2.11.13 Nor'Easter Analysis: How We Did - FOX44 - Burlington / Plattsburgh News, Weather & Sports

Alright, storm analysis, here we go.

Overall I say we did 'okay' in forecasting this storm. In my opinion I wouldn't call it a hit, but I wouldn't call it a miss either. Here's what I mean by this.

When bigger than average storms coming to our area yielding more of an impact than just a dusting to a few inches of snow, I typically do a pre and post analysis of the storm. I write down what my predictions are, what the National Weather Service's forecast numbers are, what the numerical weather models say, and then go back and see how everything panned out and where I can improve.

So let's start with happened with this storm and why it was a difficult one to precisely predict.

We had two systems merging together. See left for the pictures which you can click on and increase the size. A low moving east from the Great Lakes combined with a coastal low pressure system moving northeast from the Carolinas to Cape Cod. This was a tricky forecast because we had an initial burst of snow Friday morning with system #1, then snow in our southeastern forecast area from system # 2, which ultimately became the nor'easter as both systems merged.

Leading up to the event two major weather models were very different in terms of how much liquid precip would be coming to our area. For instance, just two days from the event the NAM predicted much more precipitation than the GFS. Eventually they came together on a similar amount which was a half inch to one inch of precipitation for *most* areas. Ultimately the GFS won this one. It has twice as many verified predictions as the NAM when looking back at their individual predictions for our area.

Well how did we do? I say 'okay' because there were some bulls eye hits in areas but others that completely missed. The Northeast Kingdom's picture didn't come into a clearer view until just before the storm. We had forecast a widespread 8-14" accumulation but then knocked it back to 4-9" and 4-8" the day before and day of. Those latter predictions were much better as the NEK mostly got 4-8", though some Caledonia County locations received up to 10-11 inches.

The following data is for my forecast made Thursday morning 02/07 at about 9 a.m., roughly 18 hours before the beginning of the event and 54 hours prior to the storm's ending.

When looking at my forecast of 8-14" for most places, 62% of towns within the forecast area fell somewhere between 8" and 14". Worst forecast areas were Orleans, Essex, Caledonia, and Addison counties. As I  mentioned earlier the Northeast Kingdom forecast did come down the day prior and day of which helped those towns. Best forecasts for areas go to Essex, Clinton, and Franklin counties in New York, as well as Orange, Washington, and Lamoille counties in Vermont.

I verified two ways. I counted number of towns falling within forecast range by county
                                                _____________________________________________
                                                 number of total towns reporting snow by same county   

The other way was a pinpoint forecast. I wrote down 12 towns and cities and put an exact number on it. For instance I put Montpelier at 13". Montpelier reported 8.5". That's a 65% accuracy on my end. I wrote my numbers down and the NWS Burlington's from Thursday 02/07.

I'm happier to say here that the pinpoint forecast did pretty well. Overall the accuracy was 73% which is pretty good with snowfall total. I know it's not perfect, but I've seen these %'s as low as 20-30. The NWS matched at 73% on their Thursday morning forecast for Friday. That forecast changed multiple times through the event, which I didn't verify those #'s.

Areas which got much less than what I thought were Bennington and Rutland areas. Per my verification I only grabbed a 25% and 20% accuracy respectively. That was an area (which actually stretches north to Addison county) where snow got blocked by the mountains and there was a lull in overall precipitation.

Well onto the next storm system! Hopefully this blog wasn't too confusing for you, but I like to do them. Why? Well because I hear it all too much. "You have the best job. You can be wrong all the time and still get paid." Technically true. But here's data to back me up when asked, 'How accurate are you?'

-Meteorologist Steve Glazier

Comment Agreement