Following up a continuing conversation with hurricane expert Jim Elsner, this post presents an analysis of Atlantic May-June SSTs versus normalized damage 1950-2005, but only including storms which had >$100 million in damage and storms of hurricane or greater strength, as recommended by Jim. As the graph below shows [10-25-06 update -- analysis superceded by Part 4 here], the results of this analysis show no relationship.

[10-25-06 graph reposted in part 4 with >$250M threshold]

I’d welcome Jim’s response, but for now I remain unambiguous in my conclusion that there is no relationship between SSTs and normalized damages. If Jim provides his data, I’d be happy to reconcile the different results, and perhaps my views will change. Until then, I necessarily must go with what the available data shows, which is quite unambiguous.