Not that I am any sort of expert, but I completely agree with Clark that it is premature to proclaim a track shift to the west if Dennis goes underneath Jamaica. Although it is true that a jog left or right now often extrapolates out down the road in the sense that the models might be over or underinfluenced by some environmental feature, it is not necessarily the case that the models being off by 50 miles now turns into it being 250 miles off 4 days from now.

This is one of the things that frustrates me most about atificial "straight line models" -- if the forecast point is off by half a a degree today, then people want to "redraw" the same straight line they just had on their charts, using the new point to run from. The consequence is that the error now translates into a massive move down the line when, in reality, the storm isn't following a straight line to begin with and can shift back the moment you go to bed that night.

No pun intended, but experience tells us that an error now does not have a linear effect on the storm beyond now. .