Wednesday, October 21, 2015

Will This Be a Ski Free October?

Now is the time of year when people ask me repeatedly when it's going to snow.  Actually, I get that almost all season long, but it is times like these, when we are in a snow drought, that people start asking things like "what about next week" or "how does next month look"

The problem with answer questions like these is simple.  The atmosphere is chaotic and weather is unpredictable, at least with current tools, at such long lead times.  This is true despite the fact that Accuweather issues a 45-day forecast.  In some situations, one can predict a shift in the climatological odds of above or below average conditions (an example would be the expectation of a warmer than average winter in the northwest due to El Nino), but if you want to know what the weather is going to be like 10 days from now, good luck in most situations.

Here's a good example.  I confess that when in a snow drought, I often look deep into the extended range hoping for the possibility of a storm.  I did that with yesterday's GFS and here's what I found for 0000 UTC October 30.  A ridge with dry weather over nearly the entire western U.S.  Skiing in Utah?  Fat chance.

Then I did it with today's GFS and here's what I found.  A nice broad trough with precipitation over the western U.S.  Skiing?  Well, we'd not a lot of snow, but this gives us a bit of a chance.

When we see a model shift of that type in the extended range forecast, should we interpret that as a trend?  Probably not.  At such long lead times, one can see large flip flops from one GFS run to the next due entirely to the chaotic nature of the atmosphere.

Here's the proof.  Instead of looking at a single model forecast, let's look at an ensemble.  Below are last night's GEFS forecasts for 0000 UTC 30 October.  These are produced with the exact same model, but slightly different initial conditions.  Over time, the solutions diverge and if we're looking at the forecast many days out, we see dramatically different forecasts.  Some put a ridge over Utah, others a trough, others different flow configurations.

Source: Penn State E-wall
So the weather later this month is pretty unpredictable.  Nobody can reliably predict what is going to happen, so no point getting excited about today's GFS run.  Check in next week when the lead time is shorter.


  1. What about November? It has to snow in November.

  2. I think I remember something about long range forecasts being, shall we say, BOGUS!

    1. Yeah, that was a guest post by Jeff Spicoli ;-).