Saturday, October 27, 2012

October 27, 2012: The Beast of the East

We're starting to get down to details:

1.  Model Track/Landfall Location

With a bit of waffling north and south (as expected) the models appear to be converging on a solution that brings the storm into the strip of real estate from Delaware-New Jersey.   I expect that the track forecast will tighten up in the next 24 hours.   As long as this is the case, Baltimore-D.C. stay in the less energetic left-of-track semicircle:

National Hurricane Center 11 AM Track Forecast
Obviously, we would prefer to be farther away from the center...but this vortex is so large, you'll need to be 100-150 miles away before significant effects are no longer felt!    Here is the afternoon spread in track forecast (still highly divergent):

South Florida Water Management District
2.  Intensity Forecast

I believe this aspect will remain problematic.   The tropical core itself has become disorganized and has opened up;  Sandy is barely clinging to hurricane status.  The energy-yielding thunderstorms are no longer sitting over its core.   The storm is becoming tilted over by strong vertical wind shear, and dry air is invading from the south and west.   The storm is likely in its early phases of extratropical transition.   The models historically do not demonstrate as much skill with intensity as they do with track.  I'm going along with the Hurricane Center's general idea, which suggests a bit of intensification is likely prior to landfall. This is because of the anticipated jet stream interaction, which adds energy, and also the energy contained in the strong contrast between air masses (cold continental air abutting with warm tropical air).  Will the storm remain organized enough to efficiently tap this new source of energy?  I would still plan on ex-Sandy maintaining tropical storm status at landfall.

Here is a portrayal of the three elements beginning to interact, creating a hybrid vortex out of post-tropical Sandy:  (1)  strong cold front moving across the Mid Atlantic;  (2)  large trough (southward dip) in the jet stream (over the Upper Midwest);  (3) the tropical vortex itself:
Surface Weather Map, 11 AM Saturday (Intellicast)
Jet Stream Chart, 8 AM Saturday (NCAR)
3.   Wind Forecast

The left semicircle usually experiences weaker winds, because of several factors:  (1)  the storm's forward movement, or translation, which is in a direction opposite the counterclockwise swirl of wind (the storm is moving W-NW, but the winds are blowing from the W-NW - the opposing effects cancel somewhat);  (2) in the right semicircle, E-SE winds add to the direction of movement, which is from the E-SE;  (3) on the left side, winds are blowing over land (where surface friction is large) but on the right side, winds continue to scream in off open ocean, where they have built up momentum;  (4)  gustiness is reduced on the left side, because of the cooler, stable air mass in place over land.   When the air layer near the ground is stable, it resists vertical stirring, meaning less of the higher momentum air aloft gets mixed down to the surface.

Here is the NWS maximum gust forecast during the window of maximum storm impact (Tuesday morning-evening).  Notice how the Bay really amplifies the wind speed, and the highest winds shift to north-central MD as the storm tracks by to the N&W:

National Weather Service
National Weather Service
How fast will the winds diminish?  (1) If the storm stalls inland over PA, strong winds remain longer;  (2) if the storm makes progress toward the N&W, the winds will begin tapering Wed AM.   Winds will start to taper once inland, as the hybrid vortex "occludes" (the low pressure fills in and loses its jet stream support) and the supply of oceanic heat and moisture is lost.   However, this large vortex will take a while to wind down, because of its tremendous angular inertia (think of it as an enormous atmospheric flywheel).

4.  Big Rains

We're still on track for significant rainfall, in the 6" range.  This could begin falling as early as Sunday afternoon.  When large, landfalling storms approach the East Coast, the strong onshore flow of warm ocean air induces a frontal boundary along the coast.  In Sandy's case, the cold front presently moving across the Mid Atlantic will set up shop along the coast, and will become reinforced...becoming the coastal front.  There are myriad factors contributing to heavy rain:  (1) Heavy rains tend to concentrate not just around the core of this storm, but also focus along the coastal front, where there is strong, sustained uplift of air.  Heavy rains will likely spread all the way up to New England well in advance of landfall, along this front;  (2) if storm movement slows, the higher the local rain accumulation; (3) uplift of tropical air along the east slopes of the Appalachians will create "orographic enhancement" of heavy rain;  (4) the strong uplift of air in the jet stream trough will generate heavy rainfall over a wide area;  (5) a mitigating effect for Baltimore may arrive Wednesday, in the form of a surge of dry air from the southwest...shutting down precipitation over parts of the interior Mid Atlantic.

NOAA

5.  Tidal Surge

Surge effects, coinciding with astronomical high tide, should be in the 3'-4' range, mainly in the northern reaches of the Bay and large main stems feeding toward Baltimore and Annapolis.   As the storm approaches from the east, W-NW winds will initially push water out of the river mouths and toward the eastern shore of the Bay.  As the storm passes to the N&W, the water will slosh back, and E-SE winds will push water into the mouths of main stems, and along the Bay's western shore.
NOAA
This is NOT another Isabel, provided the storm track remains to the north of Baltimore.




Friday, October 26, 2012

October 26, 2012: Sandy's Baltimore Impact Becoming More Certain And More Severe

What The Prediction Models Say:

The prediction models continue a westward trend in the point of landfall;  the official hurricane center forecast issued at 11 AM today calls for landfall of a Cat 1 storm over Delaware early Tuesday morning:
National Hurricane Center
As we discussed yesterday, note the very broad cone of uncertainty.  The various models still remain widely divergent:
South Florida Water Management District
...as do the various ensemble runs of individual models:
Weather Underground
Nonetheless, because of the consistent wholesale westward shift in track guidance - closer to Baltimore - confidence is growing for increasingly severe impacts in our region.   Bear in mind a couple things:  (1)  don't focus on the center point of the hurricane center prediction, as this is a very large storm, and the most severe effects will not be confined close to center;  (2) we still have another 5-6 cycles of model runs to go before landfall.  These runs are made every 12 hours, and I would not be surprised to see further shifts in the landfall point, perhaps even back toward the eastern side of the guidance spread.  Some back-and-forth movement of tracks is typical for this timeframe in the numerical track guidance.

Likely Impacts In the Baltimore Metro:

Again, because we are 2-3 days out from passage of the vortex past our region, we can't get too specific here.   But some generalities are starting to emerge.

(1)  Tropical Storm force winds.  If we remain on the LEFT SIDE of the track, the overall energy will be less than on the right side.   This means lower wind speeds and gusts.   This applies as long as the storm is moving briskly.  Here is a very preliminary estimate of the sustained winds on Tuesday morning and early afternoon, the likely period of peak activity:
National Weather Service
These winds are tropical-storm intensity (40+ mph sustained, with higher gusts, perhaps in the 50-60 mph range).  The highest winds will be over the Bay and along the eastern shore of the northern Bay.  These winds will initially push water OUT of the tidal mainstem rivers draining along the western shore of the Bay, but pile up a surge of several feet along the Bay's eastern shores and mouths of major tributaries.  Winds will also be stronger in the high elevations of the Maryland Piedmont and Blue Ridge;

(2)  Heavy rain is expected, to the tune of 6"-10".  Here is the latest graphic illustrating likely amounts:
National Weather Service
These rains will be fairly steady with intermittent, heavy bursts, for a 24+ hr period.  So there is certainly the possibility of urban and small stream flooding, as well delayed flood crests in the main stem rivers

(3)  Timing of impacts.  Because this is a large storm, impacts will begin long before the center moves onto Delaware.  Conditions will begin deteriorating Sunday night...meaning arrival of rain, and a freshening of the breeze.   Winds and rain intensity will continue to ramp up through Monday, peaking early Tuesday AM.   As long as the storm remains on the move, winds should begin to abate early Wednesday AM, and rain will slacken.    The winds may die down particularly quickly, from a combination of the storm moving away from Baltimore, and fairly rapid weakening of the storm once inland.

All of this information is predicated on the latest (11 AM) hurricane center track and landfall information, which is the best guess offered by the hurricane prediction pros - a mental consensus of the dozens of track and intensity forecast models.  There are models that suggest alternate scenarios, including (1) a stronger storm at landfall than the hurricane center is forecasting;  (2) storm tracks further north and east of our region;  (3) the storm becoming stationary for a time along the coast, and even executing a tight loop - which would greatly prolong the window of severe weather.   So much of the outcome depends on how the post tropical phase of Sandy interacts with the large, intense jet stream trough, and the vagaries of its poorly-understood extratropical transition process.  This interaction has not yet started.  There are few historical precedents for this type of unusual tropical-extratropical merger to gain insight from.  There is one model (the European Center model) that continues to portray an absolutely frightening storm, the intensity of which has never been experienced in our region - but there is some reason to discount this scenario as unlikely.

Thursday, October 25, 2012

October 25, 2012:  Countdown to Hurricane Sandy


After a long hiatus (I have been on sabbatical) I have decided to re-activate the UMBC Storm Page, for obvious reasons.  I will be updating this daily as long as Sandy remains a threat.

Track Uncertainties 

 

Let's start with the Hurricane Center's 11 AM advisory:

NOAA Hurricane Prediction Center



Sandy is a very late season hurricane, and an intense one at that - presently a high-end Cat 2, with tropical storm force winds extending across a 200 nmi diameter vortex.  The forecast is for Sandy to accelerate northward this weekend and gradually weaken (from a combination of cooler ocean water and increased wind shear).  It is expected to dip just below hurricane intensity once it passes north of the Carolina Outer Banks.



But DO NOT focus on the exact center-line of the forecast track, but rather on the large envelope of track uncertainty, which expands with time.   The numerical prediction models have been struggling for days to reach consensus on exact track, and point of landfall.  Some of this is to be expected, since there are a great many different types of track models being used, and large errors are typical in the 5-7 day range of prediction.   To provide an example of the disagreement, the plot below (called a "spaghetti diagram") illustrates the possible track solutions from a large number of these models, compiled at 2 pm today:

South Florida Water Management District
Most models curve Sandy, as a post-tropical storm, back TOWARD the U.S. mainland.   But there is tremendous divergence, or spread, in where the storm will make landfall.

Another type of track prediction comes from what's called an "ensemble model".  This is where a single forecast model is run with many different meteorological starting points, in order to better characterize the error inherent in the simulation.   An ensemble of individual runs is generated.  An example of this is shown below, which is the ensemble model forecast for a widely used model called the GFS:

Weather Underground


Once again, lots of possibilities exist for various points of landfall, ranging from the Mid-Atlantic, New England and Canada.   The white track indicates the "center of mass" of all these various ensemble solutions.  Notice the odd kink over Long Island.  This is created when the storm slows to a stop (stalls) and essentially loops on itself, before making final landfall.

What Will The Impacts Be? 

 

With all the uncertainty in the track, it's still premature to nail down specific impacts for the D.C.-Baltimore region.   Right now we can only speak in general terms:

1.  Timeframe will be Sunday evening through Tuesday, with greatest impacts on Monday;
2.  Expect, at a minimum, periods of heavy rain (squalls) with sustained windy conditions

It's still too early to foretell (1) exact amount of rain, and (2) speed of maximum winds - there are too many variables involved without knowing (a) exact track;  (b) vortex intensity as it enters the post-tropical phase;  (c) storm speed (forward motion);  (d) vortex size.

Suffice it to say, given the POSSIBILITY of a landfall as close as Delaware (bringing the storm center very close to Baltimore), it's prudent to begin preparing for rapidly deteriorating conditions Sunday night, the possibility of urban and small stream flooding, and extended power outages.
The area utilities (Dominion, BGE, Pepco) have already started their initial preparations.

The Nature of The Beast 

 

It's important to understand that the storm that impacts the Mid Atlantic and New England is not expected to be purely tropical, but rather a hybrid type of system that combines elements of a hurricane and a nor'easter (nor'easters are intense wintertime coastal cyclones).    As Sandy approaches our area, it is expected to interact with a vigorous trough in the jet stream (currently over the western U.S.).   So while we have to not only forecast the track and intensity of the tropical storm, we also must accurately predict how the jet stream trough evolves over the next few days, and the manner in which it will phase with post-tropical Sandy.   Some hurricane-trough interactions can intensify a vortex, while others can lead to its demise.   Sandy's interaction with this trough is expected to (1) maintain a strong post-tropical storm against factors that would otherwise cause it to spin down (decay);  and (2) draw the storm westward toward the East Coast.

The process by which a tropical cyclone morphs into an extratropical system is termed extratropical transition, and a lot about this process remains poorly understood.   Transitioning storms change character in many ways - for instance, they (1) acquire weather fronts;  (2) develop strong asymmetries in the wind and precipitation (heaviest rains shift to the north and west side, highest winds develop on the east side);  and (3) expand considerably in diameter.   Some storms accelerate rapidly once they become embedded in the jet stream.  If a constructive type of phasing between a trough and vortex occurs, the storm can even re-intensify, as it begins feeding from dual energy sources (both the warm ocean and the jet stream);  this process is called rejuvenation.  These transitions are tricky to forecast because meteorological data near the surface and aloft is non-existent, or sparse, over the open ocean.

Some of you may remember Hurricane Hazel from 1954, which created quite a mess across the Mid Atlantic.  While the meteorological situation with Sandy is different in many respects, the Hazel that struck Washington-Baltimore was a storm that underwent extratropical transition in mid-October, interacting with a very strong trough in the jet stream.   The post-tropical vortex over land drew a tremendous amount of energy from the jet stream, and allowed Hazel to generate hurricane-force winds all the way to Toronto, Canada (the storm made landfall in the Carolinas).  Again, I stress that what we will experience with regards to Sandy is not a Hazel repeat, but Hazel serves as a reminder of how potent a tropical-extratropical interaction can be.

For additional meteorological perspectives on this storm, check out the blog of my colleague and hurricane expert Dr. Michael Folmer at NOAA:

http://goesrnatcentperspective.wordpress.com/






Tuesday, March 6, 2012

March 5: Surprise Snowstorm Narrowly Misses Washington-Baltimore


Perhaps unbeknownst to some, the Washington-Baltimore region narrowly escaped a major snowstorm on Monday morning, March 5.    Had the track of the parent low pressure system jogged 100 miles further north (a minor deviation given the large size of these weather systems) our region could have been blanketed with a fast 6”-10” of dry, powdery snow…in addition to hours of severely reduced visibilities on the highways.
The NWS correctly predicted that a fast-hitting “clipper” snowstorm would track south of DC, dropping up to 2”-4” inches across western and central Virginia.   The appropriate winter weather advisories were issued.   Then, in the midst of the mid-morning snowfall, a small, nearly-stationary band of heavy snow set up across several central Virginia counties.   Snowfall rates intensified.   The NWS rushed out a winter storm warning for the affected counties.  Not long after the warnings were posted, the heavy snow came to an end.    To be fair to NWS, the very latest model predictions that morning hinted that a heavy snow band might set up across central VA.   As soon as it became clear that mother nature was “cooperating with the models” (radar trends indicating an increase in snow rates), NWS acted as quickly as it could to elevate the level of warning.
Below is the Monday morning synoptic chart showing the low-pressure clipper system tracking across Virginia:
 
Here are the warnings issued by NWS (notice how small of a geographical area was impacted):
  And here are the snow reports:
 
Even with fairly sophisticated forecast models, we still struggle to predict smaller-scale atmospheric processes, such as those that set up narrow bands of intense snow. 
So…what happened?   The regional analysis (below) shows the details.   This is an analysis at the 5,000 foot level, where the atmosphere manufactures snow.   The solid black lines are isobars of pressure.  The large dip across the Mid Atlantic is a trough of low pressure, causing vigorous ascent of cold, moist air over central VA.  The narrow band of solid green (oriented west-east across central VA) is the heavy, nearly-stationary snow band identified on radar.  This band formed along a warm frontal boundary stretching across the southern part of the state.   But the key here is that this front was rapidly intensifying, meaning it was lifting cold, moist air more vigorously through the morning.   The technical terms is “frontogenesis”, and the three purple contours surrounding the snow band show where the rate of frontal intensification was greatest.   In other words, the snow band set up within this small bulls-eye of frontal strengthening.
  In closing, it is rare that a fast-moving clipper system such as this produces as much as 10” of snow.   It seems that this storm, for a brief few hours, was able to pull in additional moist air off the Atlantic as it tracked east of the Blue Ridge.   Combined with a strengthening front, and sufficiently cold air during the early morning hours, nature assembled a winning combination of ingredients for big snow.

Thursday, March 1, 2012

Feb 29, 2012: A Day of Unexpected Flooding in the Mid Atlantic


The same potent weather system that brought severe weather to the central Plains on Feb 29 (including the Branson, MO tornado) created a very rainy day in the Mid Atlantic.   Periods of moderate to heavy rain lead to localized flooding in the Baltimore region and prompted the NWS to issue several flash flood warnings during the afternoon.  Thunderstorms were even reported in our region during the late afternoon.
The weather map below portrays a well-developed wave cyclone (extratropical cyclone) over the Great Lakes, approaching the Mid Atlantic.   Three distinct air masses are brought into juxtaposition around this low:  (1) cool, humid air streaming from the northeast off the North Atlantic ocean (the air mass that lay over Baltimore on the 29th);  (2) a “warm sector” of warm, humid air streaming toward the north from the Gulf of Mexico, blanketing the entire southeast U.S., ahead of the cold front;  and (3) a cold, dry Canadian air mass that swept from the northwest around the back side of the low.
 
The warm sector spawned widespread severe activity over the eastern U.S.  Lines of severe thunderstorms generated tornadoes and damaging straight line wind across the central Plains early on Feb 29.    Severe thunderstorm watch boxes (red) and tornado watches (yellow) are shown on the map above (7 pm Feb 29),  covering most of Appalachia and the mid-South. 
Over Baltimore, batches of heavy rain swept repeatedly from west to east through the day.   This type of “training” was facilitated by the warm front (shown in red in the chart above).  Along this front, warm and humid air from the warm sector streamed over cooler air at the surface.   Continuous uplift of humid air established a corridor along the front,  oriented west to east across our region.   Winds aloft were blowing from the west, parallel to this boundary, pushing new batches of heavy rain toward Baltimore from the west.
 
This chart shows the rain accumulation from yesterday, derived from rainfall estimates.   You can get a sense for the west-to-east corridor of heavy rain stretching from WV across VA, MD and PA to the Eastern Shore.  There were widespread 1”-2” amounts with some pockets in excess of 2” just southwest of Baltimore.

Tuesday, February 21, 2012

February 19: A “Nostorm” for Baltimore


That’s right, we had a nostorm, not a snowstorm.   For days, the computer models went back and forth on this one.  Early on, it looked like DC and Baltimore would be ground zero for a significant Mid Atlantic snowfall.  But then the models began tracking the storm further to our south, and moving it quickly out to sea.   And the bulls eye for heavy snow turned out to be southwestern and central Virginia.
Let’s look at the factors that contributed to a “near miss” for Baltimore.
Not An Ideal “Snowmaker” Setup
To get heavy snow in Baltimore, we need three key ingredients:  (1) A rapidly intensifying coastal cyclone (nor’easter) moving slowly up the coast (tracking from south to north); (2) a deep layer of cold air, and (3) abundant oceanic moisture.
As the surface synoptic chart below shows ) (7 pm, February 19), a cyclone did develop along the Gulf of Mexico and track across the southeastern U.S.  However, the jet stream – which has been stuck in a zonal or “west to east” configuration most of this winter - never developed a deep loop or “trough” along the east coast.   These troughs tend to draw coastal cyclones northward along the coast.   The February 19 storm instead tracked nearly due east, emerging into the Atlantic south of the Outer Banks.   It moved quickly, meaning fewer hours for heavy snow to accumulate, nor did not undergo the rapid deepening that characterizes truly heavy snowfalls.

One of the key reasons why this low failed to rapidly deepen hinged on the configuration of the jet stream (see chart below).   This chart shows the winds at the 30,000 foot level.   The jet had split into two separate streams:  A weak northern stream draped across New England, and a vigorous southern branch with a core of 150 mph winds across Georgia.   This “split flow” pattern is often a precursor of big East Coast snowstorms.   However, the trough in the northern branch failed to link up or “phase” with the trough in the southern branch.   Had this occurred, a much deeper trough would have set up over the East Coast.  This would have promoted a rapidly intensifying storm that moved north along to the coast (the classic “snowmaker” type of nor’easter for the Mid Atlantic).

Cold Enough For Snow
The second main ingredient, a fairly deep layer of subfreezing air from surface through at least 5,000 feet, was present in this case.   In fact, one of the key elements is called cold air damming, in which a cold, dense air mass gets wedged up against the eastern slopes of the Blue Ridge.  The figure below shows the characteristic surface pressure ridge east of the mountains when damming is in place.   Note how the isobars (black solid lines) sag to the south east of the mountains, indicating dense, cold air flowing southward.  Superimposed on the isobars is the distribution of precipitation as revealed by regional radar.  While the precipitation fell as rain over North Carolina, the heavy batch over Kentucky, southwest Virginia and streaming east across Central Virginia is snow, forming within a deep, cold air mass.


The figure below shows the actual air temperatures at 5,000 feet, the critical snow-forming layer in the clouds.   The closed vortex of the coastal low is centered over North Carolina’s Outer Banks.  To its west, a tongue of very cold air is being drawn southward (blue and purple colors), over the Appalachians and into the cold damming region east of the mountains.   Temperatures at 5,000 feet were -4 to -8 °C, plenty cold for vigorous snow formation.



On A Knife Edge:  Insufficient Moisture Over Baltimore
The southern track of this storm put the Baltimore region on the extreme northern fringe of the storm system, where clouds were being undercut by very dry air in the lowest several thousand feet.   This resulted in a sharply delineated border or northern edge to the heavy precipitation, which remained mainly south of DC.   The sharp cutoff in snow is shown on the regional radar image below.   Superimposed on the radar image is the amount of moisture contained in the air (solid lines and green shaded regions) – a quantity called “precipitable water”.   You will note over an 1” of liquid available to the storm along the Outer Banks, but the precip water drops off rapidly moving north toward Baltimore.  Because the air was so dry over us, any snowflakes that managed to develop completely sublimated before reaching the surface. 

Major Impacts Over Appalachia
The snow that did fall north and west of the low’s center was heavy and wet, thanks to large precipitable water content close to the storm’s center.  Snow totals ranging from 6” to 10” were fairly widespread across southern WV and southwest VA, with locally higher amounts at the highest elevations.    The snow accumulation map below reveals a fairly compact region of heavy snow:

 
Over 1000 traffic accidents were reported across the snow region, and as predicted, the blanket of heavy snow brought down tree limbs onto electric utilities.  More than 60,000 power outages led to an extended period of power restoration.  The map below shows “ground zero” for these outages on the day after the storm:

Lucky This Time…
In a nutshell, Baltimore dodged a bullet, thanks to a weaker storm that moved rapidly and too far to our south.   But I find it instructive to do a careful post-mortem of these “near misses” because there are always important meteorological lessons to be learned!
 

Baltimore’s Only Brush With Winter: February 11, 2012 Arctic Front and Snowfall


This has been an unusual, but not unheard of, mild and snow-free winter.   Heavy snow in Baltimore can best be described as “episodic” in nature, a feast-or-famine in terms of total seasonal snowfall.   Recall the record-setting 2009-2010 winter in Baltimore, with a season total snowfall of 77 inches.   This season, cold air masses have sequestered well to the north of the Mid Atlantic and the jet stream has been largely zonal, running flat in a west-to-east flow.   Without big, cold meanders in the jet building down the east coast (these are called jet stream troughs), storminess and winter air outbreaks have been minimal.    Whether we can blame this behavior on La Nina, and/or the North Atlantic Oscillation, remains to be carefully assessed once the season ends.
Baltimore endured a brief winter blast on February 11-12, when a chunk of that cold air mass over Canada pushed south into the Plains and Great Lakes.   The anticyclone that propelled this cold air south and east was intense, with a surface pressure of 1046 mb (fewer than 5% of all anticyclones over CONUS ever reach this intensity).   The leading edge of this cold surge invaded the Mid Atlantic in the form of an arctic front – essentially a cold front on steroids.   The synoptic weather map at 7 AM on February 11 (below) shows the front (blue line running through West Virginia) poised to blast through Baltimore:

Ahead of the front, there was sufficient moisture and an unstable (overturning) air mass.   As this air was lifted along the front, a narrow band of convective snow showers erupted east of the Blue Ridge Mountains.   Around 2 pm, this sudden squall slammed through the area.   Visibility plummeted as a heavy fall of snowflakes and graupel (tiny white icy spheres) fell in a volley, accompanied by gusty winds, rapidly falling temperatures and even lightning.   Many areas picked up a quick ½”-1” of accumulation in a matter of 20 minutes.   The regional weather chart (below) shows this squall as detected by weather radar, along with snow streamers blowing off the Great Lakes and a patch of upslope snow along the Allegheny Plateau.



In this image, isobars (lines of constant pressure) are drawn at 2 mb pressure intervals.  The very compact squeezing of isobars implies an intense pressure gradient, which drives the wind.   Winds from the northwest rapidly increased trough the late afternoon as the heart of the cold air mass began to filter in.  The arrival of cold air from the northwest was very swift.  The diagram below shows this cold air advection at about 5,000 feet above the surface.  Cold air advection is a calculation of how rapidly the temperature changes at a location, due to the influx of cold air by the wind.  Dark blue shades in this diagram show where the chilling effect was most pronounced.   Note that the air temperature at 5,000 feet above Baltimore was 18° F!  


But really cold air was only part of the story that night.   Wind speed steadily increased through the evening, as surface pressure surged in Baltimore.    The winds were also very gusty, peaking at 49 mph.  The diagram below, called a time series, plots hourly observations of wind gust and air temperature at Dulles airport.  Note the arrival of the arctic front at 3 pm on February 11, followed by the precipitous drop in air temperature and surging winds.
 

Intermittent snow squalls continued through the evening, exemplified by the patch of moderate snow captured by weather radar just before midnight on the 11th:


The air temperature never climbed above 32° F the next day!
What’s interesting is that none of this extreme weather – arctic chill, thundersnow, high winds – were actually produced by an area of low pressure, but rather by an exceptionally strong anticyclone.    We normally think of low pressure regions or cyclones as bringing disturbed and stormy conditions, while anticyclones or regions of high pressure herald fair weather.  But in the case, the tables were turned!