crpeavley -- a building ridge should keep Irene moving more towards the west than towards the north through time. I think everyone is picking up on that, from the NHC on down, but perhaps not quite to the degree that HPC might be. It is the HPC's job to put out long-range forecasts, including when a tropical cyclone may impact things, so their thinking is usually the first you'll see from an 'official' source that talks about landfall regions. They often disagree with the NHC, and vice versa, as always happens with different forecasters; this isn't necessarily a bad thing. Really, the time frame they are talking about right now is beyond the NHC forecast period and more towards the long range (5-7 days), where most forecasters try to shy away from when it comes to tropical cyclone forecasts. It's just a wait-and-see thing right now, but not at all inconsistent with what the general consensus seems to be with this storm.

Saw the mm5 posting before about the track, followed by the wind plot -- I'm not sure how much stock I'd put in the intensity forecasts right now. With the change to the convective parameterization in the model, we tried to get rid of the spurious vortex development near Central America, something that appears to be working. Unfortunately, in this change, we weaken the ability for the model to recognize and grow shallow cumulus clouds/convection; these are important in transferring energy to the large scales and in the development of a storm beyond a minimal TS. There is still such a scheme in there, it is just not as sophisticated. So, while it remains to be seen as to whether or not this is a true forecast of intensity or a conservative forecast, I would tend to think it is at least slightly the former and thus, as always, should be treated with caution.