Sabtu, 02 April 2011

About Last Winter

Winter is pretty much over after yesterday's April Fools' snowfall so it's time to dust off those faded memories of last winter's forecasts and see how everyone did.



The first graphic above is the original TV station forecasts from last November, followed by our  forecast for the winter across the Delaware Valley.   From a snowfall standpoint, nobody did well at all -- the final winter tally of 44" was double to triple what was forecast by everyone -- and is a brutal reminder that snowfall forecasting is a complete and total wildcard as placement, snowstorm track, etc. are all variables that are tough to pin down from one end of the winter to the other.   Needless to say, the winter snowfall forecast was a miss.

Let's compare temperatures.     The actual temperature departures from normal were as follows:

December -- 4.7 degrees below normal
January -- 3.0 degrees below normal
February -- 2.4 degrees above normal
March -- 0.9 degrees above normal

Channel 6 scored on two of the three months in temperature -- missing on January.   We missed on December (rather badly) but were close on January (not cold enough though) and correctly predicted a mild February.   Channel 10, who issues a forecast for March, did not do too badly on their March prediction (one degree below average isn't a bad call in November).

Channels 3 and 10 did not correctly predict any of December-February in terms of predictions from November although both stations revised their predictions midstream.  Channel 10's revised prediction was good on snow but missed the mark on temperatures, especially in February.

If anything, seasonal predictions are tough to pin down and no one forecaster can nail the entire winter perfectly -- whether it be snow or temperatures or overall pattern.   Each forecast has its hits and misses...some of those misses are quite significant (our December 2010 outlook was atrocious).   It is something that is "fun" to do in the sense of trying to get a handle on the winter pattern and to try to make sense out of the varied and crazy months of winter around the Delaware Valley.   Unfortunately, long range forecasting overall has a ways to go to get to the accuracy of a three or five day forecast.    Someday, though, it will be there...someday!

Tidak ada komentar:

Posting Komentar