Satellite imagery above late this afternoon showing the mature and rapidly strengthening hurricane Dorian.

There have been noticeable trends in numerical guidance that started yesterday and continue in the models today with odds of the storm crossing into the Gulf of Mexico continuing to go down while the odds the center or eye of the storm could remain offshore to some extent rising.

The North to Northeast movement MAY also speed up some LATE in the forecast period after very slow movement toward Florida.

However, there remains significant differences in forward motion speed to impact timing and direction so that the outcome remains more uncertain than normal even while the model output has converged to some degree.

BOTTOM LINE... nobody is off the hook yet not by any means, but changes in data Saturday could result in the forecast getting worse for some and better for others FROM SOUTH FLORIDA TO NORTH CAROLINA!

First, the official NOAA/National Weather Service/National Hurricane Center Outlook (then some background explaining):

NHC STORM DISCUSSION (underlines are mine):

From the git go the emphasis here has been about how the uncertainty factor with Dorians track was much higher than normal. So there's no surprise that things continue to be in flux, and with this SOB (much like Matthew in 2016) that may be the case the whole way through to the end.

The current trend which may or may not be sustained going forward is for a rightward adjustment to the future track evident in both dynamic and statistical models (WxBell and StormVista):

So we are nowhere near figuring out this tricky forecast with a close approach to Vero Beach-West Palm Beach-Melbourne or just a little North or South. The forecast track could easily shift BACK left into Florida OR pull even more to the right farther away from the coast. So it's too soon to say definitively that Dorian will or will not make landfall but I can say as of now at least the range of outcomes has narrowed some.

But it only takes a minor adjustment in future data to go from landfall to no landfall in Florida. We also don’t know if it will make a landfall in or near North Carolina, Georgia or South Carolina but it could still be a strong tropical storm or hurricane at that point if it does!

Because it stands to reason if we’re unsure of Florida closer to the storm now in the short-term, then we certainly can not be more clear on locations farther away at a much later time and date.

We are MORE certain that Louisiana, Mississippi, Alabama, and the Florida Pan Handle are for now of little concern. Yet AL/NW FLA can’t quite be given an all clear yet but the focus has shifted away from there.

Small changes in the atmosphere hundreds of miles away upstream or downstream from Dorian can and do have great influence over Dorians eventual outcome (Lorenz Chaos Theory aka “The Butterfly Effect”). Google if you’re not familiar.

For example the continued shift east becoming more likely can be traced in part of a cluster or severe thunderstorms (MCS) that developed over Kansas and Nebraska. The MCS generated a Mesoscale Convective Vorticity center (mid-level disturbance in jet stream flow) that may have pushed odds of Dorian staying offshore up to around 40% for at least the moment.

From Weathermodels site you can see the situation described:

In the ECMWF Ensemble data for example there were 31 members that showed landfall in the previous cycle, now it’s down to 17 out of 51 in the current model cycle. However within those members there is still a 300 mile spread in forward speed by just 48 hours in the future and that’s significant enough to call into question the timing and location beyond two days regarding the eventual path.

For updates follow me on Twitter @MellishMeterWSB.