Guest Post by Paul L. Vaughan, M.Sc.
Does the sun tweak the odds of busy Atlantic hurricane seasons on decadal timescales?
ACE = accumulated cyclone energy (based on duration, intensity, & number of storms)
SCL’ = rate of change of solar cycle length, calculated by applying a complex Morlet wavelet [at 4 different wavenumbers] to sunspot numbers:
ftp://ftp.ngdc.noaa.gov/STP/SOLAR_DATA/SUNSPOT_NUMBERS/INTERNATIONAL/monthly/
Please read the cautionary notes regarding the early Atlantic hurricane record:
http://www.aoml.noaa.gov/hrd/tcfaq/E11.html
(Scroll down on that page for Atlantic hurricane data.)
Fotheringham & Rogerson (1993) assert that scale-dependency “[…] presents us with the challenge of reporting on the reliability of parameter estimates in the light of changes in scale […]”
Fotheringham, A.S.; & Rogerson, P.A. (1993). GIS and spatial analytical problems. International Journal of Geographic Information Systems 7(1), 3-19.
Alert statisticians may recognize the preceding as a cautionary note about “Simpson’s Paradox”.
Supplementary:
AMO (Atlantic Multidecadal Oscillation) can be decomposed into a multidecadal wave and an interannual component, to arrive at the following:
[1a] indicates simple [i.e. boxcar-kernel] 1 year smoothing.
So another question arises:
Is the interannual variation simply ENSO (i.e. as many might suggest)?
Caution is advised with this notion, as follows:
a) Regional/hemispheric/interannual-scale spatiotemporal turbulence is constrained (by neighbouring pattern & process on a finite globe).
b) Interannual variations of various climate indices show intermittent nonrandom coherence with the highly regular QBO (quasi-biennial oscillation).
Bear in mind that as the solar cycle accelerates and solar cycle length shortens (SCL’ negative), beats of decadal-timescale component features (e.g. solar wind) with the terrestrial year & QBO lengthen on average – i.e. it takes longer, on average, for particular phase-alignments to recur. Conversely, as the solar cycle decelerates and SCL lengthens (SCL’ positive), beats with terrestrial temporal modes such as the year & QBO shorten – i.e. particular phase-alignments occur with higher frequency.
Speculation:
The near-stationary QBO is lunar-driven and solar cycle phase acceleration (or something confounded with it) regionally modulates terrestrial cloud, insolation (not to be confused with irradiance), pressure, circulation, & temperature patterns.
Earlier WUWT articles in this series:
1) http://wattsupwiththat.com/2010/08/18/solar-terrestrial-coincidence/
2) http://wattsupwiththat.com/2010/09/04/the-north-pacific-solar-cycle-change/
3) http://wattsupwiththat.com/2010/09/11/solar-cycle-length-its-rate-of-change-the-northern-hemisphere/
In considering the preceding, the role of north-south & continental-maritime terrestrial asymmetry should not be underestimated.
Non-statisticians would appreciate a non-technical discourse putting these complex but indicative computations in context and perspective.
I’m reminded of the pony story popularized by Ronald Reagan. In that vein, I say, keep digging. I am yet to be convinced.
I would agree the earth’s heat engine must have some relationship, probably many, to events in the fluid dynamic systems call climate, weather and yes even storms. I would agree that major ocean currents and cycles of one system will have some influence on the happenings in the other. Looking only at the data presented some correlations look good other not so good. It is clear that other influences are present but we do not know what they are. I also think great deal more calibaration is required for all these influences and effects. Like John F. Hultquist wrote “…keep diging, I’m not yet convinced.”
Hmm… Simply eye-balling the ACE, a five-year cycle seems apparent. Oh, it’s a bit longer, or shorter sometimes. Occasionally, it skips a beat or has a double peak. Seems to be absent, or there’s inadequate data to compute a complete ACE prior to 1950.
Rate of change of the solar cycle length with a time resolution much finer than the cycle itself does not make much sense [regardless of how fancy the technique is], and in addition the SCL itself is nebulous as cycles overlap.
The observational record can only tell you so much about the convoluted relationships between climate, natural and anthro, as well as the sun. It is a considerable feat to extract the ENSO signal from a dataset — and then to correlate something like solar cycles and ACE. In my opinion, there are too many degrees of freedom and mechanisms that I would have a difficult time even deducing the sign of the impact.
This is a good idear for a idealized climate model where you can turn cranks and switches to even hint at quantifying or even qualifying sensitivity.
Thanks for mentioning Simpson’s paradox. Fascinating! Another interesting statistical pitfall into which the naive may fall.
Is not the real problem with this that rate of change of SCL is not a force? What is the mechanism the leads from this number to hurricanes? There must be many assumptions and steps from one to the other.
I wasn’t going to say this but with Ryan Maue introducing the phrase
“you can turn cranks and switches to ” . . . this seems Rube Goldberg –like: http://www.anl.gov/Careers/Education/rube/rubeolive.html
Gene:
ACE is an artificial metric. which is fine is you want to do year to year comparisons but meaningless if you are trying the understands what drives a process in a rigorous fashion. Actually none of these indices has any strict physical meaning.
Ask yourself what the units on AMO are.
Steven Mosher says:
October 11, 2010 at 9:30 pm
Actually none of these indices has any strict physical meaning.
And that includes the solar cycle length
The perigees of the moon’s orbit could well be a big factor
http://www.predictweather.com/ArticleShow.aspx?ID=221&type=home
” “There should be more tropical cyclones in 2011 than in 2010.””
Okay, there is a correlation between two sets of data. Until the physical process that links the two is identified, it is an interesting mathematical exercise.
Statistics, data analysis, etc. is interesting and could possibly find a causal relationship between two thins but is nothing more than a starting point.
http://www.nationalforestlawblog.com/Global%20Warming%20By%20Paul%20Pierett.pdf
So yet another “it’s the Sun stupid”? I am no critic of this post, I am not a “climate scientist”! William Herschel, come back, all is forgiven. How could you have won bets on the price of wheat by measuring the rise & fall of Sunspot activity?
We don’t know what true effect element “A” has on element “B”, (very low level of scientific understanding etc), but we know for certain (IPCC AR2/3/4/5/6/et al & all the others that haven’t been written yet, as we know the conclusions of course), that element “C” overpowers it! That’s great science indeed! I know AR4 claims a mere “low level of scientific understanding”, but I suspect although our knowledge is improving about how the Sun works, & it’s interactions with the Solar System, the effects on the Earth, (Gaia, home, fill in whichever), it’s magnetic field, it’s atmosphere, etc ,etc, we are probably either not that much closer than we were, or we are tantalisingly close to discovering what lies ahead, good or bad!!!!! Discuss?
My thanks to TimM for the link to a fascinating article “Watch out for perigee”.
During my time with the UK Met Office several colleagues had researched lunar cycles as a spare time interest. Some articles had been written on the relationship between lunar cycles and the planet’s weather. Unfortunately the Met Office’s vetting process before publication could be permitted in those days – and possibly still present today – meant that such papers didn’t get further than first or possibly second line management. Depressing evidence of closed minds that ultimately took me out of professional meteorology for good.
In my earlier days I had worked with a senior forecaster who had similar interest in planetary conjunctions and their effect on Earth’s weather. It was fascinating to work closely with a few people whose minds were open and full of exciting theories and ideas.
That’s what I admire about the many excellent contributors and contributions to Watts Up With That – open minds and exciting ideas and propositions!
As I read comments in this thread, I’m particularly interested in knowing (a) whether readers believe that the interannual variability in AMO is mostly due to ENSO, AO/NAO/NAM, …or something else and (b) [if so] what they think accounts for exceptions.
Steven Mosher-“Ask yourself what the units on AMO are.”
Unless you normalize it, the units are still in degC. The AMO is the difference between a temperature anomaly series, and the long term OLS trend line, both have units of degC, so their difference is also in degC.
It’s not really physical, but units nonetheless.