It has always puzzled me to see this bump upward in storm numbers amidst the trending (smooth) decline after the first 1/3 of September. I would like to think that the small sample we have of storms; <125 years, is the reason and that a few hundred more years of data would take care of my suspicions.

I am cognizant that the possibility exists during the transition from the most active TS period to the end of the Eastern season, factors may come into play in a way that enhances storm formation during this suspect, by me at least, few weeks.

I am so far removed from the little knowledge of statistics I once had that I'm unable to run data and do the testing required for assessing statistical significance and confidence in the validity of the October outlier.

The answer I seek may already be in the literature and/or you may have ideas about this issue. Any thoughts or help will be greatly appreciated.