Chances are good, if you're into weather you've seen the declaration that Tropical Storm Humberto, on its quest for hurricane status, was assuring 2013 a slot for second latest hurricane development in any Tropical Atlantic season on record. That's not exactly correct, as soem legwork by my friend and NWS forecaster David Roth reveals.
It's IS correct to say that, *in the satellite era,* 2013 will be right up there as latest season for hurricane development. That qualifier - in the satellite era - is an important one. One of the biggest efforts for the field of meteorology in recent years has been for effective and reliable "reanalysis" - that is, trying to ensure that our historical database is as accurate and representative as possible. This is important for many parameters and phenomenon in the world of meteorology, and in a time of extreme weather and media saturation of the extremes, one could argue accurate representation of the past has never been so important for the field.
So, quite literally for the record, here's the official data:
1) and 2) Atlantic Tropical Seasons 1907 and 1914 - no storms reached hurricane status
3) October 7, 1905
4) September 18, 1941
5) September 16, 1877
6) September 15, 1922
7) and 8) September 11, 1912 and 2002 (Hurricane Gustav)
2013, of course, hinges upon Humberto. The first reaction, understandably, of many is to contend that we really should only consider the satellite era, as there was a likelihood of missing some storms before satellites helped meteorologists keep watch of the Tropical Atlantic starting in 1961. Then again, there were reconnaisance missions beginning in 1944 ("the reconnaisance era"), and ships relayed storm reports starting in 1905. So, the next question would be - where do you start "counting" the database as valid? If you argue for satellite era only, then it raises a host of other questions - for instance, our climate record referring to climate change must start at 1979 (start of satellite estimates) and the tornado record may have to begin as late as 1980 (widespread population of radar network accelerated). Of course, in the latter case, this would rule out classic tornado outbreaks that are an important part of the historical record.
So, it's important for us to include the "satellite era" qualifier, and understand that the official historical database reflects longer seasons without a hurricane.