But why? Why would a specific 30-day period that is entirely a human construct of time, separated by 61 days, have any effect on the natural weather patterns of a 90-day season, that is again defined by dates that are entirely human constructions? The fact that any September with a CET above 16°C has yet to provide a below average winter (given that there's only about 6 or 7 of them anyway in a 360-odd year dataset) is merely a coincidence - there is absolutely no causation in it whatsoever. And yes, you can say that because you can find coincidences every where you look.
This theory about warm Septembers and mild winters gets peddled every year and it never ceases to drive me mad. We really need a good record warm, 17°C+ September to be followed by a below average winter to put this rubbish to bed.
We do tend to get 'stuck' in certain broad, recurring patterns for months at a time (might be zonal, or northern blocking, or mid-Atlantic blocking, or the Jet being displaced to the north allowing ridging from the south, etc, etc)
Every year is different with different driving/teleconnecting factors, but it may be that certain patterns established in any particular month have a higher likelihood of repeating for the next few/several. I think that's where a lot of weather lore derives from, too.
As for looking at records (temp, rainfall, etc) for a specific month doesn't tell the full story. Very different broad set-ups can yield very similar temp/rainfall records. That's why I think raw records data can be a useful place to start, but one needs to examine the actual pervading weather set-up patterns for each year to get a more detailed picture to build on.
Martin
Home: St Helens (26m asl) Work: Manchester (75m asl)
A TWO addict since 14/12/01
"How can wealth persuade poverty to use its political freedom to keep wealth in power? Here lies the whole art of Conservative politics."
Aneurin Bevan