Wednesday, June 18, 2014

Modeling

(Click on the cartoon to see the full image.)
(C)Copyright 2014, C. Burke.

Because either though I don't know where I'll be or what I'll be doing three days from now, I need that fifth-gen Doppler tech to tell me the weather through the end of next week! Plans need to be made!

On the bright side, the bells and whistles curve is exponential.




1 comment:

William Ricker said...

Not quite fair. The 10-day model is now as borderline-useless as the 7-day used to be, which is why it's being run that far and shared as a forecast. If taken with a ±100mi / 8hour accuracy, at 5 days it's not terrible. That's not good enough to plan a picnic between storms, but it's better than climatology to tell you sun vs cloud and %precip. When I care about ±1mile ±1 hour, I look to the 3-hour RAP model. (When they add ACARS soundings and MESONET to its INIT, it'll get really finescale. AFAIK those are in Experimental but not Operational still.)