Our March 3 snowstorm was caused by a disturbance moving along an unseasonably chilly arctic front.   Read on to learn all the details.

A swift temperature drop and rapid change from rain to snow provided our winter-fatigued region with a bit of meteorological “shock and awe”.   Unlike the powerful February 13-14 “Snochi” storm, this system was not energized by a rapidly deepening Nor’easter.  Instead, a narrow zone of “overrunning” type snow developed across the Mid Atlantic.

In the sections below I will discuss some of the unique aspects of the system that created 3-8″ of snow accumulation region-wide.

Key ingredients come together

Figure 1 (below) illustrates several ingredients that congealed over our region.  These include (1) an arctic front that sagged south across the Mid Atlantic;  (2) a low pressure disturbance tracking west-to-east along the front; and (3) Atlantic moisture streaming into the arctic air mass.

Figure 1. Key elements of a classic, overrunning-type snowstorm.   On this surface chart, valid 7 AM March 3, note the region of low pressure, arctic front and zone of overrunning precipitation (adapted from intellicast.com).

This depression was fairly weak, with a central pressure of 1008 mb at 9 a.m., March 3.   Compare this value to Snochi’s 990 mb while it dumped heavy snow across our region.   However, the March 3 vortex was very effective at drawing moisture northward and pulling arctic air southward. Additionally, the circulation provided a source of sustained convergence and uplift.

The arctic front was remarkable for its tight thermal gradient, shown in Figure 2.   At 10-11 p.m. on March 2, area-wide temperatures hovered in the mid-30s while light rain was falling.   Temps were dropping slowly, then took a nosedive around 4 a.m., reaching rates of -2 to -4 °F/hr.   This southward surge of cold air dropped  Washington Reagan’s temp to 22°F by 9 a.m.  BWI and Dulles experienced similar drops, both sites hitting 16 °F by 9 a.m. with northerly winds gusting at 25-30 mph.

Figure 2. The cold air invades!  At an elevation of 3,000 feet, this figure compares isotherms at 10 PM, March 2 and 9 AM, March 3 (adapted from NOAA SPC).

Upper-air jet stream dynamics provided some additional uplift (Figure 3).   During Snochi, I showed how coupled jet streaks enhanced rising air over the Mid-Atlantic.  On March 3, a single jet streak was present, and our region was located beneath its right entrance region (labeled “RER” in the figure).   This is a quadrant of strong upward motion – the so-called “sweet spot”.

Figure 3. Textbook jet streak at 30,000 feet and its right entrance region (RER) positioned over the Mid Atlantic (adapted from Unisys.com).

The jet streak established a three-dimensional circulation beneath its entire entrance region, causing surface winds to accelerate from north to south.   This, in turn, caused arctic air to surge southward east of the Appalachians. (Technical note:  Meteorologists called this as an “ageostrophic transverse jet circulation” and it’s one of the classic ingredients for decent snow).

Precipitation type transition

An interesting twist to this storm was its abrupt transition from rain to snow, with little or no icy interlude.

This was due to the very steep nature of the arctic front and its strong thermal contrast – basically, a nearly-vertical wall of super-chilled air sliding south. Sometimes the cold air mass takes on a relaxed tilt, with a gently sloping surface.  This allows overrunning precipitation, forming as rain in the mild air aloft, to re-freeze into sleet or freezing rain while traversing the shallow cold layer.  A broad zone of iciness sets up between rain and snow.

But two back-to-back soundings taken at Dulles illustrate the drastic air mass change in the lowest 10,000 feet (Figure 4).   The balloon release at 7 p.m., March 2 shows that air in the lowest 8,000 feet was above freezing through much of the layer.   This layer was also saturated, as steady rain was falling.

Figure 4. Weather balloon-derived temperature (red) and humidity (green) vertical profiles at (1) 7 PM March 2 and (2) 7 AM March 3.  The lowest 10,000 feet of the second sounding reflects chilling (adapted from NOAA SPC).

Around 3-4 a.m., the core of the arctic air surged through our region.   Observations at the three area airports show about an hour’s worth of light freezing rain and/or sleet, then a permanent transition to all snow.  The rapid rate of sustained temperature decline and near-lack of icing underscore the steep slope of this particular arctic air mass.

With temperatures plunging into the mid-upper teens, the snow to liquid ratio became at least 15:1, if not 20:1, along and north of the metro region.

The heavy precipitation band

Finally, this storm came complete with a heavy snow band.  For 24-36 hours prior to the main event, forecast models began to trend southward with a heavy snow or mixed precipitation band, predicting that it would set up immediately south of the D.C. Metro region.

This was a good prognostication, as evident in the morning radar observations (Figure 5).    A fairly narrow band sagged slowly southward across northern, then central Virginia during the early morning.   Additional features of note include (1) the depression located over the Smokies;  and (2) cold air damming (“the wedge”) created when arctic air surged south and east of the Appalachians.

Figure 5. This more focused analysis, on the mesoscale, reveals details such as a heavy snowband (standard radar overlay) and Appalachian cold air damming (purple arrow).  Time is 9 a.m., March 3 (adapted from NOAA SPC).

In past CWG stories, winter winter expert Wes Junker and myself have stressed the importance of frontogenesis in establishing these snow bands.   Frontogenesis means, quite literally, creation of a frontal zone.   On March 3, the surge of arctic air plowing into milder air over southern Virginia rapidly tightened the thermal gradient aloft.

Mild, humid air rises more quickly along a convergent frontal zone, enhancing the production of heavy precipitation.   This process was well underway during the morning of March 3, as shown in Figure 6.   This computer-generated graphic shows where frontogenesis was occurring (tightly packed purple contours) at 5,000 feet and 7 a.m.   The heavy precipitation band lies dead-center beneath this pocket.  This snapshot does not convey the dynamic nature of the process;  an animation of this analysis (not shown) revealed how the small, intense zone of frontogenesis moved southward through the morning, taking the snowband with it.

Figure 6. This is a mesoanalysis at 9 a.m., March 3 made at 5,000 feet.  A pocket of intense frontogenesis (purple contours) coincides perfectly with the storm’s principal snowband (adapted from NOAA SPC).

Polarized doppler radar reveals a precipitation transition

In the past two years, NWS Sterling has equipped its Doppler radar for polarization diversity.   Polarization is a complex subject and merits its own article.  But in a nutshell, the radar now transmits and receives both horizontally- and vertically-oriented microwave pulses.  Precipitation particles come in a variety of shapes, ranging from spherical to conical to flat.   The ratio of horizontal to vertical energy received by the radar can thus be used to infer the type of hydrometeor target.

When conditions are right, polarized radar can detect a rain-snow transition line, using a parameter called the correlation coefficient.   Correlation coefficient compares how similar the power returns for the vertically and horizontally polarized signals are.  A value of “1” means the signals are identical.  The lower the correlation coefficient, the less similar.  A mix of oblate raindrops, small spherical ice pellets and flat snowflakes can create this dissimilarity.

In Figure 7 (below) we see values of correlation coefficient in the widespread precipitation at 12:30 a.m., March 3.  In this image, the rain-snow line is a discernable, yet subtle feature, as shown by the arrows.   The transition from rain to ice to snow occurred along a very narrow line, oriented west-east, across Frederick and Howard Counties.   Within this band, correlation coefficient values fell to about 0.85 (burgundy shade), compared to nearly 1.0 across the rest of the region (magenta shade).

This rain-snow transition band was positioned just ahead of the core of coldest, low-level air, as it began to surge south of the Mason-Dixon Line.

Figure 7. Weather radar interpretation of the rain-snow transition zone through the lens of dual polarization.   Time is 12:30 AM, March 3.  Arrows identify a subtle band of mixed rain, snow and sleet using a new parameter called correlation coefficient (adapted from weathertap.com).

Summing Up: It’s all a matter of scale

From the foregoing discussion, there’s a lot to consider even in a moderate-impact snowstorm.    We began our discussion with the large scales (synoptic scale), moving down through mesoscale (i.e. the heavy snowband) and ended up discussing microscale attributes of polarimetric radar measurements.  The scope of this discussion reflects a broad range of spatial scale, from 1000 km down to micrometers, or nearly nine orders of magnitude!

The meteorology of these events, although appearing complex at the outset, is often best understood by adopting this multi-scale approach.   As we dissect the storm, we proceed down a funnel of understanding.   Meteorologists take separate courses focusing on each scale aspect (e.g. synoptic meteorology, mesoscale meteorology and cloud physics) – the rubric of “stovepipe science”.   The trick, in the real world, is to understand how these disparate scales interact.

In and of itself, each snowflake is a delicate, microscopically-formed thing.  But when they gang up in large numbers, look out!