Forecaster Colby wrote:I have found that though the NHC has very smart and skilled people there, they make some foolish policies:
-Erring on the side of caution for T.Ds/weak T.Ss and underestimating the big ones.
-Relying on the models when their instincts tell them otherwise. I called Katrina and Rita's nutso strenthening based purely on instict and the conditions, and someone at the NHC had to have felt the same.
-Trying to validate previous forecasts. The NHC leans towards what they forecasted 24 hours ago (like with Gamma, they didn't forecast dissipation), even with evidence to the contrary.
Don't get me wrong, I have a good deal of respect for the NHC, they did a great job when most storms followed climatology and norms in a given situation, but that isn't true recently...and I think it's time they recognize it.
There is reasoning behind every one of there desicions, and basing forecasts on continuity is to ensure that the models didn't go crazy for one or two runs, or that a strange flux in a particular feature isn't temporary. If they were to immediately change their forecast every time they or the models saw a change in what is most likely going to happen, then the forecasts would be flying all over the place and people would tend to trust the forecasts less and less, as a sudden shift from a landfall in Miami up to SC and then back to Melbourne and then keeping it fish would just look very unintelligent to most people, especially the public and media. While the overall forecasts might be a little slow to change, the NHC will make the calls as they see fit and do it without making sudden changes to a track until it seems that the new track is not some weird run. They do this because they must account for error, and trying to make suddent changes to forecasts will, in the long run, not effect the error at all, or possibly make it worse, as I'm sure you remeber the times this season the NHC has not changed there forecast to comply with strange model runs that have lasted for over a day. The NHC turned out to be right in those cases, and better warning was provided for the locations in the path of the storm. Sure, sometimes it works the other way too, but we can't eliminate all error, and being slow to change from something that makes sense and works seems like a much better strategy than jumping around like a second grader playing hopscotch.