric model. Even today, the absence of data at crucial points can result in an inaccurate "blown" forecast, one major reason why we occasionally get unexpected snow or ice storms around here.

Posted by: Bombo47jea | December 17, 2010 2:05 PM | Report abuse

The comments to this entry are closed.

RSS Feed
Subscribe to The Post

© 2012 The Washington Post Company