January 12, 2013: Winter's Lies, Damn Lies and Statistics

The title of this blog is made in the context of Mark Twain, who stated simply "Lies, Damn Lies, and Statistics".   This is relevant to understanding the capricious nature of Washington and Baltimore's winters.

You often hear the statistic that our region receives an average of 18"-20" of snow each winter season.   This is indeed true.  But does this average truly convey the actual behavior of a typical snow season?   Far from it.

This past week, the NWS released a simple graph, showing annual snowfall recorded at D.C.'s Reagan National Airport, from 1992 through 2011.  A graph for BWI looks very similar, with slightly higher amounts across the board.  Here is the graph, which is called a time series chart:

Washington's annual snow has varied anywhere from nearly zilch (1997-1998, 2011-2012) to 50"-60" (1995-1996;  2009-2010).   There is tremendous annual variability in our seasonal snow - so much, in fact, there there is very little predictive power in trying to glean one year's expected snowfall from the previous.   A "Snowmageddon" season is very often preceded or followed by a "Snow Drought" year.   Basically, it's all or nothing here.  Meteorologists term this type of behavior "episodic".  In fact, most of our years appear to hover around the 10" annual snowfall mark.

The 18"-20" snow average is really an artifact of averaging lots of very lean years, with a very few episodes of heavy seasonal snowfall.   The average creates a false expectation that our "normal" snow should be in the 18"-20" range.  In fact, 17 of the 22 years above fall closer to 10" than 20".  That's a big difference.   So the intermittent nature of our blockbuster snow seasons has inflated an average value, to the point of conveying little meaning.

Beware, you must, of those wily statistics!

December 31, 2012: The Wild Winds of December

December, 2012 closed out with a series of three wind storms, from December 22 through the 30th.  Each time, the region endured several hours of cold winds gusting between 45-50 mph...during which the NWS issued Wind Advisories, and area utilities dealt with 100's and even 1000's of regional power outages.

Here are the maximum wind gusts (mph) reported at our region's three airports for each of these events:

Winter Storm           Date         BWI        IAD       DCA

Draco                       Dec 22     44            51          43
Euclid                      Dec 27     46            46          45
Freyr                        Dec 30     46            47          49

It is certainly unusual to experience three Advisory-level wind storms in the space of a single week.  But the series testifies to the ferocity of this season's early winter storms, and the overall pattern of a very energetic jet stream. 

In each case, the winds were generated on the back side of a retreating mid-latitude cyclone (low pressure region).   Draco was a powerful storm that generated a blizzard over the Great Lakes, then moved over interior New England.  Euclid was a Nor'easter (coastal low), as was Freyr.

As each storm moved out of our region (to the northeast), it intensified - meaning the central pressure dropped.  At the same time, a cell of strong high pressure approached from the west.  The difference in surface pressure between each low and its high - called the pressure gradient - determines wind strength.   The gradient became very strong for several hours over the Mid Atlantic during each of these weather systems.   An example of this is shown on the surface weather map for Winter Storm Draco:

Surface weather map December 22, 2012 showing Draco's pressure gradient. Adapted from Unisys Corp

Additionally, clearing skies behind each storm allowed the surface air layer to warm up a bit, while the upper atmosphere cooled (as cold air was swept in from Canada).  This destabilized the air layer, causing it to bubble and stir.   Pockets of fast wind were mixed down to the surface, adding momentum to the already rapidly streaming air, in the form of high wind gusts.  Meteorologists call this effect "mixing down of high momentum from aloft".  Consequently, tree limbs snapped onto power lines, leading to periods of brief power outages across the Baltimore-Washington region.

December 29, 2012 Disorganized, Light Snow Event

Light, intermittent snow has erupted across the metro region this morning.  This event, well-predicted in terms of timing, may fall short in terms of expected intensity.   As the surface chart below shows, the storm is actually two separate low-pressure centers - one spreading rain over the Outer Banks and Tidewater, VA...the other spreading light snow across the northern Mid Atlantic.

Intellicast

The lows are embedded in relatively fast, west-to-east jet stream flow.  Accordingly, they will move quickly through the region today, and the jet stream troughs supporting their development are relatively low-amplitude.   While lacking upper-air support over the Mid Atlantic, the southern low is expected to intensify into a Nor'easter as it tracks away from our region...pummeling the New England coast with heavy snow...and will be named Winter Storm Freyr by The Weather Channel.

With the Atlantic Tidewater storm stealing moisture from the more northern storm, the rapid movement and weak intensity...snow totals will be light across our region.

Here is a more regional view of the two storm systems, as viewed by weather radar:

Adapted from NOAA

The fast, west-to-east jet flow will stick around through the coming week.   When the jet stream "superhighway" runs fast and lacks curvy twists (troughs), intense storms cannot develop at the surface.   The computer models are predicting a bit of storminess during the Wednesday timeframe - but any impact will likely be on the light side. 

December 26, 2012: Anatomy of Winter Storm Euclid

This storm is receiving lots of attention from me...because winter storm systems rarely get as complicated as this, around here.   Today we have experienced a mix of sleet, freezing rain, snow, moderate to heavy rain...and tomorrow, strong winds...all from the same mid-latitude cyclone.

My earlier posts have illustrated the complex thermal structure of this storm, tied to our region's geography, and the influence of that thermal structure on evolving precipitation types through the day.

In this post, we step back and look at the whole enchilada of this classic, textbook winter storm.

First, let's look at the synoptic surface map from this afternoon.  Winter Storm Euclid is a complex, sprawling and intense storm consisting of two low pressure centers.  The primary low is located over Kentucky.  Through the day, a secondary or coastal low has begun to develop over the Carolinas.  Eventually, Euclid will transfer its energy, and its moisture, to "Son of Euclid" as the new center tracks up the East Coast.

Euclid's primary and secondary low pressure centers, and associated precipitation pattern.  Intellicast

The precipitation shield surrounding Euclid's counterclockwise whirl is immense, wrapping from the southern tip of Florida, northward along the East Coast, then back west into the Great Lakes.   Green colors show rain and embedded thunderstorms (Euclid's "warm side") and blue colors indicate ice and snow (the storm's "cold side").

Cyclones such as Euclid have strong temperature contrasts; they bring air masses together from different source regions.  The juxtaposition of air masses creates weather fronts within the storm system.  The weather map below, showing the thermal pattern at 5,000 feet, illustrates the classic "ying yang" pattern whereby warm, tropical-source air (yellow) wraps into the system from the south and east, and cold, Canadian-source air enters (blue-purple) from the north and west:

Winter Storm Euclid's temperature contrasts.  Unisys Corp.

These air mass contrasts are what power large cyclonic storms in the mid-latitudes.

The next graphic shows Euclid from the vantage of weather satellite, using the moisture channel.  Dry areas in the atmosphere appear black, while moist areas show up in varying chromatic shades - from blue to red, dark red indicating the greatest degree of saturation.

Adapted from WeatherTap

I have added a few annotations that reveal the complex 3D anatomy of this storm.  First, note the primary low center over Kentucky, and the secondary center beginning to form near Tidewater, VA.  Next, the massive precipitation shield is this storm's distinguishing feature.  Euclid has an impressive feed of moisture from the Gulf and Atlantic.  The moisture influx assumes the form of a narrow, high-speed conduit or "river" of air entering the system from the south, and is called the Moist Conveyor (solid green arrow).  All along this corridor of high water vapor, clouds condense.  Along the East Coast, severe thunderstorms have erupted in the warm, unstable air.  Over the Mid Atlantic, moderate snow-sleet-rain has fallen from the Conveyor for over 12 hours.  And back through the Ohio Valley, heavy snow has fallen along the Moist Conveyor where air temperature has remained below freezing.  The Moist Conveyor is the one element that connects all the disparate forms of heavy and severe weather in this single storm system.

But along the back edge of the storm, very dry air has descended from high levels over the upper Midwest.   Along the track of this descending current, called the Dry Conveyor, clouds and precipitation have evaporated.   Dry air is wrapping deep inside the inner core of Euclid.  This is the hallmark of a cyclone entering its most mature and intense phase.  Eventually, the dry air will sequester the core completely, and the primary low will dissipate.   Meanwhile, the new coastal low will deepen as it moves toward New England, effectively robbing the old core of its Moist Conveyor.

Finally, and most importantly, it is important to understand where these intense cyclonic storms come from, and what sustains them.  The answer is the polar jet stream, a river of intense air blowing from west to east between 30,000-40,000 feet.  Below is the jet stream chart from this morning, which I have annotated:

Jet Stream level winds and their relationship to Euclid.  Adapted from Unisys Corp.

You can see the core of high winds zipping along from California to the Gulf Coast.  However, there are numerous undulations along the way.  Most prominent of these is a deep trough of low pressure over the Great Lakes (labeled).  The airflow around a trough has counterclockwise curvature, and the trough contains a core of very cold air.   As the air streams along the base of the trough, it speeds up as it exits and moves over the Mid Atlantic (think of a trough as a tight curve on the Beltway - you have to slow down through the curvy part, but once you are free, you accelerate).   I have shown a hypothetical red box in the exit region of the trough.  Imagine air exits the north edge of the box, faster than it arrives from the south.  This creates a "divergence" of air flow within the box.  Because the same amount of air must always remain in the box, air must be drawn up from below, from the surface, to help fill the void.   When we remove air from the surface, we create low pressure there.   Air rising up to fill the imaginary box creates widespread cloud and precipitation.  Thus, areas of low pressure and storminess - such as Euclid - develop near the "exit" region of large troughs in the jet stream. 

There is something else helping to sustain low pressure at the surface.  Note that in addition to the flow speeding up through box, it also spreads outward, away, in a fan-shaped pattern.   This is a consequence of very intense troughs such as the one shown here, which take on a certain tilt from NW to SE.  (In meteorological parlance, the spreading apart of airflow at high levels is called "diffluence").   Nevertheless, as the air fans away, more mass is trying to leave the box from the north side, than can enter from the south side.   Once again, extra air must rise up from below to fill the box.

Finally,  note a very intense pocket of high wind in the core of the jet stream, at the base of the Great Lakes trough.  This is called a jet streak.  When jet streaks zip around the base of a trough, they enhance both the divergence and diffluence of air in the trough's exit region.   This helps surface low pressure regions deepen even further. 

December 26, 2012: The Transition from Snow To Ice To Rain

Today's storm is a classic example of a winter storm starting off with a wedge of subfreezing air stuck up against the mountains - called Appalachian Cold Air Damming - creating several hours of snow and icefall across the region.

But as this storm moved closer to the Atlantic coast, it began drawing in relatively mild air off the Atlantic ocean.  At the critical 5,000 foot level of the atmosphere (the level in the cloud layers that determines precipitation type - rain or snow), temperatures warmed rapidly from several degrees below freezing, to several degrees above freezing.  By late morning, rain - not snow - formed in the clouds.  But air in the lowest few thousand feet still remained below freezing (courtesy the cold air wedge).  So rain drops froze into sleet grains - and sleet began to accumulate on all surfaces.

The 5,000 foot analysis map shows how the 0 C isotherm has pushed north and west of the metro centers (compare with the map from early this morning) - coincident with a changeover from snow to sleet:

Early Morning - Cold Air Damming;  Early Afternoon - Warm Air Invades from Atlantic.  Adapted from NOAA

You can see the change in precipitation types of D.C.-Baltimore on the radar between morning and early afternoon - noting how the snow-sleet line shifted north and west of the metro centers:

Morning:  Snow-Sleet Line south of D.C. - Baltimore.  Afternoon:  Snow-Sleet Line North of D.C.- Baltimore.  Adapted from WeatherTap

Around noon, winds at the surface transitioned from cold northeasterly to milder easterly flow.   The wind direction shifted as a new center of low pressure began developing along the Virginia coastline, setting up more of an oceanic fetch of wind into our region.  The arrival of slightly milder air at the surface began scouring out the cold air dam in place in the lowest few thousand feet over Washington- Baltimore.   As surface temperatures warmed, rain that forming at 5,000 feet, stayed rain all the way to the surface.  As of 2 pm, precipitation area-wide became plain rain.

This progressive transition from an air layer deep enough and cold enough for all snow, to an intermediate structure supporting sleet, finally to an atmosphere warm enough for all rain, took place over 12 hours.   The warming of this layer proceeded from top-down.  The final graphic below illustrates how meteorologists quantify the rate of temperature rise at 5,000 feet - a process called warm air advection.  The map shows a very strong pocket of warm air advection (solid red colors) over VA-MD early this afternoon.   Note the southeasterly winds, up to 75 mph, transporting mild air inland off the Atlantic.   D.C. and Baltimore lie within the northern edge of this intense, warming pocket.

Warm air advection early this afternoon changes the region over to all rain.  Adapted from NOAA

December 26, 2012: Cold Air Damming Bringing Frozen Precipitation This Morning

Overview of the Storm

This winter storm, called Euclid by The Weather Channel, is creating its heaviest snows and high winds across the Ohio Valley, where blizzard conditions prevail this morning along the back, cold side of this mid-latitude cyclone.  The center of the storm is over the Tennessee Valley, and the system is moving toward the northeast.  These enormous, intense, winter cyclones have "cold side impacts" and "warm side impacts" - a case of a storm with a split personality!  The band of heaviest snow usually develops 100-200 miles to the northwest of the storm center.  But in the southeast segment, called the "warm sector", the combination of warm, humid air and vigorous uplift often creates pockets of severe local storms, including thunderstorms and tornadoes.   Winter Storm Euclid has created both a blizzard on its northwest side, and tornadoes in its warm sector.   Here is the radar snapshot of this impressive storm from earlier this morning, clearly showing the cold and warm side precipitation features:

Surface weather in Winter Storm Euclid.  Adapted from WeatherTap

In this diagram, you can see the center of the cyclone, marked with a large, red "L".  To its north and west is the heavy snow band.  To the southeast, over the Carolinas, is a warm air mass containing bands of thunderstorms - some of which may produce another round of severe weather later today.
Note the orange-colored wedge of ice over the Appalachians - this is the region of overlap between warm and cold air masses - with the warm air sliding over dense, cold air trapped in mountain valleys.   The heavy magenta line roughly bisects the storm into its warm-side and cold-side.

Baltimore Impacts:  Why Geography Matters

 

Closer inspection of the morning radar shows a curious wedge of ice and snow over central and western MD, NOVA and the WV panhandle - extending along the spine of the Appalachians in Virginia:

Snow and ice wedge over the Mid Atlantic.  Adapted from WeatherTap

"The Wedge" as meteorologists refer to this setup, is classic Mid Atlantic, wintertime meteorology.   The Appalachians frequently trap cold, sub-freezing air flowing in from the northeast.  A shallow, frigid layer called Appalachian Cold Air Damming abuts the Appalachians over the Piedmont, and also gets trapped within the long, linear valleys of the mountains (such as the Shenandoah Valley).   This cold air is overrun above by warm, humid air flowing off the Atlantic.  The vertical layering of air masses provides a recipe for cloud layers containing various forms of frozen precipitation - freezing rain, sleet, snow.  Thus far, we have experienced all of these on this "multiple choice" winter day.

Cold air damming develops when a ridge of high pressure (caused by dense, chilly dammed against the mountains) sets up across the Mid Atlantic.   The pressure ridge east of the mountains is shown on this morning's surface map.  Solid lines are isobars (lines of constant pressure):

Surface map showing pressure pattern within Winter Storm Euclid.  Adapted from NOAA

The pressure ridge tightens the pressure gradient, accelerating a chilly, northeasterly breeze across the Piedmont.  This brings in more cold air at the surface, reinforcing the cold air wedge.

Let's see what this chilly air current does to the temperature field.   The morning surface temperature map (below) shows the cold air wedge in the pattern of isotherms (lines of constant temperature).   Red isotherms are temperatures above freezing;  blue are those below freezing;  the purple isotherm is the critical freezing line:

Surface temperature map showing critical freezing line (purple isotherm) and cold air wedge.  Adapted from NOAA

You can see the wedge-shaped pattern of sub-freezing air sitting over the Maryland Piedmont and mountains.   As is quite typical, the surface freezing line trends SW to NE, and this morning it bisects the major cities Washington and Baltimore.

Finally, the map below indicates that this morning's cold wedge extends upward through 5,000 feet.  This is a deep, sub-freezing air layer and it guarantees that snow will fall:

5,000 foot temperature map showing deep cold air wedge.  Adapted from NOAA

In fact, compare the location of this deep, cold wedge with the region of radar-observed snow, and you see excellent correspondance:

Location of snow wedge.  Adapted from WeatherTap.

You don't see correspondance with the ice storm extending south along the mountains.  This is because sleet is generated from a much shallower layer of cold air damming...temperatures at 5,000 feet are above freezing, meaning precipitation starts in the clouds as rain.  As it falls into a sub-freezing air layer perhaps only 2,000 feet thick (or less), it quickly freezes into sleet grains.

December 24, 2012: Surprise Christmas Eve Snow North and West of D.C.

Well, I do get it wrong from time time.

My call yesterday was for light precipitation this afternoon, mainly rain, mixing w/ pockets of sleet.  Instead...counties north and west of D.C. are being treated to an afternoon of light snowfall, and we will likely see 1"-2" of accumulation before the storm pushes out between 5-6 pm tonight.

I believe the NWS was also taken by surprise!  They were in "reaction" mode, rushing out winter weather advisories as the flakes began to fall.

First, the surface weather map (below) shows a weak region of low pressure approaching from the west.  A stationary front separates mild air and rain to the south, and cold air with snow to the north:

Intellicast

This is a fast-moving, weak system with limited moisture.  Early this morning, the cold (sub-freezing) air in the lowest 5,000 feet of atmosphere lay well to our north, across northern PA.   It certainly appeared that the precipitation, when it did break out over Baltimore-D.C., would be rain.

Alas, the early morning weather balloon launch at Dulles revealed very dry air in the lowest 20,000 feet of atmosphere, with temperatures hovering right around freezing at 5,000 feet.  When precipitation did begin to form in the cloud layer - perhaps as a mix of rain and snow at that level - it evaporated into the dry air.  Evaporation extracts heat from the air.  This chilled the air layer at the critical snow-making level (5,000 feet) several degrees below freezing.   The cooling continued, in spite of southerly winds at 5,000 feet pushing warmer air in from the south.  In the parlace of Meteorology 101:   Evaporative cooling won out over warm air advection.   Snow, not rain, was able to efficiently form in the cloud layer, and remain as snow all the way to the ground.

I've highlighted this pocket of evaporatively chilled air in the diagram below - showing isotherms of temperature at 5,000 feet.  Isotherms at or colder than freezing are shown in blue, those above freezing are shown in red:

Modified from NOAA

Here is a radar show that shows how the wedge of moderate snow (blue shades) line up with this cold pocket at 5,000 feet:

Weathertap

Now for the Wed-Thurs storm:   My thinking on this is basically unchanged from yesterday's blog post.   However, stronger cold air damming than forecast yesterday may lead to a more prolonged period of iciness (sleet, freezing rain) at the onset of this event.   I still see a strong push of warm air as the bulk of moisture moves through.   Upwards of 80%-90% of the precipitation in the metro will likely fall as liquid, but the percentage of ice will be higher farther west and north of D.C. - Baltimore.   Cold air damming, once established, can be slow to dislodge.  This means subfreezing air hangs tight in the valleys.