New Image:
with the far-left 18Z GFS at the same verify time (in the 500mb layer):
Old Image:
Note that the 00Z GFS actually initializes Ivan in a deeper layer and does not treat him as a weaker system. So far we are through 66 hours and the ultimate solution could be the same...but there will undoubtedly be a shift to the right in the GFDL and subsequent NHC suite runs.
Also...will be interesting to see how the model resolves the mean steering ridge late in the period...will be waiting for that.
MW







