qanda, where does F&R involve collinearity? Did they use a linear time trend or a linear trend? Is the trend used/found by F&R for GISS 1.024 and the overall trend 1.024 + other variables such as solar, volcanoes and ENSO? If so does this not mean that - 0.024 Trend(GISS) = other variables so that in effect F&R have simply fitted GISS to itself?

Feel free to have the same amount of audacity in your answer as I apparently have in my posts.

Take a deep breath and read, very slowly:

http://iopscience.iop.org/1748-9326/6/4/044022/fulltext/

with an open mind, it would help.

Good night Mr Cox.

qanda, q&a; both sactimonious, supercilious and vacuous; and a waste of space and money.

Mr Cox,

Quote:
>> F&R is a bad paper; looking at their EQUATION 2:

GISS = -91.43 + 1.024Trend + 0.0761MEI(4m lag) + 0.06694TSI.PMOD(1m lag)- 2.334AOD (7m lag)

Yada, yada, yada …

F&R have not solved for GISS. By including Trend(GISS) as an independent variable they have eliminated GISS. What they have shown is that the Trend in GISS can be fully explained as a linear result of MEI, TSI, and AOD, without any reference to CO2.

If you think I'm wrong in agreeing with that conclusion explain why. <<

Wait, there is more:

>> qanda; look at Table 1 from F&R; add the linear trend for GISS of 0.167 deg C/Decade to their multiple regressions of each of the independent variables, ENSO (MEI), Solar (PMOD), and Volcanic Aerosol (AOD)along with their lags, as estimated by F&R, and you get this 2nd equation [the first being F&R's multiple regression of only the independent variables]:

GISS = -91.43 + 1.024Trend + 0.0761MEI(4m lag) + 0.06694TSI.PMOD(1m lag)- 2.334AOD (7m lag)

How is that wrong? <<

.

Response:

“Because it is not "their EQUATION 2", Mr Cox. It is yours, based on a simple cherry pick and an even simpler math - ask Hasbeen. It's more complex.

If you want to talk the talk, you have to walk the walk - OLO doesn't cut it. I understand why you think it does.”

But wait, there is more:

“Mr Cox shows elementary stupidity himself by not understanding the problems with collinearity in multivariate regression analysis (despite this being pointed out to him and despite data and methods being explained) and has the temerity to 'verbal' and claim the experts have said something they in fact haven't.”

.

Mr Cox’s retort:

>> qanda, where does F&R involve collinearity … yada, yada, yada <<

.

Response:

Mr Cox, they didn’t … you did, in YOUR equation 2.

What F & R 2011 show is that if you remove the 'noise' of El Niño/southern oscillation, volcanic aerosols and solar variability from the time series, the only other exogenous factor that explains the warming trend is CO2.

Q.E.D.

qanda,

I'm sure you'll be interested in this:

http://oncirculation.com/2013/04/30/6510/

Perhaps it's what cohenite means by "being shredded in the blogs".

Q&A gets the declared intention of F&R right; they claim to isolate the AGW warming signal by removing, not noise, but the temperature impact of the other exogenous factors such as volcanoes, ENSO and the sun.

But they make at least 2 ‘mistakes’.

The first is they include the temperature trend in the multivariate regression; they say this:

“Using multiple regression to estimate the warming rate together with the impact of exogenous factors, we are able to improve the estimated warming rates, and adjust the temperature time series for variability factors.”

But trend is NOT an independent variable, it is dependent on exogenous factors so the issue is should the trend have been included in the multivariate regression analysis?

The short answer is no because the trend itself is contributing to the finding of the trend; as I said before F&R are using GISS to prove GISS.

That is statistical chicanery.

The 2nd issue is they have defined at least ENSO wrong and misunderstood the effect of solar.

F&R say this about ENSO:

“This confirms that the influence of ENSO is greater than that of volcanic forcing and much greater than that of solar variation, and that both ENSO and volcanic forcing affect LT temperatures much more strongly than surface temperature.”

LT is lower troposphere; but ENSO is a coupled surface and atmosphere effect; you cannot assume ENSO only affects LT temperature and reduce ENSO to an index as F&R have done for that reason.

Likewise with solar; it’s effect is a cumulative one which an estimation of lag does not deal with; by assuming the effect of solar is finished at the expiration of an assumed lag as F&R do is to miss the effect of solar on temperature as this paper explains:

http://vixra.org/pdf/1108.0004v1.pdf

F&R also do not find a break or step in the temperature trend which contradicts a large number of papers; see:

http://arxiv.org/abs/0907.1650

http://www.pas.rochester.edu/~douglass/papers/Douglass_Knox_pla373aug31.pdf

https://pantherfile.uwm.edu/kswanson/www/publications/2008GL037022_all.pdf

http://www.arl.noaa.gov/documents/JournalPDFs/Seidel&Lanzante.JGR2004.pdf

http://www-eaps.mit.edu/faculty/lindzen/203_2001GL014074.pdf

http://rossmckitrick.weebly.com/uploads/4/8/0/8/4808045/2011-09.pdf

It is not my equation; it is a fair representation of what F&R have done. Q&A’s failure to understand that is his problem.