It first rips up the Florida coast and northward toward coastal South Carolina Friday and Saturday. After it retreats from coastal South Carolina on Saturday night, it drifts east and then southwest — passing near the western Bahamas for an encore Monday. Then, Tuesday, it’s in the proximity of Florida for a second time.
This scenario is plausible. Several models simulate this kind of loop. Matthew performs this loop as high pressure — which had been steering it westward — gets pushed out to sea by a cold front that nudges the storm eastward initially. But then a new area of high pressure builds over the eastern U.S. and western Atlantic Ocean early next week, which pushes Matthew south and then westward back toward Florida.
Whether Matthew performs this kind of loop will depend on the strength of the high pressure system that builds in to its north next week. If it turns out to be weaker than forecast by the GFS model, Matthew may find an escape route out to sea early next week — which the European model shows in its latest forecast.
A wild card here is Tropical Storm Nicole — forecast to be east of Matthew early next week. Tropical cyclones sometimes repel or orbit around each other — this is known as the Fujiwhara effect — which could also foil any effort Matthew makes to escape.
If Matthew manages to hit Florida twice, it’s unlikely it would be as strong the second time around because it would have to pass over cold water left in its own wake. But even the prospect of such a double land strike was “worthy of profuse profanity” in the words of Weather Underground meteorologist Jeff Masters.
Capital Weather Gang tropical weather specialist Phil Klotzbach told us via email that such a strange track has precedent in historical records. His list of “real weirdo tracks” includes 2005’s Hurricane Ophelia, which made not one but two loops off the Southeast U.S. coast. Klotzbach created a map showing his “five favorite looping Atlantic tracks”:
With Matthew, it’s wise to expect the unexpected. Hurricanes are extremely sensitive to small changes in atmospheric steering currents. Forecast models — especially beyond 48 hours into the future — have sometimes profoundly changed the storm’s projected course due to the most subtle changes in the very fickle winds above.