Re: 18Z GFS streaming in
Posted by Chris in Tampa on 9/5/2017, 8:47 pm
From what I understand about ensemble models, and I could be wrong, various things are tweaked in it, perhaps initial conditions and/or conditions around. If the individual members are closer together later in the forecast, then the forecast isn't as influenced by changes. If they are further apart, then the storm is more influenced by various changes.

So if the storm isn't analyzed very well by the models, or other things that would steer the storm are not, and the ensemble members are all over the place, there is more uncertainty.

Again if something isn't analyzed very well, but the members are closer together this time, there is more confidence as the various things tweaked didn't alter the various tracks of the members that much. So if one of those alternate scenarios happens, the actual track the storm ends up taking won't be much different than if that alternate scenario didn't happen.

I don't really know for sure.

This seems a little complicated:
https://en.wikipedia.org/wiki/Ensemble_forecasting
I'm not sure what explains it well off hand.
141
In this thread:
18Z GFS streaming in - Gambit7, 9/5/2017, 5:58 pm
< Return to the front page of the: message board | monthly archive this page is in
Post A Reply
This thread has been archived and can no longer receive replies.