Most Read: Local

Posted at 06:48 PM ET, 11/01/2011

How well did we predict Snowtober?


CWG’s snowfall forecast (top) issued Friday morning versus actual snowfall amounts around the region (bottom, source: National Weather Service)
When we planned the release of our winter outlook for the first week of November, little did we know the season’s first winter storm would occur before then! To sum up our forecasting performance for Snowtober (October 29), we were overly skeptical of a snow event prior to two days before the storm struck, but did a very good job of forecasting the storm within 48 hours.

The above comparison of our accumulation forecast (map) and how much snow actually fell demonstrates how accurately we predicted snow amounts starting Friday morning.

Let’s take a look back at the evolution of our forecast...

When models first hinted at the threat of snow more than a week before Snowtober struck, we were dismissive. Even as recently as the Tuesday before the storm, when one of the models suggested the possibility, we were skeptical. Wes Junker, our winter weather expert concluded at the time:

Right now — even though the new European model coming in continues to show a wintry Saturday morning — it looks like the European is long shot at best.

The European model turned out to be quite right.

On Wednesday, we came a little closer to buying into the snow possibility, but not quite. I wrote:

The latest computer models continue to suggest the odds of seeing snow Saturday morning in the Washington metro region are small, but not non-existent especially in outlying western areas. A cold, light rain remains the most likely scenario and it’s still possible much of the precipitation misses to our east.

By Thursday, we had the general idea right. Wes Junker wrote:

The odds of a highly unusual, though not unheard of, period of October snow have increased for northern and western parts of the D.C. region on Saturday.

As I said up front, I think that once we got to Friday, we pretty much had it nailed - which is an achievement given the complexity of the storm. Here’s the forecast summary I posted with our accumulation map, timeline, and answers to frequent questions Friday morning (which were all more or less on target):

A rapidly developing, intense East Coast storm will bring unseasonably cold temperatures and heavy precipitation to the entire metro region Friday night and Saturday. For the immediate metro region, rain is likely to be the predominant precipitation type, but snow may begin to mix with the rain late Saturday morning and possibly change to all snow briefly during the afternoon. Accumulation is unlikely but cannot be ruled out.

Towards western Montgomery, Frederick and Loudoun counties, a changeover to snow is more likely to occur and occur earlier, with the potential for some accumulation, especially at higher elevations.

We did miss a couple minor details, however:

1) We did not discuss a transition to sleet in between rain and snow. Most areas got some sleet.
2) We didn’t mention that precipitation type would sometimes depend on intensity - i.e. when it precipitated hard enough, it fell in the form of sleet and snow, but would switch back to rain when the intensity waned.


A satellite image showing snow cover on Sunday from NASA.
Based on our difficulties in the extended range forecast, but success with the more important short term forecast, overall, I would grade our effort a B.

What do you think?

By  |  06:48 PM ET, 11/01/2011

Categories:  Latest, Winter Storms, Weather Checker

 
Read what others are saying
     

    © 2011 The Washington Post Company