Snow covers cherry blossoms around the Washington Monument as a snowstorm passes through Washington on Wednesday. (Harrison Jones via Flickr)
The proper atmospheric ingredients finally conspired to give Washington a moderate, late-March snow after a snow drought that lasted our entire winter. The models took us on a roller-coaster ride in the days prior, but observations collected during the storm now let us piece together the storm’s evolution.
Big storms such as this always have their roots in the upper-level river of winds, called the westerly jet stream. Giant, wavelike meanders ripple through the jet, distorting it into undulating troughs and ridges that migrate eastward. Where the air exits a trough (southward dip in the jet), air is forced to rise from below and replace it.
The figure below shows a large amplitude trough over the eastern United States at 8 a.m. Wednesday morning. The pink tones in the base of the trough indicate winds blowing at nearly 180 mph.
The red circle shows the dynamic “sweet spot” where air rises most vigorously, exiting the trough. This lowers the pressure at the surface, helping to birth a type of cyclone called a nor’easter. At 8 a.m., this storm was taking shape off the Virginia Capes and intensifying.
The next diagram illustrates the coastal low-pressure center (red L), surrounded by a concentric pattern of isobars, lines of constant pressure. At 8 a.m., sea level pressure in the storm was dropping below 994 millibars.
Note the patchy region of precipitation — mostly moderate snow — falling to the northwest of the nor’easter’s center (blue and gray shades, derived from weather radar). Also, note the second region of low pressure crossing Appalachia; this is, in fact, the primary area of low pressure that was in the process of “handing off” its energy, across the mountains, to the coastal storm.
As the coastal storm strengthened and pulled away, the primary low-pressure center dissipated. These types of handoffs often confound our snow forecasting, as the models may not agree on exactly where the new coastal will form, how strong it will be, and where it will track.
So we’ve established a zone of rising air over Washington, into which moisture streams off the Atlantic, courtesy of easterly winds circulating around the nor’easter. The last ingredient we need for snow is the most critical: cold air, and not just a fleeting supply. For this, we turn to the best possible source around here — a cold high-pressure cell to our north. Its northerly winds push a subfreezing wedge of air southward, damming the cold up against the eastern slopes of the mountains. Appropriately, we call this Appalachian cold air damming.
The figure below shows this frigid wedge of air at two different levels. We want the air in the lowest 5,000 feet to be below freezing for snow to form in cloud layers and survive all the way to the surface.
The left panel of the above image is at an altitude of 3,000 feet at 10 a.m. Tuesday. The air here was plenty cold, 25 degrees Fahrenheit (minus-4 Celsius) right over D.C.
The right panel of the image shows a pool of air chillier than 32 degrees at the surface and partly dammed up against the terrain (green shades). Locations immediately west and north of the District were in the 30-31 degrees F range. The dotted purple line in the figure represents the boundary of between above-freezing and below-freezing air.
Finally, the moderate to heavy snow in this storm was most persistent just east of Interstate 95 and across our northern regions. These heavy zones can be seen in the snow accumulation map.
Significant differences in amounts over short distances can happen when a persistent (stationary) snow band sets up to the west and northwest of the coastal low’s center.
Trying to forecast banding features more than about a day in advance is nearly impossible, given the present state of the prediction models. Instead, we rely on “nowcasting,” where we let radar trends over the past one to three hours guide our thinking with regard to heavy band placement, intensity and movement. The highest resolution forecast models such as the HRRR can also provide some guidance, although there is no guarantee that these models will adequately capture a particular band’s evolution.
