What I find intresting is the continuity between the model runs. In years past, you'd have models all over the place, sometimes as soon as <36h out.

With Harvey (and so far with Irma), the models are pretty tightly clustered. While there is overall movement, all of the packages seem to trend the same way at the same time.

Have these two storms been easier to predict? Has the NHC reigned in their algorithms? Am I missing something?

It seems odd that for the last 48 hours most models are honing into and maintaing one general location, 5-7 days out.

D!