Totally unimpressive forecasting performance
Yeah, I’m not impressed either…especially with my certainty that the city of Fredericksburg would see a few inches of white stuff from yesterday’s storm. To be fair a few forecasters – notably those at the Capital Weather Gang – were rightfully suspicious of how far north the snow would reach and they were correct. So how far did the snow reach? Check out this satellite image from 8:30 a.m. this morning:
So what happened? Two reasons: First the track of the storm itself wound up 75 miles south of where it looked like it was going. That’s enough to really mess up a snowfall forecast in itself. Secondly, given the warm surface temperatures Wednesday the reason forecasters thought things would cool off enough to snow yesterday came from something called “dynamic cooling,” which is a mechanism by which an intensifying system will cool middle and low-level temperatures via a significant pressure drop. This dynamic cooling did occur nearer the path of the storm, which is why locations in southern and southwestern Virginia got hammered by snow even though temperatures in those areas looked like they would be too warm. But since this storm weakened as it approached the coast – instead of intensifying – the overall effect was to completely bust the forecast for the ‘Burg.
So where do weather folks – including rank amateurs like “moi” – go from here to rebuild the public’s confidence that weather forecasting is more accurate than flipping a coin? One step at a time, learning from each busted forecast, and refining the interpretation of the model results per local conditions.
Meanwhile the weekend looks to be sunny with a bit of a warmup after today’s chilly high temperature in the upper 30s. Saturday and Sunday will actually top out a bit above average, reaching the upper 40s, but then the Arctic floodgates open behind a cold front that will push through Sunday afternoon. This very cold and dry air mass will creep in from the northwest and temperatures will nosedive to below normals for the first half of next week. Check out this surface temperature graphic for next Wednesday morning: