Tuesday, February 11, 2014
What Went Wrong, What Went Wrong, What The Heck Went Wrong.
As I said on my facebook page, nobody takes missing a forecast harder than I do. No snow forecast is going to be absolutely perfect and this one was nowhere close to being right and for that, I'M VERY SORRY. There are many lessons learned this week which will be used in the future.
So what happened? Where do I begin with that question? For quite awhile, all the guidance we use to formulate a forecast pointed towards a significant winter weather event for much of our viewing area. It's hard to ignore when that data says several inches of snow. The National Weather Service office in North Little Rock issued a winter storm watch too and I completely agreed with that given the information at the time.
If you look at previous blog posts, I talked about the models having troubles resolving each wave of energy coming off the Pacific Ocean. Would it comes out in separate waves or maybe come together with another piece of energy to combine for a bigger storm?
Models are only as good as the data you put into them. These small pieces of energy flying over open water can only be sampled using satellites so all the parameters are estimated without true measurements. This adds uncertainty. Once these systems move ashore, they can get better sampled using weather balloons launched twice a day at weather service offices around the country. Theoretically, this is when the forecast should improve.
Let's go back to last Sunday. The data all of the sudden started to shift indicating less moisture and changing the timing as well. The swath of snow with a disturbance we were expecting across west central through central and east central Arkansas looked less and less impressive. One of the models, NAM, even went so far as to show NOTHING. You might remember my Sunday night broadcast when I said that winter storm watch may get "adjusted" and I mentioned a great deal of uncertainty has entered the forecast after the modeling indicated some agreement in previous runs. This is a higher resolution short term model (NAM) which has not performed well lately, but it turned out to do a better job with this situation. That piece of upper level energy was clearly seen Monday night dumping significant amounts of snow across Kansas and Oklahoma and was traveling towards Arkansas. However, the guidance insisted it would fade and indeed it did. What's left of that is over northern Arkansas this morning with very light snow flurries.
The models then started to key in on another wave of moisture Tuesday afternoon and night over southern Arkansas and that still has the possibility to bring some icing there, but this will really increase east of the state. Those two waves of moisture I talked about above will come together and produce a very significant eastern United States winter storm over the next few days. So if you have travel plans, beware of that. You're going to see a ton of news reports out of that region. Atlanta, GA could end up with another big storm!
Once again, I'm very sorry. That's the nature of weather forecasting. We have had a very good run so far this winter forecasting the onslaught of cold, snow, and ice. We got this one storm (or lack of) wrong. "Onward through the fog" as my mentor always told me. It's time to get back to work and move on. Thanks!
Posted by Todd Yakoubian on Tuesday, February 11, 2014