PDA

View Full Version : Model Accuracy



adam
12-20-2007, 10:42 AM
John, I heard that the 12z and 00z model runs are the most accurate because they are the two that get to use the NWS balloon data since they are launched at 6am and 6pm. Any accuracy to this statement? And since you do your forecast by 9am do you use the mornings 6z model run? or which one? Thanks

john
12-20-2007, 11:52 AM
Adam-

I find some unique quirks with each model and in some cases each model run. I don't think that the 06z and 18z runs are significantly less accurate than the 0z and 12z runs that incorporate that balloon data.

As far as my techniques go, I never use any one model output exclusively. I look at many models and many model runs, including some that are 18 hours old to make my forecasts.

-John

adam
12-20-2007, 12:16 PM
OK thanks, thats the info I was looking for. I find everything about weather forecasting and how it is done intriguing. (especially snow forecasting) Thanks for the input.

eric
12-20-2007, 04:56 PM
I'm with you there Adam. I'm very curious with the approaching system, and, given what there is to work with, JD, what track or events would be ideal for the U.P to receive the most amount of snow?

john
12-20-2007, 07:02 PM
Eric-

For the western UP, pretty much what is currently indicated by the last 2-3 runs of the European and NAM.

-John

eric
12-20-2007, 07:45 PM
thanks.