In response to suggestions that we look at the several hurricanes in
the summer/fall of 2004, I did try an exploratory look.
Hurricane Ivan was in the news for more than a week, and wrought
havoc in several places at many separate times. We are today
(18th) getting rain in the NY area as the last gasp from Ivan. What
would be a good time period to specify in a prediction?
About the closest I could come to a sensibly limited segment
would be a few hours following 3 am on the 16th (when
Pensacola was wrecked). But why not when Grand Cayman was
hit, or Jamaica, etc? It is like the middle east war.
The following figure shows the data of Sept 18, and the time Ivan made
landfall in the northern Gulf coast is marked. Perhaps one could
construe the precipitous fall of the trace following that time as
correlated with the strong focus (in the US) on the hurricane, but of
course this is not a formal prediction and hence cannot be rigorously
interpreted. It is, however, interesting to look at.
|