From the ABC 7 Weather team

Washington Snow Storm: Really "Wrong"? Why?

March 7, 2013 - 12:01 AM
Text size Decrease Increase

OK I and others busted the "big" snow forecast for D.C.  Close but why no cigar? How could this happen?

It's true - the forecast for 3 to 6 inches of wet snow in D.C. itself was a bust.  The District got some snow, but only enough for a brief white coating until quickly turning to rain.

In fact, the more than 1 inch of rain would have made for 5 to 8 inches of heavy, wet snow. So, why was the forecast for possible "major" storm wrong?

Well, 1) It was and is a major storm. Look at the snow accumulations. Here's a great picture form Adam Caskey of the ground and snow from the Iwo Jima Memorial (near sea level) and 4 miles west (about 200 feet higher).

ZZZZZ

Elevation does, and can make a big difference in snow forecasts.  We know  the air near the ground was warm after the strong March sunshine of Tuesday. 

The forecasting tools we now use are largely based on "Numerical Weather Prediction." But for us "oldtimers," we also rely on experience.  Most of the signals I saw told me this would be a "dynamic" storm. 

Rapid development in just the "right" spot in the lower Chesapeake Bay could have drawn in cold air, and with rapid rising air, it should have overcome the low mild air and changed early rain to snow. 

One of the "what could go wrong" parts was the strong east winds shown on these simulations.

ZZZZZ
ZZZZZ

The prediction during the Tuesday evening forecast for the Wednesday morning rush hour and snow amounts was right on!

ZZZZZ

I should have kept that all day.  But I also remembered the famous Veteran's Day storm of November 1987. Several days before that storm it was 70°, but the "dynamics" in the atmosphere made for a historic snow, even if the calendar said November.

I knew the rain/snow line would be close. The late Tuesday vertical profiles of the air above us for Wednesday showed enough cold air for snow.  But that low level east wind was there.

ZZZZZ

Our local hyperlocal microcast model/simulation was the "outlier" from everything else, but proved to be reasonably correct about the snowfall.

ZZZZZ

The storm formed in the "perfect" area for a big snow around D.C., which is in the lower part of the Bay.

The only other "weak link" was the lack of real cold air to our north. I thought the snow forming above us (all the rain that fell was snow until 1,000 to 2,000 feet up) would be intense enough to "beat"  the mild air near the surface. Even with at mild east wind,  the morning rain around D.C. should have changed to snow. 

Wrong. 

But look at the snowfall reports in the area:

ZZZZZ

So would I make the same forecast again?  Probably not. That east wind did turn to the northeast and heavy snows continued not far away. 

But I sure will look for colder air to our north and try harder next time.

Apologies?  No. I did my best just like you do. We'll edo more homework and study a bit more.

Here is a quick look at our hyperlocal simulation that for several days was showing more rain than snow. It proved pretty good. I also explain why this "outlier" was hard to go with.

Short URL: http://wj.la/XX01Bd
comments powered by Disqus