Wednesday, March 14, 2012

Forecast Uncertainty

It's easy to succumb to temptation, but one needs to be cautious with regards to forecast uncertainty.  All model forecasts are physically realistic and plausible, but the issue is, will they verify.

I am not producing a detailed forecast for this weekend yet because I'm not sure how things are going to play out on the local scale.  Let's take a look at the last two GFS forecasts as an example.  The run initialized at 1200 UTC this morning, puts a pronounced front and frontal precipitation band over extreme northwest Utah at 0600 UTC (0000 MDT) Saturday 17 March.  There's some precipitation over the Salt Lake Valley and adjoining Wasatch Mountains, but it is not especially heavy.


The run initialized at 1800 UTC this afternoon, however, has the front about 100 km further south, with heavy precipitation right on the doorstep of the Salt Lake Valley and adjoining Wasatch Mountains.


That's a modest change in position, but a major change in the forecast if you are say trying to predict what will happen at Snowbasin.    In the scenario at the top, you are warm, prefrontal, with perhaps a few snow showers.  In the scenario at the bottom, you are right in the frontal band and it is nuking snow.

Colder weather and snow is coming, but I'm still holding off on bold declarations for where, when, and how much.

2 comments:

  1. Is whatever falls going to be the good stuff, light and dry, or is this a denser type of snow with this system?

    ReplyDelete
  2. http://www.inscc.utah.edu/~alcott/forecast/bufr_output.txt

    That's our snow density algorithm, although it assumes the model forecast is dead on.

    ReplyDelete