Snow total reports from NWS | Snowfall total map *

The St. Patrick’s Day Storm poured “salt in the wound” of a harsh winter seemingly without end. In this story we review ten aspects that conspired to create a major snowstorm across the Mid Atlantic.

1. This was a classic “overrunning” type snow event. Like the March 3 snowstorm, the setup was similar: Abundant Gulf and Atlantic moisture overrode a wedge of arctic air trapped against the mountains, as a wave of low pressure approached from the southwest. Figure 1 below highlights the main “synoptic scale” (large scale) elements contributing to this snowy recipe.

Figure 1. Surface weather analysis at 1 AM March 17 while moderate-heavy snow was underway across our region. Adapted from NOAA.

2. An intense cold air damming episode became established during the event. During the winter months, the cold “wedge” almost always spells trouble, even in mid-March. In the figure above, this feature is very apparent. The graphic below shows the thermal pattern (isotherms) that define this wedge at 5,000 feet. Note the “ridge” of extremely cold air spilling down the Piedmont from the north, blocked by the rampart of the Appalachians.

Figure 2. Temperature analysis at 5,000 ft, 5 AM March 17. Sub-freezing isotherms are shown by blue dashed lines. Adapted from NOAA.

3. The pattern’s evolution resembled a “Miller Type B” setup. This is textbook meteorology, but we have been discussing these types of patterns all winter long at CWG. Figure 1 (above) shows that the main region of overrunning precipitation was created by a low pressure system over the Mid-South, but the pattern transitioned to a developing coastal low. Instead of tracking up the East Coast, the coastal low followed a “flat” trajectory, carrying it out to sea. This likely spared the region of significant, additional snow accumulation.

4. Upper air dynamics played a secondary role. The pattern of “split flow” between northern and southern branches of the jet stream lacked timely phasing that is often the hallmark of a blockbuster snowstorm. Additionally, the upper level wave supporting the storm over the Mid South was not that impressive; in fact it was “shearing out” and weakening. However, there was a classic, coupled jet streak circulation over our region. As explained in my article on the February 12-13 snowstorm, this generates vigorous uplift over a more focused region. The enhanced region of dynamic uplift is shown in the upper-air analysis chart below.

Figure 3. Jet-stream level flow analysis at 8 PM March 16. Note that the two pockets of fast-moving air, called jet streaks (pockets of yellow-orange), merge right over Washington. Adapted from Unisys Corp.

5. A sustainable supply of sub-freezing air fed the storm. This element is often lacking in many potential D.C.-region snowstorms, but as this past winter has demonstrated, arctic air has been plentiful. Although not shown in Figure 1, a strong anticyclone was positioned along the northern tier of the eastern U.S. A lobe of this arctic air mass became funneled down the eastern slopes of the Appalachians as a cold air wedge. The upper-air sounding from Dulles, shown below, reveals that the layer of sub-freezing air was quite deep.

Figure 4. Balloon sounding (temperature profile) obtained from Dulles Airport, 8 AM March 17. Shown is a deep sub-freezing air layer, from 10,000 feet down…with exceptionally cold air a few thousand feet above the surface. Adapted from NWS.

6. Early onset of snow helped boost total accumulations. The precipitation lacked an early transition from rain to snow. Most meteorologists in our region were caught off-guard by this aspect. The culprit: Very dry air above the surface evaporated the rain, which began falling mid-afternoon. Evaporation extracts heat from air, and this quickly chilled a deep layer of air beneath the cloud to below freezing. Precipitation began as snow, then quickly intensified. However, meteorologists correctly identified warm ground temps as an early mitigating factor in “stickage”.

7. Snowbands. In a word, these small-scale (“mesoscale”) features have been the bane of accurate snow forecasts all winter long. D.C.-region meteorologists struggled to identify a priori if and where these dynamic features would set up. The models were giving mixed signals as to their location. Trying to outwit a snowband is the “holy grail” of heavy snow forecasting, because their intense snow rates, up to several inches per hour, spell major bust potential for a forecast.

This particular storm created a succession of exceptionally narrow bands, along the northern periphery of the storm system. Snow rates at times produced near-whiteout conditions, with visibility at Washington Reagan plunging to less than 1/8 mile.

The figure below is a radar screen shot of these bands in action. They took the form of elongated, narrow ribbons of heavy snow moving from southwest to northeast. A succession of these bands developed throughout the duration of the event. This is a somewhat different model than the “classic” snowband, which often remains stationary for hours on end.

Figure 5. Radar at midnight, March 17 showing two parallel snowbands (elongated, dark green shaded regions) producing moderate-intensity snow. Arrows are placed parallel to and just above each band. Adapted from Weathertap.

8. Frontogenesis helped sustain snowbands. OK, more textbook meteorology here, and this is a technical term that I, Wes Junker and other CWG staffers have bandied around this winter. Frontogenesis literally means “creation of a weather front” and this process often plays out in the middle atmosphere, on the northwest side of a rapidly deepening low pressure system. However, a detailed analysis on this event suggests that while present, its effects were not as pronounced as in other 2013-2014 winter events. The reason for this probably relates to transition between low pressure systems. As the system over the Mid-South was weakening, the coastal low was taking over. This created disorganized flow in the mid-levels, not conducive to a persistent, strong pocket of frontogenesis – nor a single, intense snowband anchored in place.

9. Snow:rain ratios were increasing throughout the event, upping the accumulation as snow rates intensified. I make a big deal out of these ratios, and they are important to the snow impact. The most common ratio is 10:1, and this is what all the models all “assume” for their snow depth prognostication. However, ratios not only vary widely across a snowfall region, but are also dynamic, changing character throughout the course of a storm.

Early in this event, ratios were near 8:1, on the wetter side of the spectrum. While it was below freezing a few thousand feet above the surface, the residual warmth of the afternoon kept surface temps above freezing during the first few hours. There was partial melting of flakes with a thin liquid coating. The snow was packable.

Around sunset, as the deeper wedge of cold air became established, surface temps plummeted below freezing. Without partial melting, snow ratios shifted toward the drier end of the spectrum, surpassing 10:1 then 12:1 and even higher.

10. Luck of the Irish. I tried to think of a tenth unique aspect to this storm, and came up short. But this should suffice as a place-holder. Perhaps readers of the CWG blog can “fill in the blank” here – both technical and humorous entries are encouraged!

 RELATED: Relentless snow prompts DC weather haiku