The GFS didn't even back the surface winds at all... even as close as the 18Z run yesterday, only a few hours out from "go time"...
It didn't until last night's 00Z run at 00hr (mid event), only because it was FORCED to by the observational data it was initialized with...
It didn't until last night's 00Z run at 00hr (mid event), only because it was FORCED to by the observational data it was initialized with...