As we know by now, this storm wasn't extremely well predicted for people in parts of New England. I think it's very important to go back and review the mistakes that were made in the forecast process, so these mistakes aren't made again. Meteorology is an inexact science - and always will be - but I still think it's important to go back, and review. Especially after a busted forecast. I've gone back and looked at each snowfall forecast I made for my personal weather web site - www.atkinsonweather.com - focusing on the Merrimack Valley of southern New Hampshire and northern Massachusetts. Unbelievably (or...maybe believably in this case!), I modified my snowfall forecast five times in between Monday afternoon and today. Here are some of the notes that I made when going back, and looking at my previous forecasts.
* Monday afternoon: There were indications we'd see a healthy dose of snowfall around here, on many of the computer models. My first forecast, of 4-8" in the Merrimack Valley and southern New Hampshire, was issued.* Early Tuesday morning: Overnight Monday night, the models backed off quite a bit. They indicated a more southern axis to the heaviest snow. I became much less confident that we'd see a half foot or more. I downgraded the forecast to 2-5", a pretty wide range, but the uncertainty that existed justified it, in my opinion.
* Tuesday evening: The switch was flipped back on later Tuesday. A big shift in the models occurred; some were even indicating up to 1 foot of snow in the Atkinson area! I modified my forecast again, returning to 4-8".
* Early Wednesday morning: Incredible uncertainty still existed Wednesday morning, when updating around 5:30am. One computer model (the NAM) predicted only an inch or two of snow for the Atkinson area. At the same time another model (the GFS) had about 6" for the Atkinson area. With so much zigging and zagging of the computer models, I decided to stick with what I had Tuesday evening. Radar trends looked good early Wednesday morning, with lots of moisture to be had. The question was - where would this moisture set up - over our area, or south of here? I only made minor adjustments to the map - shrinking the 4-8" band to only include extreme southern New Hampshire, while places like Manchester and Concord were under a 1-4" band.
* Wednesday afternoon: By late morning or midday, we knew we dropped the ball. Yes, it was snowing, but it was at or above freezing! Therefore, the snow wasn't accumulating. A mid-storm update was issued; the snowfall map completely done from scratch. See the final map here, which did verify.
Hopefully this gives you a little bit of an insight as to my thought process with this storm. That all said, let's move on...and enjoy the quiet weather for the next few days.










