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 11:25 AM ET, 11/ 2/2010

Why was last year so snowy? Part II

By Wes Junker

Winter outlook preview

* Why was last year snowy? Part I | Varying predictability of snow *

(Please welcome Wes Junker, who will serve as Capital Weather Gang's winter weather expert moving forward. Read Wes's full bio. This is the second in a series of articles leading up to our 2010-2011 winter outlook, which will be published Thursday)

By Wes Junker, Winter Weather Expert

cwg_Junker.jpgYesterday, I discussed how El Nino played a key role in establishing a storm track and moisture feed for generating last winter's historic snows in the mid-Atlantic. But I concluded by emphasizing El Nino did not lead to the epic snow totals by itself.

The help from the north last year came from the Arctic Oscillation (AO) and its cousin the North Atlantic Oscillation (NAO). Both these fluctuations have two phases and they are predicated by the differences in the pressure between the polar vortex and the pressures farther south in the mid-latitudes.


When the two indexes are positive and the pressures at high latitudes are lower than those at mid-latitudes, the increased pressure differential between the polar region and the mid-latitudes leads to an increase in the strength of the low level winds that blow from west to east in the mid-latitudes. This causes weather systems to typically move faster. Also, having higher pressure to the south of a high latitude low keeps a southerly component to the wind which prevents Arctic air from being swept southward into the U.S. since air flows from higher to lower pressure.

Importantly, note that a positive AO and NAO index is usually associated with above normal temps across much of the U.S. including the mid-Atlantic States (see below, the warm colors on the bottom two panels indicate areas where the surface temperatures are usually above normal with a positive AO and NAO)

A positive Arctic Oscillation (AO) and/or North Atlantic Oscillation (NAO) is correlated with warmer than average temperatures over much of the eastern half of the U.S.


The opposite is true when the two oscillations are in the negative phase. Pressures then are above normal across the higher latitudes and lower to the south which leads to the winds having a northern component more often than when the opposite is true. Therefore, more Arctic air masses enter the U.S. and temperatures tend to be colder than normal during the negative phase.

The colder air over the Southeast and mid-Atlantic regions leads to a stronger than normal temperature difference between the warmer air over the oceans and the colder air over the land. The anomalously strong temperature differences (gradient) along the coast helps fuel more frequent "Nor'easters".

Last winter, the AO and was more than 4 standard deviations below normal, the NAO was more than 1 standard deviation from normal indicating that the negative phases of each oscillation were unusually strong.

The colder than normal air across the mid-Atlantic States and Ohio Valley and the increased thermal gradient along the coast are two reasons why negative NAO winters tend to be snowier than periods when the oscillation is in its opposite phase. A negative NAO also increases the chances of getting a major snowstorm.

Paul Kocin and Louis Uccellini in their book, Northeast Snowstorms Volume 1, investigated thirty snowstorms that produced at least 10 inches of snow. Twenty-two of the thirty storms occurred with a negative NAO, only 5 were associated with a positive index. They also found a negative correlation between the NAO index and the seasonal snow across the Northeast (including Washington and Baltimore).

Another study done at Rutgers University indicates that there is a 55% increase in snowfall at Reagan National Airport (DCA) when the NAO is negative versus when it is positive. However, the increase in snowfall is 347% when the NAO is strongly negative versus strongly positive (see table below for other locations). Last year both the NAO and AO were extremely negative during the snowstorms (see below)

NEGNAO-snowfall surplus.jpg
Source: Rutgers University

negnao-snowfall surplus2.jpg
Strongly negative values of the NAO coincided with last winter's (2009-2010) snowstorms.

The December 19th and Feb 5th snowstorms are great examples of how El Nino and a negative NAO can act synergistically to produce a major snowstorm. Last winter, both storms leaped onto the list of Washington's top 10 heaviest snowstorms. The figure below compares the December snowstorm forecast pattern (on left) to the February storm (on right). Note that the above normal pressure (also known as heights, warm colors) are at high latitudes on both with a high in the vicinity of Greenland. Below normal pressures (or heights) are found to the south with a low near or just off the Canadian Maritimes. The configuration is a classic negative AO and NAO.

Forecasts of pressure patterns at around 18,000 feet (500 hPa) for the December 19 and February 3 snowstorms.

The high pressure system around Greenland (closed upper level anticyclone) acts as a block that helps hold the intense low near the Canadian Maritimes in place. Note how the lines over New England on both forecasts become closer together as you move from west to east. Meteorologists call such a look confluent flow which is usually an area where air molecules are converging which causes the weight of the atmosphere above the earth's surface to increase. This causes the surface pressures across the northeast to rise and stay higher than the pressures farther south with the low.

When the surface pressure is high to our north and northwest, cold air continues to feed into the area as the surface low is forced to our south. The confluence across New England is one mechanism for keeping the cold air damming across the Metro area making it more likely to be an all snow event rather than one that mixes or changes to rain. A perfect track taking the storm across the south to near Cape Hatteras also was optimal for keeping cold air in place during both storms.

Both maps also show another area of below normal heights across the south. That feature is more likely to occur with an El Nino than during other years. That feature is our southern stream storm that gives us moisture and brings a surface low towards the east coast while the negative NAO helps provide adequate cold air for snow.

The February 1983 blizzard, another top 10 snowstorm during another El Nino winter, had the same basic NAO look with a distinct southern stream wave.

Random luck comes into play in a couple of ways. While it's not unusual to have a season where the NAO and AO are negative for much of the winter, much of the 1960's featured a predominantly negative oscillation, having both indexes so strongly negative for such long stretches is rare and the reasons why may be due to random chance or serendipity. The Arctic oscillation reached record negative levels last winter. Why these levels were so negative is anyone's guess.

Also, while the phases of the southern and northern oscillations favored above normal snow, there have been other years when both were in the same phases that produced significantly less snow than last winter. How much snow is produced by an individual storm is partly governed by luck. Both these storms were associated with impressive plumes of moisture that came from very low latitudes making them extremely wet.

Later in February and March, the heavy snow was confined to our north because the timing and the location of the various players were a little out of phase for the DC area to get a big storm. During most years, that is the norm. The timing of the system in December and again in early February were perfect for getting major snowstorms last winter. We were lucky, or unlucky depending on your point of view about snow.

By Wes Junker  | November 2, 2010; 11:25 AM ET
Categories:  Capital Weather Gang, Snowmageddon, Winter Storms  
Save & Share:  Send E-mail   Facebook   Twitter   Digg   Yahoo Buzz   StumbleUpon   Technorati   Google Buzz   Previous: Election 2010: Where will the weather matter?
Next: Tropical storm Tomas regroups, eyes Haiti


(Long time reader, first time commenter. Love the site and the discussions)....

So what's the NAO outlook for this year?

Posted by: josh28 | November 2, 2010 11:48 AM | Report abuse


Thanks for the comment. We're going to talk about this in our winter outlook and Wes can comment more on this when he's online, but NAO is very tough to predict. That's always the big wild card in seasonal outlooks for our region.

Posted by: Jason-CapitalWeatherGang | November 2, 2010 12:17 PM | Report abuse

Josh28, the NAO is tough to predict more than a few weeks in advance during most winters. Unlike last year when there were quite a few factors pointing towards a negative NAO leading up to winter, this year the signals are more mixed as to what might be the predominant phase. I'm more a short to medium range forecaster than one with much skill at seasonal forecasts. The CWG winter forecast will be issued later this week by Matt Ross. His accompanying discussion will address what the NAO might bode for this winter's snowfall.

Posted by: wjunker | November 2, 2010 12:21 PM | Report abuse

Thanks to you both! I missed all the big storms last year, so I'm crossing fingers for an opportunity to catch up - at least once!

Posted by: josh28 | November 2, 2010 12:35 PM | Report abuse

Since this year is a moderate/strong La Nina we're not quite so likely to get the huge snowstorms we had last winter...though the NAO/AO configuration looks rather similar to last year at present, and the Greenland block seems to be hanging around, at least for the near term.

My hunch...we'll see more of the changeable "wintry-mix" type systems...

Posted by: Bombo47jea | November 2, 2010 12:42 PM | Report abuse

Well at least that's the conclusion in 2010. In five years it will be something else.
Science is a great subject but after all, it's a science and is subject to interpretations and variations..........

Posted by: tjmlrc | November 2, 2010 1:23 PM | Report abuse

"Changeable" winter storms are all too common in these parts. This includes the so-called Storm of the Century in 1993, which changed over to sleet and or rain and back to snow, depending on locale. IAD got 12", DCA less than 7" and where I live in NW DC, about 9".

When the precip changed back to snow in the evening it got pretty wild.

A heavy Christmas Day evening/snow in 1969 changed over to sleet and then rain over and east of town, whereas about 15" fell in the far NW 'burbs.

Last year was exceptional in that except for a few hours in the early a.m. during the Feb 5-6 storm, the precip stayed as all snow during the big storms. However, even this changeover, to sleet, impacted snow totals and we "only" had 26" in Glover Park, compared to IAD's 32" inches.

I suspect this year, the magnitude will be smaller... a couple of inches of snow and then the dreaded changeover.

Posted by: JerryFloyd1 | November 2, 2010 2:31 PM | Report abuse


you correct about us often having mixing or changeover issues. My memories of the Feb 5-6 storm were different from yours. I live east of the city and we had a period of heavily rimed snow but no sleet. The hydrometeors remained white rather than being translucent which indicates that the crystal never melted but ended up with lots of riming for supercooled in cloud water. Also the sounding never had a layer above freezing. The second Feb storm however did changeover prior to the temperatures crashing. The typical La Nina storm track is to our west towards the Ohio Valley which is a track than often produces storms with precipitation type issues.

Posted by: wjunker | November 2, 2010 3:47 PM | Report abuse

@Wes, my recollection is that a number of CWG posters reported sleet on Dec. 6, because that's what we thought it was (there were quite a few diehards up at 3 a.m.!). But not being meteorologists, we certainly could have mistaken heavily rimed snow for sleet. Whatever it was, it wasn't flakey. Thanks for clarifying.

As for the second storm, if there was a changeover, I missed it because I didn't get up until 7 a.m., just as the storm started intensifying.

That Christmas Day 1969 storm was a colossal bummer. All the great snow that fell, predictions for a foot or more and then it unexpectedly changed over at least NE of town.

However, there was more snow to follow that winter, including a storm in January when a significant storm was predicted, initially fizzled, then after shot of really cold air came in and we ended up with a decent accumulation.

The next two days in early 1970 saw highs of 17 and 12, two of the coldest consecutive days I've experienced hereabouts (surpassed only by 1994? when consecutive daily highs were 8 and 18 and the city shut down because of the cold).

Posted by: JerryFloyd1 | November 2, 2010 4:49 PM | Report abuse


I could be wrong about the Feb 5-6 event but think it likely it was heavily rimed rather than sleet. Depending on your location, the second big Feb storm might not have had a changeover. At my home , we even changed to freezing rain before midnight before changing back to snow later at night. It was snow by 4:00Am or so and then stayed snow as the temperatures really dropped during the remainder of the day. Both were really neat storms.

Posted by: wjunker | November 2, 2010 5:58 PM | Report abuse

Another great post Mr. Junker!

Does the PDO have any significant effect on our winters? I read that depending on the phase the PDO is in, it can either enhance or hinder an El Niño/La Niña. Isn't the PDO like the El Niño Southern Oscillation, but larger and farther north? T

Posted by: Yellowboy | November 2, 2010 7:27 PM | Report abuse

Sorry! That last word was supposed to be "Thanks!".

Posted by: Yellowboy | November 2, 2010 7:29 PM | Report abuse


Matt will discuss the Pacific Decadal Oscillation (PDO) in his article that will be posted on Thursday. Right now we're in the negative cycle of the PDO and we're in a moderate to strong La Nina which means the PDO should remain strongly negative through the winter.

Posted by: wjunker | November 2, 2010 8:33 PM | Report abuse

@Wes, it was around 3 a.m. and there was something icy falling when I went out for a trudge through my neighborhood. It could just as easily been rimed snow as sleet and now that I think about it, it didn't ping off my parka the way sleet usually does. The key factor you note may have been temps overhead.

Posted by: JerryFloyd1 | November 3, 2010 11:23 AM | Report abuse

I hope to get *really* lucky this winter.

Posted by: ennepe68 | November 3, 2010 4:05 PM | Report abuse

Thanks ever so much for supplying all this really fascinating analysis!

Was particularly interested in your last paragraph of this "Part II" story concerning how in February the pattern changed and "...the timing and the location of the various players were a little out of phase for another snowstorm..."

After the early Feb. blizzard, I for one was waiting for another at least garden-variety dump to REALLY wipe the record books. But it never happened. In other words, the snowiest winter ever for DC included the last 45 or so days being virtually snowless.

Would be really interested in what happened to bring about so drastic a change in a single season (even if it was a shift more to "the norm"). Thanks again...

Posted by: kenkimg | November 5, 2010 10:14 AM | Report abuse


Sorry I didn't respond earlier. It's hard to say why the rest of the winter was largely snow free. The next couple of chances for snow were Miller B type storms that impacted to our north. That's pretty typical of most Miller B type storms though occasionally the reformation of the surface low is far enough south to put us in heavy snow (the Feb 9-10 storm). There also was a storm that went to our south in early march. The NAO remained negative through mid March but for some reason, maybe random luck, the storm track was not quite right for us. One notable difference was that during the big storm period there was always a strong vortex somewhere near Nova Scotia. That was not nearly as evident later in the season so for the Miller B type storms they had more room to come northward and be a little too far north for us to get significant snow.

Posted by: wjunker | November 9, 2010 11:42 AM | Report abuse

The comments to this entry are closed.

RSS Feed
Subscribe to The Post

© 2012 The Washington Post Company