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 1:15 PM ET, 12/13/2010

Analyzing pre-Christmas snow odds

By Wes Junker

A D.C. white Christmas is rare, what about this year?

Some readers may have heard rumors of a possible snowstorm this Thursday and another around December 19th from reading various weather chat forums. Others may be wondering: what are chances of a white Christmas?

I'll start off by discussing the odds of getting a white Christmas from a strictly climatological perspective (using data provided to me by Ian Livingston) and then will discuss what the models are suggesting about the upcoming period leading up to Christmas.

The odds of seeing snow any time during the week prior to Christmas (Dec 19-Dec 25 inclusive) based strictly on climatology varies depending on the amount you analyze. For the snow-starved who want any amount, snow has accumulated at least one tenth of an inch 57 times in the past 123 years (a 46% probability) between Dec 19 and 25.

However, on any given day in the week leading up to Christmas, snow fell an average of just eight times in 123 years. So there is about a 6.6% chance of seeing a few flakes on either Christmas Eve or Christmas Day (or any one of the days leading up to it) based on climatology.

The frequency of events drops off quite a bit when looking at one inch or greater accumulations. Only 25 such events have occurred during the week prior to Christmas in the past 123 years (about once every five years). The average number of one inch or better snowfalls per day during that week is only three or four giving a probability of getting an inch or more on Christmas Eve or Christmas Day (or any particular prior day) of just 2.9 percent.

Santa's reindeers want more snow than that but they know the probability of getting a snowstorm of over four inches during Christmas week is only around seven percent. Moreover, the odds of getting a such a snowstorm on either Christmas Eve or Christmas Day is only 1 in 100 based on climatology.

So a major snowstorm during the week ahead of Christmas is rare and getting one on Christmas Eve or Day is even rarer. Last year's Christmas snow cover from the December 18-19 storm was the exception not the rule.

Prior to Christmas week this year, there is a chance of light snow on Thursday as a weak low scoots by to the south. The NAM model gives us almost .25" liquid and the GFS model around 0.08" which would equate to under an inch to a couple inches of snow. The system has very little upper level energy with most of its lift being supplied by warm advection (winds from the south). The upper system will also be shearing apart and the low will be weakening as it comes east so its snowfall potential is pretty limited. Right now I prefer the less robust GFS as this is the type of system the NAM typically overdoes and I'm not impressed with the upper level support.

The December 19th threat has more potential though the probability of getting a storm still is relatively low. The models have been alternating between having a storm close enough to the coast to give us snow and one too suppressed to give us any as shown in the figure below. The first model to identify the threat was the European model. Now it has drifted away from the snowstorm solution. Two of the last three runs of the GFS have shown a storm. But the latest GFS model has trended back to a flatter look suggesting the odds are better for a miss than a hit.

Two runs of the European (Euro) and GFS models from yesterday and last night simulating the position of low pressure Dec 19-20. The Euro model shows a storm off the East Coast in yesterday's run. The GFS model shows a storm off the East Coast in last night's run. Graphics courtesy Raleigh Weather.

Two other global models are split in their take of the storm suggesting the pattern is again a fairly chaotic one making any definitive call about the storm dubious.

The crux of whether a storm develops close enough to shore to produce snow really is dependent on what happens in the steering pattern aloft. The figure to the right below shows a "spaghetti diagram" which can help us analyze the upper level pattern (at about 18,000 feet) and range of possible storm tracks.

A "spaghetti plot" showing the range of modeled steering currents in the atmosphere at about 18,000 feet (500 mb) Sunday, December 19.

The white line (amidst the green lines) with the big dip (trough) in the East is the operational GFS model from last night - a run that supported snow as it simulates a storm moving up the coast with cold air in place. Note that there are a number of green lines near it representing additional runs of the same model (with slightly different inputs) known as ensemble members. Many have less of a dip than in the operational model (i.e. the white line). These ensemble members have a weaker, flatter trough and end up with the surface low being suppressed and well out to sea. That's essentially the European solution. The green lines that are east of the white lines also are members that have the low too far east for snow.

The important thing to notice is that the differences between the white line and the other more amplified green lines and the flatter less amplified green lines are relatively small. The differences in this case between a storm and a non-storm are relatively small suggesting that the potential for snow on December 19th is higher than normal.

I think the flatter idea (i.e. no storm) is probably right because the timing and the evolution of the northern (in green) and southern jet (in red) streams needs to be almost perfect in this case to get a storm. That can happen but is sort of like threading a needle. You often miss the eye. However, the evolution of the pattern is certainly worth watching and the probability of snow on the 19th while probably only being in the 20 to 30 percent range is still above normal for the day.

The probability of getting any significant snow decreases and pretty much reverts towards climatology (i.e. average) beyond the 19th. The pattern will be colder than normal suggesting that any precipitation that occurred would have a better chance of falling as snow than normal. If it is to snow prior to Christmas this year, it will likely stick around through the holiday as cold air should be in abundant supply. But the pattern should be drier than normal as the upper flow looks like it will primarily be from the northwest. The Climate Prediction Center precipitation 8-14 day forecast ending on December 26 is for drier than normal weather.

With respect to the weekend (Dec 19) storm, we'll be watching it closely over the next several days. In this pattern, we expect the models to go through several more flips before converging on a solution. The game is far from over. I'll offer updates about the models in the comments section during the next couple of days unless another full-scale article is warranted.

By Wes Junker  | December 13, 2010; 1:15 PM ET
Categories:  Latest, Winter Storms  
Save & Share:  Send E-mail   Facebook   Twitter   Digg   Yahoo Buzz   StumbleUpon   Technorati   Google Buzz   Previous: With climate change, long-term trends are key
Next: Charting 36 wind-chilled hours


Wes, thanks 4 the update, this looks like the usuual set up 4 this area, 40% to far S, 20% perfect path & 40% to far N. In the 50+ yrs I've following the weather in this area it seems like about 20-25% of the time we get the prefect path, with last yr being the exception. Would be nice 2 get a nice snow b4 the 25th.

Posted by: VaTechBob | December 13, 2010 1:43 PM | Report abuse

Thanks for the detailed explanation, Wes.

Posted by: Rcmorgan | December 13, 2010 2:25 PM | Report abuse

Wes, thanks for this clear, concise, intoxicatingly interesting description of the ways to measure the chances of a snowstorm on the 19th! I've never looked at satellite runs or computer weather models in detail before, but have sometimes wondered what all the squiggly lines and colors signify. Your article here does a super job of describing this complicated info in a way the common man can digest! Have you ever heard of Bill Nye the Science Guy?? I think you have a skill set that would be perfect for an adult version of that show dealing specifically with weather :)

Posted by: kolya02 | December 13, 2010 2:35 PM | Report abuse

just a reminder to readers who read this. while we may or may not have a white christmas, the time to stock on stuff is a period like we have now. if you wait until it happens or after the fact, things run out or you pay higher prices. last season, i waited too late to get a snow shovel and gloves. shovels were gone and gloves were pricey. snow scrapers. check on those now. non perishable food and a manual can opener. even toilet paper will keep if you dont feel like running around in snow to get some. extra bread can be frozen. and remember a little donation to a food bank is a great way to say thanks during the holidays.

Posted by: rhitchens1 | December 13, 2010 2:45 PM | Report abuse

i'll consider this the "7-days out" forcast. anybody know what jb is saying for washington for dec 19/20? for 1) chances of snow, and 2)quantity of snow? let's catalog his prediction now and compare it to the actual results. wes seems to be saying 25% chance and doesn't give amounts.

if that 25% (20-30%) chance does materialize, what can you say about the amount of snow?

Posted by: walter-in-fallschurch | December 13, 2010 2:49 PM | Report abuse

As always, a great pleasure to read your very informative articles.

I am out in the Central Shenandoah Valley (Augusta County). Season to date, received 1 inch of snow. Areas 30-40 miles west, 12" plus.

I notice with interest the northward jog of the last two runs of the NAM and GFS, with some support from the GGEM for Thursdays possible event. A blend of the NAM/GFS gives my area 2-3 inches.

There also seems to be a shift to the west with some of the latest guidance for the weekend potential. The next several days will be interesting!

Posted by: AugustaJim | December 13, 2010 2:52 PM | Report abuse

Wes, I think a good part of the difficulty (uncertainty) with the potential storm on around the 19th is the complexity over the central Pacific of the evolution of blocking ( ) and associated circulation.

Bottom Line: 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 ( )

Posted by: SteveT-CapitalWeatherGang | December 13, 2010 3:04 PM | Report abuse

Thanks for this analysis--clear and helpful.

Also, it's interesting to me how the possibility of snow brings out the commenters. It's good to see some of the screen names from last winter's storms here again.

Posted by: mensan98th | December 13, 2010 3:16 PM | Report abuse

Last year, I got spoiled. It was my first full winter in DC (pcs'ed from AK) and I loved every second of it. I go a paid week as change off work and my power only quit once. I'm not asking for epic greatness but a nice dumping on a Sunday night/Monday morning so I can snuggle my honey under the new down comforter would be wonderful. He missed all the fun last year.

Posted by: HappyArmyWife | December 13, 2010 3:28 PM | Report abuse

Thanks for the heads up, Wes. Bread: check. Milk: check. T-paper: check. Beer: ... hold off on that snow, I'll be right back!

Posted by: blackforestcherry | December 13, 2010 3:33 PM | Report abuse

Steve, I think the Pacific is one of the problems in why the models have been having so much trouble pinning down a solution. I also think anytime you need to both amplify the pattern in both streams as well as time the various waves, it's a recipe for having lots of uncertainty.

Walter I have no idea what JB is forecasting with the event. Any forecast of amounts has to wait until we actually know whether the storm will develop and its track.

Posted by: wjunker | December 13, 2010 3:48 PM | Report abuse

you mentioned beer, and that's important, but for big snow events it's important to have something to spike the hot chocolate and/or coffee with. ;-)

obviously, it's not ususally like last year.

the normal pattern is lots of chatter about the impending storm, forcasts wavering to and fro btwn lots of snow and no snow, t.v. weather people always hyping storm chances - often beginning a broadcast with the teaser "will we see snow this week?" (the answer 20 mins later usually being "no") - and ending with a cold rain storm...

Posted by: walter-in-fallschurch | December 13, 2010 3:55 PM | Report abuse

I think Thursday snow for the NW folks (like moi) would be dreadfully unfair, as we are looking forward to Thursday as a possible day to put our hair back on, chase down and rehang the outdoor Christmas lights, and take the garbage bins out of the garage. Please weather gods, give us a break.

Posted by: tinkerbelle | December 13, 2010 4:06 PM | Report abuse

Tinkerbelle, The NAM has backed off on the snow on Thursday and now just shears that system apart. It gives DC no snow with that system. It's earlier forecast always looked suspect given the weak upper level support it had. You might luck out though the NAM at that time range is always a little suspect. I'll post something when the GSF come in.

Posted by: wjunker | December 13, 2010 4:14 PM | Report abuse


I didn't realize you were posting. Glad to see someone from Americanwx has read the blog.

HappyAmrywife, I hope sometime down the road you get your snow wish.

Posted by: wjunker | December 13, 2010 4:29 PM | Report abuse

Thanks for the responses. At least I have my Goldschlager at the ready!

Posted by: HappyArmyWife | December 13, 2010 4:42 PM | Report abuse

Now HM says the WRF model has dc to bmore at 2-4 inches of snow on thursday :/

Is this guy just model hugging or is there a big chance this will happen?.... I thought thursdays storm was going south?

Posted by: KRUZ | December 13, 2010 4:56 PM | Report abuse

Kruz, I think there is very little chance for 4 inches. I could see the system shearing too much to give us any snow more than I could see 3 or 4 inches.The most likely amount would be nothing to an inch.

Posted by: wjunker | December 13, 2010 5:15 PM | Report abuse

Wes, thanks for all the info!

Really looking forward to your updates about sundays potential storm. Really gotta plan salt ordering for my job. Id hate to place the order for specific tonnage and nothing happens :/

Stressful time of year for public works employees:)

Posted by: KRUZ | December 13, 2010 5:20 PM | Report abuse

b> Kruz,

The 18Z gfs now has shifting back west with its low and has DCA in .50 to .75 liquid equivalent. That's a pretty good snowstorm for our parts. I still wouldn't yet raise the odds past 20 to 30 percent of it happening until the euro and/or its ensembles shift back west.

the 18Z gfs still gives us a dusting of snow. a little less precip than the 12Z run but not the massive change that the NAM made.

Posted by: wjunker | December 13, 2010 5:32 PM | Report abuse

Thanks for the detailed and comprehensible analysis. Unfortunately, you forgot to mention two important factors around here regarding white Christmases in general (not to mention decent winters in general): warmer temperatures and/or subsequent warmer precipitation (i.e. rain). Except on the occasions when we get both snow *and* cold temperatures on Christmas Day itself, one or both of the above usually conspires to eliminate any snow that we may have gotten in the week up to and including Christmas.

Posted by: sburban | December 13, 2010 5:34 PM | Report abuse

Oops, I noticed I wasnt very specific. The 18Z gfs shifted back to a snowy dec 19 solution and kept light snow/a dusting for Thursday even though the nam had gone dry.

Posted by: wjunker | December 13, 2010 7:09 PM | Report abuse

Thanks for the thorough posts- I really enjoy reading your clear and interesting explanations!

On another note- I have nothing in my house to spike anything with. Forget the run on bread and milk this season- I better run to the liquor store! Regardless snow, or no snow, this cold justifies some spiked hot chocolate. There's no way I can survive the viewing of the Geminid meteor shower in these temps without it.

Posted by: Snowlover2 | December 13, 2010 7:33 PM | Report abuse

whoa... wes,
so are you now saying dec. 19 18Z gfs gets 5-8" of snow for my house! that's real snow.

i recommend bailey's and/or vodka...

Posted by: walter-in-fallschurch | December 13, 2010 8:49 PM | Report abuse


I'm on it! Got the Baily's. Phew.

Posted by: Snowlover2 | December 13, 2010 9:41 PM | Report abuse

Walter, yes taken as gospel the 18Z gfs would give the dc area 5-8 inches of snow. The 18Z gfs ensemble mean also looks pretty good in terms of its 500h pattern. Still I'd probably only boost the probability of a snowstorm up to 30% as I don't like making changes based on one model cycle.

Posted by: wjunker | December 13, 2010 10:39 PM | Report abuse

i understand your professional need for caution.... still i'll take 30% chance of 5-8". that's the best snow news i've had in 10 months...! all the while, i understand it could disappear w/the next run....

Posted by: walter-in-fallschurch | December 13, 2010 10:50 PM | Report abuse

Walter, the two GFS runs last night gave us between .25 and .50 with the drier total coming from the later run so the model has trended towards lighter amounts. The euro is a complete miss and has much less amplitude to the upper trough than the gfs. Unfortunately the euro ensemble mean has the same suppressed solution. The odds have probably edged down slightly to around 25 percent.

Posted by: wjunker | December 14, 2010 8:40 AM | Report abuse

NOOOOOOOOOOOOOOOOOOOOO! Now that that's out the way I still won't get up hope...

Posted by: Rcmorgan | December 14, 2010 9:22 AM | Report abuse

harumph... darn that euro. i choose to ignore that and go with the gfs... so now at 25% of 2-5".... gotcha. still better than a stick in yer eye.

Posted by: walter-in-fallschurch | December 14, 2010 12:34 PM | Report abuse

From what I just saw on the Euro, the weekend storm trended more west but somehow is a little farther OTS?.... Am I seeing that right?

Ensembles west as well?... Does this improve our chances much?

Seems like if these goes ENE it could still miss DC like we would be too far west of the storm as its OTS, but be huge for New York and north?....

Look forward to your latest update Wes!!!

Posted by: KRUZ | December 14, 2010 1:23 PM | Report abuse

The comments to this entry are closed.

RSS Feed
Subscribe to The Post

© 2012 The Washington Post Company