We’ve received a lot of thoughtful comments in response to my post questioning our local (Sterling, Va.) National Weather Service (NWS) office’s decision to issue a winter storm watch Wednesday for the potential for at least 5 inches of snow Thursday (when no snow fell).

A few commenters thought I came down too hard on the NWS.

I understand that perspective and was conflicted in criticizing the office as I have in­cred­ibly high respect for the work it does. The storm was extremely challenging to forecast, the most difficult winter weather situation I can recall since the Boxing Day (Dec. 26) storm in 2010. And I want to emphasize that I know NWS had the best intentions in issuing the watch - to alert the public and decisionmakers about a potentially hazardous situation.

But my main point remains I think NWS acted too quickly and decisively on very ambiguous data and that ended up resulting in confusion and compromised decisions.

James Lee, the meteorologist in charge at the Sterling NWS office, sent me an email describing the rationale for the watch. While I disagree with his described interpretation of the model data (which informed the watch decision), I’ll give him the last word (email republished with permission)...


I appreciate your openess, and asking for our rationale in the watch.

As [warning coordination meteorologist] Chris [Strong] said, our watch criteria is 50% chance of occurence. So half the time we issue a watch, by definition, we do not expect to meet warning criteria. The 50% watch policy helps transportation planners, emergency managers, and the public to prepare for a potential winter weather hazard; particularly the transportation planners who need to staff plows and put blades on trucks. In my 15 years of winter forecasting here and in Boston, I’ve seen 4” snowstorms with no watch draw metro areas to a standstill, whereas 18-24” snowfall with a 36-48 hour watch be transportation non-issues.

In a society that responds well to winter weather hazards, it’s all about preparation and planning. It is easy to get the purpose of a watch confused with the purpose of warnings, with the goal of 90% accuracy in our warnings. For the record, our office has produced winter storm warnings at or above 90% accuracy in 7 of the 8 past winters.

The model runs leading to watch issuance showed us a trend upward. All of these forecasts are for KDCA [Reagan National Airport]:

For the GFS, 16 Jan 00Z 2.1”, 16 Jan 06Z 4.5”.

For the NAM, 16 Jan 00Z 5.8”, 16 Jan 06Z 4.3”, 16 Jan 12Z 6.1”.

For the SREF, 16 Jan 03Z mean 1.9” with range 0-23”, 16 Jan 09Z mean 5.0” with range 0-24”.

Our watch was issued at 11:23 AM EST on 16 Jan, after the 16 Jan 12Z NAM run came in with 6.1” for KDCA. We knew we had to hustle to get the watch out quickly as it was approximately 24 hours before the first flakes were to fall.

I believe that we would have been negligent of our duty in not issuing a watch Wednesday morning, given the facts of the model trends and the watch policy of the NWS.

Again, I appreciate you giving us an opportunity to comment.


Jim Lee