Sunday, October 09, 2011

Our History Of Winter Forecasts

This will be our sixth winter forecast at Phillyweather.net and over the past five years we've had our share of hits, misses, and flat out follies in winter forecasting.   Long range forecasting has some merit and some accuracy, specifically to trends as to whether one will see colder or warmer than usual conditions for the winter ahead, as well as the potential for snowier conditions.   The specifics are harder to pin down...as is timing...as a later or earlier-than-expected cold snap can ruin two months of one's winter forecast although you end up getting the winter "right" overall.   Instead of hitting up only our "best" features, we'll pull out our dirtier forecasts as well...and no lipstick will be needed on some of these pigs since the ugliness cannot be avoided!   

As a point of comparison, I've also included the rest of the local media as a point of comparison.   A lot of the local weather forecasters and meteorologists have had similar forecasts over the years -- some of them really good, some of them worthy of the blooper reel.   I think I can roll with the variant punches, shoe busting, and  yucks that will follow.



2006-2007 was our first real forecast.  On the whole, it went well.  We weren't warm enough in December or cold enough in February but we got the idea of a relatively milder winter on the whole pretty well right.   This was the winter of extremes in temperature (low 70's in January) and precipitation type (sleet).  We came up a bit short on forecast snowfall (13" in reality compared to our forecast) but were lower than many other forecasts in a Nino winter.

2007-2008 was a misfire as we incorrectly timed a cold snap for December (it hit in mid November) but accurately predicted milder conditions in January and February.  We forecast average snowfall in a La Nina winter but it ended up being less than half of what we predicted (6.3" on the whole, much of it in one snow event in February).  Most everyone had the right idea of the winter being relatively snowless on the whole.

2008-2009 had the right idea of being a "normal" winter but we misfired on temperatures (coldest to average was January, not December) and we were a bit short on snowfall (22.8" in reality, 14-19" was our forecast).  This was one forecast that I wish to have back from a specifics standpoint.     We were pretty close to the mark in terms of seasonal temperatures though as winter averaged out to exactly "normal" on the old climo normals...on our current normals it would have averaged out to a half degree below average.  Thankfully nobody else was that hot in temperatures either.

The winter of 2009-2010 was one where predictions were thrown out the window early and by January the rest were being used in dumpster fires.   Our predictions, well, weren't that hot...at least we got the cold idea right overall, as well as in December but misplayed February quite badly from a temperature and precipitation standpoint.   At least we got the right idea of going above average on snowfall.   The rest of the locals struggled on snowfall...as should not be surprised as nobody came close to predicting 78.7" of snow that winter.

Last winter was another one where we whiffed on one month's forecast (December was almost five degrees below average and we forecast an above average temperature that month), which made for a pretty bad winter forecast on the whole despite hitting January and February on trend (colder in January, warmer in February).  Two out of three ain't bad on temperatures but missing one month by seven degrees will ruin your whole winter.   Snowfall was a mess and a miss as well...not just for us, but for everyone else.

The past several winters have been wild rides, with extremes in ocean state and atmospheric blocking that have been harder to predict.  It goes to show that long range forecasting is still very much a work in progress.  There is a good bit of work to be done in research, study, and methodology to improve the forecast science in the longer range.  There's a good bit more accuracy overall than there used to be despite the misfires of the past two winters.  Science will help continue to improve the method to the madness that comes about each and every winter.