Article: Snow Lovers Share Frustration on Chat Board

Millions of schoolchildren along the I-95 corridor from Washington to New York went to sleep Thursday night with visions of a snow day, or at least a late opening, swirling in their heads.

But well before bedtime, savvy weather watchers knew the sad truth.

At 8 p.m., “TerpEast” had posted it on Easternuswx, the hot chat board for a gathering assortment of weather enthusiasts, from seasoned professionals to so-called snow weenies.

The storm was going to fizzle, wrote Nabeel Keblawi, a Maryland meteorologist who posts daily as “TerpEast.” His technical treatise ranged from the North Atlantic Oscillation to a misplaced high-pressure system, and led to an inescapable conclusion: Here comes another another dud in a deadly dull winter.


Why the Feb 1-2, 2007 East Coast Snowstorm Failed to Materialize

Brief Summary:

– No negative North Atlantic Oscillation (NAO), no blocking, no 50/50 low
– Upper level flow too flat and fast
– Northern stream s/w didn’t “pick up” the subtropical jet wave, instead went towards the east
– Lack of real phasing
– Cold air was there, but no high to the north, and thus eroded by the first wave
– Storm “split” into 2 frontal waves without a defined cyclone
– Dry slot between 2 waves, convection robs moisture

Here are a few thoughts on what happened with this snowstorm, and why my forecast was a bust for Washington, DC… but may still verify for North Carolina and Southern New England.

Five Things That Went Wrong:

1) Following radar trends, it looks like the precipitation between Texas and Georgia are falling apart, and northward progression too slow to start the precipitation here on time.

2) A line of strong convection in the immediate Gulf shore robbed northward moisture transport. Diabatic (latent) heating due to that convection interfered with northward isentropic lift and moisture transport towards the Mid-Atlantic, leaving us high and dry.

3) Against my forecast, this system did develop a dual wave system. I was correct in that there would not be dual low pressure centers, however I did not expect a dual wave along the stationary polar front. The first wave produced some good precipitation in the North Carolina mountains and piedmont – and my forecast for them pretty much verified nicely. That was expected. What was not expected was that there was a large dry slot between the first and second waves, killing any chance of my forecast verifying north of the Charleston-Fredericksburg-Dover line.

4) Speaking of this aforementioned dry slot, the second wave doesn’t appear to be able to overcome that dry air, and strong convection to the south is not helping either. While the strong convection may help organize a surface low in the south and have it move northeast along the east coast, it’s not enough to bring enough moisture to the Mid-Atlantic. For Southern New England? Barely enough for my low-end forecast to verify. New York City isn’t going to make it to my predicted snowfall amounts, I don’t think.

5) With the first wave making passage towards the east, it brought warm air aloft to the Mid-Atlantic but not enough lifting to produce snow. And when the second wave passes by, any precipitation that may fall will be mixed with little or no accumulation because the air aloft had been warmed by the first wave.

Five Previous Warning Signs That This Would Go Wrong:

1) The northern stream shortwave was programmed by models a few days ago to dive south-southeast or almost due south to pick up the subtropical jet stream wave and take it up the east coast as one consolidated low pressure system. This was the warmest and wetter solution depicted by models. But the warning sign I saw that told me it wouldn’t quite pan out this way – I discussed this in my initial forecast. Using the satellite wind data, I saw that the mid-upper level winds associated with the arctic jet wave was more westerly than northerly. That was the thing that tipped me off and I knew something was wrong. That was the reason I went against the model consensus at that time that the storm would do an inland run and give the Ohio Valley, northern Mid-Atlantic, and Southern New England significant snows. I knew it wasn’t going to work out this way, so I stuck with the more suppressed coastal track, and I was correct with that. And when the models finally caught on to the northern s/w going more to the east than to the south, they suddenly suppressed the track – much to the surprise and chagrin of this board.

2) The models, especially the NAM, had a difficult time resolving the subtropical jet wave as it was going over the southern Rockies at the US-Mexico border. I noticed that the 500 mb height contours were squiggly in those runs, and I knew something was screwy about these model runs. I still don’t know what it was. Maybe it was gravity waves? Maybe it was the model’s representation of the mountains messing up the flow? Still don’t know, but I will be wary from now on should a feature like this show up again.

3) I should’ve paid more attention to the jet structure of this storm. The jet streak was directly over the southern Mid-Atlantic, with the entrance over the Gulf. I was looking to see where the left quadrant of the jet exit would be, but I did not see it anywhere near where the low was forecast to develop off the coast. The flow was flat and very fast out to sea. This should’ve tipped me off to the first wave passing out to sea rather than developing off the coast and tracking northeast to within the benchmark towards Southern New England. That was one thing I missed – not enough upper level support for a decent storm here.

4) Another thing I should have addressed was that the subtropical jet wave associated with this system were sheared and ripped apart by the time it got to the lower Plains. It looked good going from Colorado to Missouri, but then it fell apart due to the lack of upper air support. Should’ve noticed that in previous runs the shortwave became almost imperceptible over the east on the H5 charts compared to when they were in the west.

5) My knowledge of model biases is very limited, so it would probably help if I learned more about them to make an educated guess whenever the models totally disagree or are all over the map like they were this week.