Blogging With The Data You've Got

While looking over who has been dropping by TIM lately I came across this discussion about my post on Hurricane Myths.

In an attempt to refute my work the following was stated:

Misleading Because he limits his analysis to Hurricans that hit the United States," and not the sum total of ALL Atlantic hurricanes, many of which never strike the US.

I didn't address this issue in my original post because I believed the answer to that question was self-evident, but maybe for many it isn't. If one wants to compare data on hurricanes into the 19th century it makes sense to limit the discussion to only landfalling hurricanes. Why? For many reasons:

A) Without benefit of satellite imagery many non-land falling hurricanes in the 19th century would have been unknown. Failing to take this into account would seriously skew results in favor of saying there were many more post 1950 hurricanes. A similar thing happens when you hear of a "record number of tornadoes" striking the U.S. Well, with the tremendous increase in the number dopplar radar sites finding storms that might have never been seen by human eyes (e.g. remote rural areas, or in the dead of night), that should come as no surprise. It certainly doesn't indicate that the number of tornadoes are increasing.

B) Logically, there is no reason to think that the incidence of U.S. landfall should change statistically speaking from the 19th century to today. In other words, if there are more hurricanes across the Atlantic basin as a whole, there should be a corresponding increase in the number of U.S. land-falling storms. Why one would wish to believe otherwise escapes me.