5PM advisory on October 1 just came out. Forecast looks like it is going to go a little further east. Good news, iow.
Late in the period the track model spread remains considerable, as
the global models continue to have issues depicting the synoptic-
scale pattern over the eastern United States and western Atlantic in
4-5 days. The evolution of the mid/upper-level low currently
centered over the Ohio Valley and how it interacts with the western
Atlantic subtropical ridge appear to be critical to the long-term
track of Matthew. The UKMET and ECMWF tracks have shifted westward
and slower by day 5, while the GFS has trended east. This has
narrowed the guidance envelope somewhat, but given the lack of
run-to-run consistency I'd hesitate to say that confidence in the
long-range track forecast has increased by any appreciable measure.
The new NHC track at these times leans heavily on continuity, and by
day 5 is close to the GFS/ECMWF blend and left of the latest
multi-model consensus aid TVCN.It is important to remind users that average NHC track forecast
errors are around 175 miles at day 4 and 230 miles at day 5.
Therefore, it is too soon to rule out possible hurricane impacts
from Matthew in Florida.