New research on AMO and hurricanes

[x-y-no]

This new statistical study (pdf) examines the alleged connection between the AMO and Atlantic tropical storm activity and finds no correlation.

I've only skimmed the paper so I can't comment in any depth yet.

[Aquawind]

Pretty Lippy discounting Atlantic Multidecadel Oscillation and of course anthropogenic is in the less than 2 pages of text 9 times.

[curtadams]

Actually the study is mild in its criticism of the AMO/SST correlation. The secular trend to correct with should be the global temperature changes rather than a linear regression. The global temp changes - http://en.wikipedia.org/wiki/Image:Inst ... Record.png - are a dead ringer for both the Atlantic SST changes and the smoothed Atlantic cylone rate - a big rise just before the early 40's, a slight decline to the mid-70's, and then a brisk rise ever since. Since the p value for the AMO association is only .01 after correction for just a linear regression, after correcting for the known temperature background it's going to be far from significant, even with a 150 year time series.

Incidentally, the chart shows how that recent publication disputing Emmanuel's power increase is bogus. It claimed the changes were statistically insignificant from the mid-80's to now - but if you look in the chart the cyclone rate has almost doubled over that period. So all the naysayer really showed is that his own study was so hopelessly underpowered it couldn't detect a doubling of cyclone rates (given how variable cyclone incidences are, I'm not at all surprised you can't find anything with a 20-year study. Except the recent acceleration in monster majors is still p < .001!)

[x-y-no]

Pretty Lippy discounting Atlantic Multidecadel Oscillation and of course anthropogenic is in the less than 2 pages of text 9 times.

Lippy? I'm not sure what that's supposed to mean. By my quick purview, their methodology seems good, but even though I was a math & computer science major statistics wasn't exactly my first love. So I'm going to have to slog through it more carefully to be sure.

BTW, they don't say there's no such thing as the AMO - what they're saying is that there appears to be little if any significant correlation between the AMO and SSTs in the tropics.

[btangy]

Pretty Lippy discounting Atlantic Multidecadel Oscillation and of course anthropogenic is in the less than 2 pages of text 9 times.

The article doesn't discount the AMO. It just shows that after separating the Atlantic Main Development Region (MDR) SSTs into 1) global mean SST increase due to mean global radiative forcing changes, 2) N Hemisphere aerosol forcing, and 3) "natural" noise which is assumed to contain the AMO, the power spectrum of 3) does not contain a significant AMO-like peak (multi-decadal time scales) after factoring in the aerosol forcing (2). The MDR SSTs are not affected by the AMO and since the MDR is so strongly correlated to both the power dissipation index (or ACE) and the TC counts, the connection between the AMO and TC variability in the Atlantic is called into question.

[Aquawind]

Righto the AMO and SST relationship of course.