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 1:30 PM ET, 01/ 4/2011

First half of January trending cold, any snow?

By Wes Junker

Watching two winter weather opportunities

A few people have been wondering about the Capital Gang Winter Outlook and whether January will closely follow the script. The original CWG call was for December to be cold and for January temperatures to average one or two degrees above normal January.

Now, it's looking increasingly like that at least the first half of January temperatures will average below normal and the 8-14 day forecast suggests that precipitation will be normal to a little below normal. The cold temperatures do raise the probability of us getting snow during any periods of storminess. However, in most La Nina winters such as this, moisture can be limited and snowstorms usually tend to be on the light to moderate side. As for the next week, models hint that there could be a little snow Friday from a clipper and another winter weather threat early next week.


Why does the first half of January look like it will be colder than normal despite the strong La Nina conditions in the tropical Pacific?

The answer lies in the Arctic Oscillation (AO) and North Atlantic Oscillation (NAO) which have been strongly negative through December. An earlier article discussed how these two oscillations could influence the temperature forecasts during a La Nina. For those who have forgotten, when the AO (or NAO) is in its negative mode, higher than normal pressures are located across the polar regions and lower than normal across the mid-latitudes. Since air moves from areas of higher pressure to lower, cold air feeds southward toward eastern North America when the AO is negative.

The Arctic Oscillation: Observed (top panel), 7 and 10 forecasts (middle panels), and 14 day forecast (bottom). Source: NOAA.

The impact of the AO this winter has been even stronger than normal because of the magnitude of areas of blocking high pressure that have been occurring at the higher latitudes. (Read Matt Rogers' article for discussion of a theory about what's causing the blocking)

Because of this blocking pattern, the AO is forecast to remain negative through the first half of January. NOAA's Climate Prediction Center (CPC) ensemble forecasts of the AO are shown to the right. Note that though the next 10 days the index is forecast to remain negative and only towards the end of the period is it forecast to rise to a little above zero. However, if you look at the 14-day forecast versus verification (bottom panel) you can also see that the AO has generally been running lower than forecast.

The left panel below shows the how much temperatures (in Celsius, a departure of 1C equals 1.7F) average below their normal values when the AO is negative during winter. The current CPC 8-14 day temperature forecast for January 11-17 is shown on the bottom right.

Temperature departure from average during negative AO phase (left panel) and CPC 8-14 day temperature prediction (right panel).

Not surprisingly, the pattern of below normal temperatures on the CPC forecast looks very similar to the typical pattern associated with a strongly negative AO. So far this year the state of the AO is trumping the current La Nina in terms of temperatures. Until the AO weakens or changes sign, our temperature are probably going to have more below normal temperature days than warmer than normal ones.

The CPC forecast also suggests that we have below normal precipitation during the period. The models are showing very cold air building in western Canada that looks like it will start plunging into the Plains next week. The Capital Weather Gang will be watching this arctic air mass to see how and when it might eventually affect our area.

What about Friday? I agree with Matt's assessment earlier today. Both the GFS and NAM models bring a clipper to the area giving the area a period of light snow. They are uniform in saying that the strongest low initially will be to our north but that a weaker low may pass to our south and then start to deepen as it scoots off the coast. The low center at around 5000 feet passes to our north with little or no warm air advection to transport much moisture into the region. Such storms typically produce a dusting to an inch at best. I'd place the odds of getting an inch at around 20% as the bulk of the 09Z SREF ensemble members give us no accumulating snow.

This morning's GFS and Canadian models have a storm early next week but they differ in the details. The GFS takes a low into the Ohio Valley by 7 a.m. Tuesday and then reforms it over eastern North Carolina by afternoon putting the D.C. area in that battle zone where precipitation type is up in the air. The Canadian model has the low going up towards the Great Lakes which would be a track favoring a rain storm. The GFS ensemble members are all over the place suggesting that either solution is possible or that the much weaker low could scoot harmlessly out to sea. The European model suppresses the storm early in the week to the south and southeast but holds back a second disturbance that later produces a little snow.

This pattern remains one in which the models are likely to struggle. The spread of the ensembles also suggests there will be little or no skill in model forecasts of any potential storm next week until we get much closer to any possible event. Later in the week, the CWG will take a closer look at the potential of any storm that might be on the horizon for next week and I'll offer my best guess for how the storm might evolve.

By Wes Junker  | January 4, 2011; 1:30 PM ET
Categories:  Latest, Winter Storms  
Save & Share:  Send E-mail   Facebook   Twitter   Digg   Yahoo Buzz   StumbleUpon   Technorati   Google Buzz   Previous: Coast to coast: Video of Dec. 26 beach blizzard
Next: PM Update: A dry winter chill


Thanks for the details, Wes!

Posted by: Rcmorgan | January 4, 2011 1:44 PM | Report abuse

Thanks, Wes! You answered my question to Matt about the lack of an expected January warm-up.

Posted by: ThinkSpring | January 4, 2011 1:52 PM | Report abuse

Sorry, I answered it late on my other posting. This is the 2nd winter in a row where Arctic blocking has trumped Tropical Pacific influences. Amazing.

Posted by: MattRogers1 | January 4, 2011 1:57 PM | Report abuse

so, basically, it may or may not snow? i'm just kidding. so far, though, i'm surprised it's been so cold. after last summer's heat, i thought we'd never see snow again!

Posted by: swishjobs | January 4, 2011 2:37 PM | Report abuse

I was out and about today and I noticed flocks of birds hanging low or on the ground which is usaully a sign of falling (snow?) weather. Keeping my fingers crossed! From a snow lover in Carroll County.

Posted by: jimmygre | January 4, 2011 2:39 PM | Report abuse

I am predicting sunny Cold weather broken by occasional cloudiness. AKA January

Posted by: pvogel88 | January 4, 2011 2:40 PM | Report abuse

I'm getting an Object Not Found error on the 09Z SREF ensemble members link...

Posted by: spgass1 | January 4, 2011 2:56 PM | Report abuse

Wes, very helpful. Thank you. Unfortunately, it sounds like we're only getting 50% of what we need for the snow lovers. We have continued cold arctic air but with drier than normal precip...ugh! Although, the models for Tues seem to be trending against an OTS pattern which we started getting used to (frustratingly!).

Posted by: parksndc | January 4, 2011 2:59 PM | Report abuse


Just add an L to the end of the url so that it says .html not .htm

Posted by: KRUZ | January 4, 2011 3:05 PM | Report abuse

wes, model experts, statisticians,
i know this storm is next week and the odds of it coming to fruition are small, but does anyone know how small? 1%? 5%? 20%? has there been any kind of "post stormum" analysis of how accurate the GFS is for DCA 192 or 168 hours out?

Posted by: walter-in-fallschurch | January 4, 2011 3:40 PM | Report abuse

Kruz thanks for answering spgass's question concerning the link.

Walter, it's hard to assess the probabilities when we don't even know whether there will be a storm or not. What happens with the vortex over the great lakes will play a big role in what ultimately happens. I'm not comfortable guessing. In this pattern, 168 hrs is way out there. There are verification studies but most have dealt with the skill on a hemispheric level or for storm tracks but the latter is for much shorter time range. Steve Tracton might be able to post something in more detail about the decay in model skill as you go out in time.

Posted by: wjunker | January 4, 2011 4:43 PM | Report abuse

thanks for the reply wes. i hope to hear from steve too - and any other people who have relevant expertise.

ok, so when (i.e., how many days/hours out) can you start issuing probabilities for the gfs? for the euro?

Posted by: walter-in-fallschurch | January 4, 2011 5:31 PM | Report abuse

It will certainly be several more days. The models are not even close to each other into terms of solutions. The 18Z gfs still has one low going to the oh valley and a second one forming off the carolinas. It would certainly give us some winter weather but the 850 low still is to our northwest which usually isn't good for staying snow. However, the euro looked much different.

Posted by: wjunker | January 4, 2011 5:41 PM | Report abuse

Some of the DC area snow lovers might be in severe flake deprivation status.
Let's all grasp at the long range forecast model straws together. A CWG "2" SPI is so much better than zero.

Posted by: FIREDRAGON47 | January 4, 2011 6:08 PM | Report abuse

Throwing caution to the icy wind, the National Weather Service forecasts a 30% chance of snow here Monday night and a 40% chance on Tuesday.

Posted by: craighowell1 | January 4, 2011 6:52 PM | Report abuse

that "coast to coast" video by jason was fascinating.

this spanned 6 days. is next week's "storm" identifiable yet?

Posted by: walter-in-fallschurch | January 4, 2011 8:45 PM | Report abuse


I'll be getting to the issue of verification as soon as I have a chance to finish a post on that subject.

A standard measure of model skill, the Anomaly Correlation (AC) score, evaluated by NCEP (program I initially wrote) for several models indicates that, for example, the average 500 mb height score over the Northern hemisphere (~18,000 ft) for December 2010 looses skill at the scale of systems relevant here between 4 and 5 days.


In the figure, the scale of interest are zonal waves 10-20 (~800 mi wavelength). Smaller waves and scores over sub hemisphere regions, such as North America (not shown) , are distinctly less predictable (on average)

Posted by: ensemblemean | January 4, 2011 9:07 PM | Report abuse


Do you mind not posting your comments in bold?? Bolding comments should be reserved for the CWG only.

Posted by: BobMiller2 | January 4, 2011 9:22 PM | Report abuse

Walter -

You coined the name "no-mageddon", right? You better start thinking of a name for the next storm! :)

Posted by: BobMiller2 | January 4, 2011 9:28 PM | Report abuse

capital weather gang- You have to admit that according to the models, next week's storm is DC's best chance yet this winter for accumulating snow. I think next week will be our biggest storm of winter followed by the coldest air of the winter. DC the worst of winter will come in the next two weeks.

Posted by: ajmupitt | January 4, 2011 9:36 PM | Report abuse

excellent, steve - can't wait. "anomaly correlation score". now that's what i'm talking about! very informative graphs. first thing: it seems to me like the "EC" is always better than all the others... what's up with that? am i understanding the graph right?

i believe "ensemble mean" is steve's alter ego or something.

i did think up no-mageddon - in my little world. but apparently, unbeknownst to me, it appeared in the print version of the paper that same day. it was just me wallowing in disappointment after our non-event of a storm. i'm philosophically opposed to all the over-naming of storms.... but, if there must be a name it should be descriptive like "the jan 12 2011 falls church bombogenesis"...

earlier matt referred to it as a "mother lode" of cold air. me likey.

Posted by: walter-in-fallschurch | January 4, 2011 10:12 PM | Report abuse

Steve, Thanks for the response. I'm looking forward to your article.

BobMiller2, Ensemblemean is one of the Capital Gang. He's our model and ensemble expert.

Posted by: wjunker | January 4, 2011 10:28 PM | Report abuse


I think ensemblemean IS a member of the CWG... based on various things I have seen him post in the past (though I think previously it was under ensembleman.)

(CWG, please correct me if I'm wrong.)

Posted by: MKoehl | January 4, 2011 10:31 PM | Report abuse

You were too fast for me, Wes.. :)

Posted by: MKoehl | January 4, 2011 10:32 PM | Report abuse

Thanks KRUZ for the fix.

00z NAM doesn't look so good for Friday snow... GFS showing some light precip... close call still I guess?

Posted by: spgass1 | January 4, 2011 11:30 PM | Report abuse

Apparently I have two user ID's for the Washington Post: "ensemblemean" and "SteveT-CapitalWeatherGang". The first is an older ID when I first registered for the WP years ago, the latter for CWG. I haven't yet figured out why on occasion commenting on CWG grabs onto the ensemblemean ID (cookie issue??).

Bottom line: ensemblemean is really ME, SteveT

BTW, I forgot to mention above that an AC score less than ~.6 is considered to have no skill.

Posted by: ensemblemean | January 4, 2011 11:50 PM | Report abuse

@SteveT--so all this time I've been reading your other moniker as "ensembleMAN" I've was misreading it?? Amazing how the eye & brain can mis-connect!

Posted by: petworthlad | January 5, 2011 12:19 AM | Report abuse

The comments to this entry are closed.

RSS Feed
Subscribe to The Post

© 2012 The Washington Post Company