Thursday, August 15, 2013

Historical Case Study of February 23, 1987 Winter Storm

Although many snow-lovers will say that the winters of the 1980s and early 1990s were mostly boring, the winter of 1986-87 featured several significant snow storms in the Mid-Atlantic, including this one in late February.

But first, a look at the pattern that produced this storm. (And another passive aggressive note about incomplete data sets: the only level above the surface that I was able to produce a map for was 250 mb, and that data only existed for 0000 UTC on 22 Feb, almost 24 hours before the worst of the storm in Philadelphia.)

250 mb map:
Ignore the "12 UTC" in the caption.
You can see a very broad trough encompassing the western half of the Continental United States, with a closed upper Low over northern California, and two more shortwave troughs embedded within the flow: one over Minnesota, and the other over New Mexico.

Surface:
The shortwave over Minnesota was associated with a weak surface Low over the upper Mississippi Valley. A cold front trailed southwestward from there. Another surface Low was located over central California, associated with the deep upper Low.

Although not pictured here, the shortwave over New Mexico moved into the Southeast, and initiated surface cyclogenesis over the Gulf Coast (Kocin and Uccellini, 558). The northern and southern shortwaves phased over the Mid-Atlantic during the afternoon on the 23rd, by which point the surface Low had already pulled away from the coast and finished dropping snow on Philadelphia.

In this storm, heavy snowfall was not widespread; the highest totals of 15 to 20 inches (with a very few outliers) existed in a narrow east-to-west band across the northern and western suburbs of Philadelphia, roughly paralleling the PA Turnpike. Some snowfall totals from this region include:

  Southampton, PA - 17.0" accumulation
  Willow Grove, PA - 19" snow depth (actual accumulation, accounting for compacting, probably exceeded 20")
  Norristown, PA - 16.0" accumulation
  Conshohocken, PA - 16.0" accumulation
  West Chester, PA - 17.5" accumulation
  Glenmoore, PA - 18.0" accumulation
  Honey Brook, PA - 16.0" accumulation
  Coatesville, PA - 23.5" accumulation

Lesser totals were found south of that band...

  Indian Mils, NJ - 11" snow depth
  Wilmington, DE - 13" snow depth
  Philadelphia International Airport, PA - 6.8" accumulation

...as well as north of it.

  Doylestown, PA - 10" snow depth
  Lambertville, NJ - 8.0" accumulation

What was interesting about this storm was not just the snowfall amounts, but the rate at which snow must have fallen to get to those amounts. Light snow had only started a little before 10 p.m., and by 10 a.m. the next morning, most stations were already reporting partly cloudy skies. That means that some places saw more than 20 inches of snow accumulate in less than 12 hours.

Kocin and Uccellini, in their Index of Northeast Snowstorms, also note about this storm:
An important aspect of this case is the widely varying operational numerical model forecasts of the location and intensity of cyclogenesis. Despite the inconsistency of the model forecasts and the unusually warm temperatures, NWS forecasters (e.g., David Caldwell) made accurate predictions and related warnings of the timing and amount of the snowfall.
Both are valid points. Philadelphia International hit 50 F that afternoon after snow ended. As an example, Conshohocken's 16" of snow melted and compacted to a snow depth of just 7" the following day. And the surface cyclone did indeed deepen quite rapidly, as you'll see below.

Hour-by-hour:

7:00 p.m.
10:00 p.m.
1:00 a.m.
4:00 a.m.
7:00 a.m.
10:00 a.m.
1:00 p.m.
A pre-existing tight temperature gradient along the Southeast coast likely aided in the cyclone's rapid development.

What you don't see in all of these maps in a surface High in Canada. A strong anticyclone north of the region is typically a key element in northeast winter storms because of the cold northerly and northeasterly flow it produces ahead of the storm.

The temperature and pressure contours in the first few images also show some Cold Air Damming east of the Appalachians, which allowed for so much snow to fall despite the otherwise warm temperatures and lack of a true cold air source.