Just two explanatory variables (GHG and AMO) still account for 93% of the temperature variance.
Dr. Leif Svalgaard sends word of this article in Geophysical Research letters by Petr Chylek, James D. Klett, Glen Lesins, Manvendra K. Dubey and Nicolas Hengartner Article first published online: 5 MAR 2014 DOI: 10.1002/2014GL059274
Abstract
A multiple linear regression analysis of global annual mean near-surface air temperature (1900–2012) using the known radiative forcing and the El Niño–Southern Oscillation index as explanatory variables account for 89% of the observed temperature variance. When the Atlantic Multidecadal Oscillation (AMO) index is added to the set of explanatory variables, the fraction of accounted for temperature variance increases to 94%. The anthropogenic effects account for about two thirds of the post-1975 global warming with one third being due to the positive phase of the AMO. In comparison, the Coupled Models Intercomparison Project Phase 5 (CMIP5) ensemble mean accounts for 87% of the observed global mean temperature variance. Some of the CMIP5 models mimic the AMO-like oscillation by a strong aerosol effect. These models simulate the twentieth century AMO-like cycle with correct timing in each individual simulation. An inverse structural analysis suggests that these models generally overestimate the greenhouse gases-induced warming, which is then compensated by an overestimate of anthropogenic aerosol cooling.
1 Introduction
During the past century the Earth has experienced considerable warming due to anthropogenic as well as natural causes. Although a substantial body of research suggests that most of the warming has been due to an increasing atmospheric concentration of CO2 and other greenhouse gases, an exact partitioning of the magnitude of the global warming due to the natural and anthropogenic causes remains uncertain. Most climate research has centered on the use of coupled AOGCMs (atmosphere-ocean general circulation models) to elucidate the climate system from near first principles representing physical, chemical, and biological processes.
Empirical statistical models have been used recently [Lean and Rind, 2008; Foster and Rahmstorf, 2011; Mascioli et al., 2012; Zhou and Tung, 2013; Canty et al., 2013; Chylek et al., 2013] to complement physics-based models and are contributing to our understanding of anthropogenic and natural components of climate variability. The method assumes a linear relation between the observed temperature and a set of selected physically plausible explanatory variables (predictors). A typical set of explanatory variables includes the known radiative forcing and an additional factor characterizing the oceanic influence on climate [Compo and Sardeshmukh, 2009; Zhou and Tung, 2013].
Major radiative forcing includes solar variability (SOL), volcanic eruptions (VOLC) [Douglass and Clader, 2002; Haigh, 2003; Scafetta and West, 2006; Camp and Tung, 2007; Lean and Rind, 2008], anthropogenic greenhouse gases (GHG), and anthropogenic aerosols (AER). The oceanic influence is usually characterized by the El Niño–Southern Oscillation (ENSO) index [Lean and Rind, 2008; Foster and Rahmstorf, 2011]. However, the AMO (Atlantic Multidecadal Oscillation) [Schlesinger and Ramankutty, 1994; Delworth and Mann, 2000; Gray et al., 2004] also exerts a considerable influence on the global and regional climate [Polyakov and Johnson, 2000; Chylek et al., 2006, 2009; Chylek et al., 2010; Chylek et al., 2013; Zhang et al., 2007; Mahajan et al., 2011; Frankcombe and Djikstra, 2011; Zhou and Tung, 2013; Canty et al., 2013; Muller et al., 2013; Kavvada et al., 2013].
In this note we show that the observed annual mean global temperature variability is captured more fully by a regression model when the AMO is added to the set of explanatory variables. Considering a compromise between accuracy and complexity, the minimal regression model that accounts for 93% of the observed annual mean global temperature variance contains only two explanatory variables: anthropogenic greenhouse gases (GHG) and the AMO. Adding all other predictors increases the fraction of accounted for global temperature variance to 94%.
(a) Radiative forcing due to greenhouse gases (red), solar variability (blue), and volcanic aerosol (black); (b) four considered models of anthropogenic aerosol radiative forcing; and (c) the observed mean global temperature (GLT) and the regression model temperatures without the AMO among the predictors (M1) and with the AMO (M1 + AMO). (d) The AMO index and residual of the regression model without AMO among the explanatory variables. (e) The division of the observed temperature variability between the statistically significant predictors. (f) The division of the observed temperature variability in two predictor models between the GHG and the AMO.
Summary and Discussion
A multiple linear regression model that uses the set of explanatory variables composed of radiative forcing due to anthropogenic greenhouse gases and aerosols, solar variability, volcanic eruptions, and ENSO accounts for 89 ± 1% of the global annual mean temperature variance. When AMO is added to the set of explanatory variables, the fraction of explained temperature variance increases to 94%. Just two explanatory variables (GHG and AMO) still account for 93% of the temperature variance. The improvement of the regression model by including the AMO is highly statistically significant (p < 0.01). For comparison, the CMIP5 ensemble mean of all simulations accounts for 87% of the observed temperature variance.
Our analysis suggests that about two thirds of the late twentieth century warming has been due to anthropogenic influences and about one third due to the AMO. This is a robust result independent of the parameterization of the anthropogenic aerosol radiative forcing used or of the considered regression model, as long as the AMO is among the explanatory variables.
An inverse structural analysis shows that all considered climate models (GFDL-CM3, HadGEM-ES, CCSM4, CanESM2, and GISS-E2) overestimate GHG warming that is then compensated by an overestimated aerosol cooling. The overestimates are especially large in models with an indirect aerosol effect. In these models a strong aerosol effect generates the AMO-like 20th century temperature variability. The apparent agreement with the observed temperature variability is achieved by two compensating errors: overestimation of GHG warming and aerosol cooling. This raises a question of reliability of these models’ projections of future global temperature. The inverse structural analysis underscores the significance of the AMO-like oscillation and therefore the need to establish its origin and to better simulate it in future climate models.
It is available as open access PDF here: http://onlinelibrary.wiley.com/doi/10.1002/2014GL059274/pdf
And as HTML here: http://onlinelibrary.wiley.com/enhanced/doi/10.1002/2014GL059274/
Also of interest:
CMIP5 multi-model hindcasts for the mid-1970s shift and early 2000s hiatus and predictions for 2016–2035
Gerald A. Meehl* and Haiyan Teng
Article first published online: 7 MAR 2014 DOI: 10.1002/2014GL059256
Abstract
Compared to uninitialized climate change projections, a multi-model ensemble from the CMIP5 10 year decadal prediction experiments produces more warming during the mid-1970s climate shift and less warming in the early 2000s hiatus in both the tropical Indo-Pacific region and globally averaged surface air temperature (TAS) in closer agreement with observations. Assuming bias in TAS has stabilized in the 10 year predictions, after bias adjustment, TAS anomalies for the 2016–2035 period in the 30 year predictions initialized in 2006 are about 16% less than the uninitialized projections. One contributing factor for the improved climate simulation is the bias adjustment, which corrects the models’ systematic errors and higher-than-observed decadal warming trend. Another important factor is the initialization with observations which constrains the ocean such that the starting points of the initialized simulations are close to the observed initial states.
http://onlinelibrary.wiley.com/doi/10.1002/2014GL059256/abstract
“An inverse structural analysis suggests that these models generally overestimate the greenhouse gases-induced warming, which is then compensated by an overestimate of anthropogenic aerosol cooling.”
OK, they’re starting to move in the right direction.
Now they need to account for fact that volcanic aerosols have already been severely ‘tuned’ down before their effects were “over-estimated” .
The only way to explain that without frigging the input data is for there to be a strong negative feedback to changes in radiative forcing. Once they get that right the effect of CO2 is pretty small and the real problem stands out: there is a long term, centennial rise in temperature which is being spuriously correlated with CO2 because there is nothing else in their list of variables which can account for it.
We know temperatures have risen since LIA and that started too early to be attributed to CO2. So there is a long term natural variability that is being correlated against a set of variables that include nothing that could account for it.
By starting their regression in 1900 they mask this problem.
The other thing they are trying to sweep under the carpet is the temperature drop in late 19th c. The temperature records go back further than 1900 , why do they cut off the earlier data? Doesn’t do the right thing to support the CO2 curve ??
JamesS Apr 8 5:25pm asks “Did I miss the point of posting this? “. All sorts of things get posted here because they are interesting, not because they are “party line”. Even papers/articles touting AGW or aspects of it get posted, often without comment, because they are interesting.
anna v Apr 8 9:05 pm says “Climate is one of the first examples of dynamical chaos. Dynamics means that the functional dependence to the assumed linearity of parameters can hold for only a limited interval on the variables, even for smooth dynamic variations as in the example.
This is the basic reason why all these time consuming and expensive computer simulations of climate fail in predicting the future. Fitted by the linear method to the past, real time dynamics takes over and destroys their future predictions.“. Spot on. Correct me if I’m wrong, but AFAIK the “limited interval” is not decades or years or even months, but just a few days.
It’s going to be a long haul. These guys will continue to produce this crap as long as the journals will print it. Comments at the journals may be an important leverage point.
Bob Tisdale, in Mann et al. 2014 the abstract reveals his article is mostly in response to Curry and the “stadium wave” paper…
“Abstract
We estimate the low-frequency internal variability of Northern Hemisphere (NH) mean temperature using observed temperature variations, which include both forced and internal variability components, and several alternative model simulations of the (natural + anthropogenic) forced component alone. We then generate an ensemble of alternative historical temperature histories based on the statistics of the estimated internal variability. Using this ensemble, we show, firstly, that recent NH mean temperatures fall within the range of expected multidecadal variability. Using the synthetic temperature histories, we also show that certain procedures used in past studies to estimate internal variability, and in particular, an internal multidecadal oscillation termed the “Atlantic Multidecadal Oscillation” or “AMO”, fail to isolate the true internal variability when it is a priori known. Such procedures yield an AMO signal with an inflated amplitude and biased phase, attributing some of the recent NH mean temperature rise to the AMO. The true AMO signal, instead, appears likely to have been in a cooling phase in recent decades, offsetting some of the anthropogenic warming. Claims of multidecadal “stadium wave” patterns of variation across multiple climate indices are also shown to likely be an artifact of this flawed procedure for isolating putative climate oscillations.”
Yet another model that presumes discarded variables and functions had no affect on the outcome, rendering its predictive value useless?
Eliza
“Do you see warming???”
————————————–
yes
if you decontextualise the last 30 years it looks like you could have an up trendline [except for the last bit]. Then if you project 100 yr prediction lines on that trendline with a few invented accelerator feedbacks then we all going to fry like its venus 🙂
See this article on the role of sun and volcanoes with respect to the temperatures in the Atlantic. It addresses this very issue.
http://www.nature.com/ncomms/2014/140225/ncomms4323/full/ncomms4323.html
Hey, I added the color of my flashlight to the explanatory variables and see, there was a contribution of 1.43785×10-25 to the forcing. So that paper is wrong! Anthropogenic effects account for almost 100% of the warming.
[/sarc]
He’s a good lad is our Leif .
Eliza
i put some trendlines and extended them to make prediction lines .
Black= on road to boiling
Blue =ice age
pretty easy to be a headless chicken several times with that chart
http://jauntycyclist.wordpress.com/
Santa Baby says: April 8, 2014 at 5:28 pm
Two thirds of the warming is caused by antroproghenic adjustment of the data?”
Bingo.
I think it’s more like one third to one half, but they’ve adjusted the data just as you’ve adjusted the spelling.
It appears they are finally getting around to admit they may have “missed” some factors in their previous models. But they are still hung up on their biases.
I just have to laugh. All this AMO, PDO, Jetstream, solar cycle stuff does is illustrate the inaccuracy of our present method of estimating the current Global average temperature. Shifting energy within the system should have zero impact on the global average temperature. The error bars should be twice as large as the affects these cyclical changes have on the system.
Of course that would mean the AGW’ers have no useful data to play with, but I can live with that.
The question is, what is driving the 60 year oscillation in temperatures then?
The Sun, GHGs, randomness, ocean circulation patterns like the AMO.
One of these appears to be a better explanation.
Are we saying that (insert something here) is such a large factor that it could halt warming for around 17 years?
So, why wasn’t it built in to a single model, will all models now be updated to now include this feature and can all previous climate predictions be scrapped?
Almost starting to look like there will be a strong summer El Nino.
(The summer ones have a tendency to weaken rapidly/blow themselves apart at least in the last several decades, so I wonder if the same thing will happen).
This is the most important thing happening in the climate right now.
http://www.cpc.ncep.noaa.gov/products/GODAS/pent_gif/xz/movie.temp.0n.gif
http://www.cpc.ncep.noaa.gov/products/GODAS/pent_gif/xy/movie.h300.gif
http://www.cpc.ncep.noaa.gov/products/analysis_monitoring/enso_update/heat-last-year.gif
http://www7320.nrlssc.navy.mil/GLBhycom1-12/navo/equpacsec_nowcast_anim30d.gif
I still don’t understand why the “Climate Models” leave out water vapor as a major radiative forcing agent. I know, water vapor molecule is short lived before it is rained out but an other one takes its place, but the radiative forcing didn’t stop. I bet if water vapor was part of the “Climate Models” CO2 radiative forcing would be hidden in the error bars.
“Our analysis suggests that about two thirds of the late twentieth century warming has been due to anthropogenic influences and about one third due to the AMO.”
I see around a 0.3°C rise from 1965 to 1995, and from 1995 to 2005, again around a 0.3°C rise. So it may be that the AMO raised global mean surface temp by an extra 0.2°C from 1995, which is one third of the total 0.6°C: http://www.woodfortrees.org/plot/hadcrut4gl/from:1965
That’s impressive for an increase in negative AO&NAO episodes, and definitely not what the models expected to happen to the AO&NAO from presumed increased forcing:
http://www.ipcc.ch/publications_and_data/ar4/wg1/en/ch10s10-3-5-6.html
anna v says:
April 8, 2014 at 9:05 pm
You point out another instance in which past is NOT prologue.
In the post itself, I was most interested in the following statement made by the authors in their abstract:
“In comparison, the Coupled Models Intercomparison Project Phase 5 (CMIP5) ensemble mean accounts for 87% of the observed global mean temperature variance. Some of the CMIP5 models mimic the AMO-like oscillation by a strong aerosol effect. These models simulate the twentieth century AMO-like cycle with correct timing in each individual simulation. An inverse structural analysis suggests that these models generally overestimate the greenhouse gases-induced warming, which is then compensated by an overestimate of anthropogenic aerosol cooling.”
That would seem to contradict any conclusion that AMO accounts for anything; it also clearly (if only indirectly) hits on the central flaw of ALL of the climate models of the past 30 years. Which is the hazard of predicting both weather and climate. Making linear assumptions in a dynamic system.
So….two more papers which suggest that the models have overestimated temperature increases, and thus climate sensitivity is lower.
Can’t wait to see the complete lack of coverage in much to most of the MSM. MAYBE, just MAYBE, Andrew Revkin, who does have a bit of spine — enough of one to get the warmists angry with him, even though he (like me) does think CO2 causes some amount of warming — might cover these.
When it comes to the models, it seems to me that they have tried to tune themselves to a certain amount of historical data that we as humans had collected. The unfortunate part of that, is that historical (or is that hysterical) data accounts for such an infinitesimal percentage of earth’s history, that it’s laughable that we think we can predict/project anything into the future. I kind of had this aha moment earlier. My analogy is this:
Look at climate like an infinite road race. The climate has been running for Billions of years, sometimes fast (Hot) sometimes slow (Cold) and sometimes a nice decent pace with not much change. Just because we have 2,500, maybe 3,000 years of actual data, we think we can build a model, tune it to some of that recent past “running” of the climate and assume that in that small past, we can determine that the climate is going to continue to run at its current pace, whether that is Fast or Slow or on cruise control and we now know everything that constitutes climate.
HA, that is the most egotistical thing I have ever heard. Just like driving down the highway going 60 MPH and a car passes you going 70 MPH. You assume in one hour he’s 10 miles ahead of you, and suddenly you pass him on the side of the road…… huh, must have been one of those “unknown” variables!
That’s my layman’s two cents worth.
Were the temperatures derived from thermometers or satellites? If thermometers then have we accounted for UHI, ‘necessary adjustments’ and the 70% of the Earth’s surface with no thermometers called oceans?
A natural long-term trend is defined by a proxy factor times the time-integral of the difference between each annual average daily sunspot number and the average sunspot number for a long period (1610-1940). The net surface temperature change of all natural ocean cycles oscillates above and below this trend. The combination calculates average global temperature anomalies (AGT) since before 1900 with a correlation of 95% and credible AGT since the depths of the Little Ice Age. Search keywords AGW unveiled.
Dan Pangburn says:
April 9, 2014 at 10:15 am
Go search it yourself, Dan. I don’t go on blind searches for any man, been there, done that, it doesn’t work. The problem is that whatever I bring back from my search, about half the time they say something like “Oh, no, that’s not what I was talking about” …
If you want to get any credibility here, don’t play the “go search” card. Instead, LINK TO EXACTLY WHAT YOU THINK IS IMPORTANT. I don’t go on wild goose chases for anyone. Search keywords “osculate my fundament”.
w.
Seems like another Terraflop job to me; with a net result of overestimating this, followed by an overestimating of that , which presumably results in an overestimation of nothing.
Next time ANY WUWT reader gets wind of mother nature caught in the act of doing any linear regressing, would they please alert WUWT, so we can all watch that happen.
In a classic letter, sometime in the 1960s- 1970s, someone derived the numerical value of the fine structure constant (actually 1/FSC); which is close to 137, by simply frakking around with numbers.
The derived result was of the form:
1/alpha = (pi^a.b^c.d^e.f^g.h^i)^0.25 where a,b,c,d,e,f,g,h all have small integer values (not necessarily all different).
Alpha is known to parts in 10^8, and this paper computed it simply from whole cloth, to within 60% of the standard deviation of the then best experimental value.
So obviously, the paper had to be correct, because you couldn’t get that close by simply frakking around with numbers.
The fact that NO observables from the physical universe, appeared anywhere in the paper, did not inhibit normally sane people from believing.
Within a month, a couple of computer nerds derived all computed values of the fourth root of the product of pi and four small integers, each to some small integer power, that came within one standard deviation of 1/alpha. The list contained about a dozen values; the best of which was within 30% of the standard deviation from experiment.
Proving that you can in fact “prove” anything you want, by simply frakking around with numbers.
Close agreement with experiment is no proof of causality.
Global Temperatures and radiant emittances are not even linearly related (in theory), so why would anyone do linear regressions of anything climate wise ??
Why does this editor not understand the word emittances ??