Guest essay by Bob Tisdale | Judith Curry and Gavin Schmidt are arguing once again about how much of the global warming we’ve experienced since 1950 is attributable to human-induced global warming. Judith’s argument was presented in her post The 50-50 argument at ClimateEtc (where this morning there were more than 700 comments…wow…so that thread may take a few moments to download.) Gavin’s response can be found at the RealClimate post IPCC attribution statements redux: A response to Judith Curry.
Gavin’s first illustration is described by the caption:
The probability density function for the fraction of warming attributable to human activity (derived from Fig. 10.5 in IPCC AR5). The bulk of the probability is far to the right of the “50%” line, and the peak is around 110%.
I’ve included Gavin’s illustration as my Figure 1.
Figure 1
So the discussion is about the warming rate of global surface temperature anomalies since 1950. Figure 2 presents the global GISS Land-Ocean Temperature Index data for the period of 1950 to 2013. I’m using the GISS data because Gavin was newly promoted to the head of GISS. (BTW, congrats, Gavin.) As illustrated, the global warming rate from 1950 to 2013 is 0.12 deg C/decade, according to the GISS data.
Figure 2
For this discussion, let’s overlook the two hiatus periods during the term of 1950 to 2013…whether they were caused by aerosols or naturally occurring multidecadal variations in known coupled ocean-atmosphere processes, such as the Atlantic Multidecadal Oscillation (AMO) and the dominance of El Niño or La Niña events (ENSO). Let’s also overlook for this discussion any arguments about how much of the warming from the mid-1970s to the turn of the century was caused by manmade greenhouse gases or the naturally occurring multidecadal variations in the AMO and ENSO.
Bottom line, according to Gavin:
The bottom line is that multiple studies indicate with very strong confidence that human activity is the dominant component in the warming of the last 50 to 60 years, and that our best estimates are that pretty much all of the rise is anthropogenic.
Or in other words, all the warming of global surfaces from 1950 to 2013 is caused by anthropogenic sources. Curiously, that’s only a warming rate of +0.12 deg C/decade. He’s not saying that all of the warming, at a higher rate, from the mid-1970s to the turn of the century is anthropogenic. His focus is the period starting in 1950 with the lower warming rate.
HOWEVER
Climate models are not tuned to the period starting in 1950. They are tuned to a cherry-picked period with a much higher warming rate…the period of 1976-2005 according to Mauritsen, et al. (2012) Tuning the Climate of a Global Model [paywalled]. A preprint edition is here. As shown in Figure 3, the period of 1976 to 2005 has a much higher warming rate, about +0.19 deg C/decade. And that’s the starting trend for the long-term projections, not the lower, longer-term trend.
Figure 3
And that’s why, when compared to the observed warming rate for the period of 1950 to 2013, which, according to Gavin, is the period “that our best estimates are that pretty much all of the rise is anthropogenic”, then climate model warming rates appear to go off on a tangent. The modelers have started their projections from a cherry-picked period with a high warming rate.
Figure 4 shows the warming rates for multi-model ensemble-member mean of the CMIP5-archived models using RCP6.0 and RCP8.5 scenarios for the period of 2001-2030. RCP6.0 basically has the same warming rate as the observations from 1976-2005, which is the model tuning period, but that’s much higher than the warming rate from 1950-2013. And the trend of the business-as-usual RCP8.5 scenario seems to be skyrocketing off with no basis in reality.
Figure 4
And in Figure 5, the modeled warming rates for the same scenarios are shown through 2100.
Figure 5
CLOSING
I’ve asked a similar question before: Why would the climate modelers base their projections of global warming on the trends of a cherry-picked period with a high warming rate? The models being out of phase with the longer-term trends exaggerates the doom-and-gloom scenarios, of course.
But we purposely overlooked a couple of things in this post…that there are, in fact, naturally occurring ocean-atmosphere processes that contributed to the warming from the mid-1970s to the turn of the century—ENSO and the AMO. The climate models are not only out of phase with the long-term data, they are out of touch with reality.
SOURCES
The GISS Land-Ocean Temperature Index data are available here, and the CMIP5 climate model outputs are available through the KNMI Climate Explorer, specifically the Monthly CMIP5 scenario runs webpage.
The presumptions and assumptions of Gavin are false.
I really don’t give a damn what Gavin means by 110% . His chart has no Y axis so it means nothing.
Shouldn’t Gavin repay American taxpayers for the time he spends blogging on the job?
Why does the US hire foreigners like Gavin & Kevin, anyway? Aren’t there lots of lower paying yet cushy sinecures on their home islands for such drones?
That made me laugh. Thanks.
Hmmm… Did Gavin actually somehow calculate the probability distribution that his graphic shows, or did it come out of some dark hole of his, as it seems to me?
Gavin? Gavin who?
Isn’t that the bloke who claimed he was too important to debate a real scientist face to face?
And now he’s trying the very old ‘Argumentum ad Verecundiam’ (argument from authority) as if Gavin has any authority or respect left in science that he could use as a basis to support his funny numbers. “Thus sayeth the lord rascal of GISS…”
Gavin is so desperate and eager to support the case for man causing warmth through CO2 emissions that he happily tailors his graphs to a very short period of time. Not forgetting that he must cherry pick.
One would think that such a high placed NOAA executive would utilize all his resources, historical documents and data available. Which makes it very odd that Gavin instead places all of his eggs into a very shallow tippy and flimsy basket.
(Good point and question Nylo!)
Gavin just took over the directorship of GISS from James Hansen the (Venus) CO2 expert. How on Earth can he deviate from his holy party line?
I have yet to see any attempt at modelling the entropy between TOA and BOA. If Trenberth’s calculations are to be believed (given the huge margins of error): If the earth as a system is retaining more energy than previously due to CO2 why would this change the mean temperature at BOA – this response would only be expected from the simplest of systems. Not one in which H2O dominates and exists in all three states simultaneously. If you haven’t determined all possible arrangements of energy in the system why on earth would you look for a signal in average kinetic energy at BOA? And why would you then manipulate the data and claim you have a signal in data where only the most obtuse would expect to see one?
I think that is rather Trenberth’s point — the ocean is in fact retaining a lot of the supposed surplus in energy, buffering the surface temperature. However, the effect is enormously marginal — we are talking hundredths of a degree here as the ocean is huge and has a gazillion times the heat capacity of the atmosphere — and one can, and probably should, question whether there is any way in hell we can resolve this sort of tiny temperature variation from statistical noise, natural drift, instrumental error, the substantial errors that result from undersampling the ocean to depth by a half-dozen orders of magnitude compared to what might reasonably be expected to yield that sort of precision where the ACTUAL record doesn’t even use homogeneous instrumentation and is even now derived from only a few thousand samples.
But if true, it is great news! The ocean could eat the heat for a century and not change temperature by a half degree even in the top kilometer only. And by then, who knows where we’ll be?
rgb
The period of the AMO is about 60 years. If you use this period in a trend analysis the oscillation due the AMO will cancel out and this trend reflects the green house effect. But the temperature rise due to the greenhouse effect is basically nonlinear in time. Therefore a linear trend analysis will run into error, if the period used for the trend calculation is too long. Better you should make a least squares fit to the data with a Fitting function TA(t)= c0+ c1*t+c2*t°2+c3*sin(wt+phi).
I have made this Analysis now. My results for Giss Loti data are listed in the following table. Temperatures are anomalies relative to 1900 (reference interval 1886-1915) in °C. Data (D) and Forecasts (F) are annual means. 5 different assumptions are made: (1) no temperature change, (2) 30 yr linear trend, (3) 60 yr linear trend, (4) TA(t) = c0+c1*t+c2*t^2, and (5) TA(t) = c0+c1*t +c2*t^2+ c3*sin(wt+phi), where ci, w and phi are adjustable parameters.
| (1) | (2) | (3) | (4) | (5)
F Jul 2074 | 0,9 | 1,9 | 1,7 | 2,1 | 1,9
D Jul 2014 | 0,9 | 0,9 | 0,9 | 0,9 | 0,9
F Jul 2014 | 0,2 | 0,6 | 0,6 | 2,0 | 0,4
I have checked the fitting procedure by using data until July 1954 for a 60 yr forecast to 2014. As you see in the third line of the table, best agreement between data and forecast is obtained for the linear trend forecasts. The reason for this is that before 1954 the anomalous temperature rise was small. Therefore the fitting functions (4) and (5) were no good description of reality.
110% certainty = 100% certainty that 10% of the warming MIGHT have been caused by human activity.
All data before 1950 is suspicious for the following reasons
a) no re-calibration done on thermometers
b) human observations (usually 4 x day) versus automatic recording every second
c) after 1970 we changed from thermometers to thermo couples
etc.
better is to stick with data only from the 1970’s
http://blogs.24.com/henryp/2013/02/21/henrys-pool-tables-on-global-warmingcooling/
and to make your own conclusions from those…
Goddard Institute of Space Study
Gavin and every other climate modelers who spends the taxpayer funds that should go towards space study instead on thermometers on the surface of the earth need to be fired for misusing taxpayer funds and the GISS needs to be put back to its true purpose, space.
Gavin is a public leach who misuses taxpayer funds and a the reputation public institution for his own pet issue. He and the rest of the climate modelers at the GISS should have been fired a long long time ago.
I raised the case of the CET summer/winter temperatures trends dichotomy on RealClimate blog, but no viable CO2 explanation was offered, instead it was dismissed as a regional coincidence.
CET instrumental record goes back to the nadir of Little Ice Age.
It shows that summer temperatures have stayed almost constant during the last 350 years, or to be more accurate it rose from 15.1C to 15.45 C, or less than 0.1 C/century.
Meanwhile, the winter temperatures have risen from 3.05C to 4.35C or nearly 0.4C/century.
http://www.vukcevic.talktalk.net/CET-s-w.gif
It is more than clear that these changes have nothing to do with CO2, and that what Gavin Schmidt, and his sorry band of cataclysmates, are advocating is a nonsense.
“cataclysmates” – Word of the Week!
Well done, Vuk.
I hope that the colloquial ‘mate’ in the American english is same as in the south London english.
Not sure – but it definitely is the same in Australian English!
I am just speculating but I wonder whether the reason for the rise of temps in winter in densely populated areas [where the meters are] has to do with [1] the quick removal of snow which would normally deflect light (energy) if we had not interfered with it. In addition there might also be [2] a more noticeable UHI effect in winter. In winter we have inversion here (South Africa- inland) i.e.warm [smokey] layers trapped underneath a cold layer when there is no wind. I don’t know if that could be [3] a factor in Europe as well. I doubt it because it seems to me in Europe there is always wind….
Hi Henry
Met office does adjustments for the UHI factor (TonyB has looked into this).
It is more complex than just UHI, I think is to do with the specifics of the N. Atlantic as I outline here:
Ocean heat transfer of a major consequence is permanently active in the far North Atlantic, to the south west of Iceland (mainly in the winter months) and to the Iceland’s north (throughout the year).
http://www.grida.no/climate/ipcc_tar/slides/04.18.htm
Fact:
Cold Arctic winds remove the surface heat at rates of several hundred watts per square meter (W/m2)
Assumed:
There is a twofold effect of this phenomenon:
1 – rising plum of warm air affects meandering of the polar jet stream, causing short term temperature (weather) variation across the N. Hemisphere temperate region.
2 – wind cooled saline surface waters sink to depths of 1-2000m. This deep water convection is the engines of the oceanic thermohaline conveyor circulation. Changes here have a long term effect, affecting the strength of the north-ward horizontal flow of the Atlantic’s upper warm layer, thereby altering the oceanic poleward heat transport and the distribution of sea surface temperature (SST – AMO), the presumed source of the (climate) natural variability.
a – Intensity of the summers’ variability is of lesser effect, mostly due to the near constant insolation (TSI) across the decades or even centuries, overwhelming any major variability in the external forcing.
b – Extent of the winters’ variability is far greater due to the absence of the solar suppressing factor, with the external forcing having the full effect.
(on the external forcing at another occasion)
This summer / winter dichotomy in the N. Hemisphere’s temperature variability is clearly shown in the CET’s 350 year long instrumental record.
http://www.vukcevic.talktalk.net/CET-s-w.gif
Effect of the CO2, before or since 1950’s if any, it is most likely.
But I doubt that anyone with a fixed agenda would take any notice of it.
correction: If any, effect of the CO2 before or since 1950’s is most unlikely..
Thx for your comment
I checked here in South Africa [inland] where we have this inversion,
trapping smoke [containing large % CO2]
whether winters were getting warmer, due to this….
No such luck, here…
It has to be the snow, UHI, or some GH effect
[note that central England is getting warmer during a global cooling period, due to GH effect]
that is causing warming winters in Europe
Is there a theory which can explain this seasonal difference?
Hi Mark
No, other than what I have written above (it is my hypothesis). When I posted the ‘summer no trend CET’ at Gavin’s real climate, apparently no one was aware of the case. Two well known climate ‘scientists’ Daniel Bailey – Sceptical Science and Grant Foster – Tamino, went as far as to accuse me of fabricating data.
Gavin had to go and look up the CET, and put them right, not that I am much welcomed there since.
So much for the expertise of the ‘climate science’ on whose advise many governments bring in laws and regulations.
Let us say that I construct models that use methane emissions from animals posteriors, orange juice consumption in america, working hours in ireland and other parameters I carefully select to precisely match the temperature record from 1950-2000. I now state that I can prove with near 100% certainty that methane emissions from the posteriors of land animals is nearly 100% the cause of the temperature variation during the 1950-2000 period. I can prove this because I can show that other factors in my model aren’t contributing. Yes, I had to homogenize data and do some renormalization, a little parameterization. Given some millions of dollars I am certain any half decent set of mathematicians can construct such a thing. It’s called curve fitting and it’s been done millions of times.
All Gavin seems to be saying is that he has models that were constructed to show that between 1950-2000 100% of the temperature increase was because of CO2. Therefore because he constructed models to show that it must be so.
He then depends 100% on the legitimacy of the models and the predictions they make outside the 1950-2000 period because his models were fitted to this data and precisely engineered so that 100% of the temperature change can be attributed to Co2. Saying that he can prove that 100% of the temperature change is because of CO2 is trivial because he baked that into the models.
The only thing that matters then is: Do the models work after 2000? If they do he may have a leg to stand on. It would still depend on the accuracy of his models but he would have a leg. Since the models fail to correspond to new data since 2000 it means the models are disproven. Pointing to how good the models work during the period he fit them to is irrelevant because as I pointed out above I could have constructed models based on animal methane emissions. I am simply regurgitating the fact I constructed a model that fit the curve. It has no legitimacy if it doesn’t match NEW data. The only new data we have contradicts the models therefore it is impossible to conclude anything using models which don’t seem to work or any of the assumptions in those models about forcings or other things.
This is just basic science folks. This is just as true if Einsteins general theory of relativity predicted mercuries perihelion would be much different than it was. There is the issue of accuracy of the models. But this is a two sided coin. If they say the models are consistent with the 17 year history and have error bars which put this well within the probability of the models then they would also be saying that the temperature in the future could be virtually anything and they would have no basis for saying their models are able to do anything or prove anything let alone how much co2 contributed to warming from 1975-1998.
I haven’t read the newest post from Gavin, but I read the earlier ones carefully when working on my Big Question essay. And there is clearly an attempt to discount natural variability through som rather vague, convoluted and indirect logic: “The final issue is whether the internal variability of the system on multi-decadal timescales has been properly characterised. For instance, it is possible that all the models grossly underestimate the internal variability, in which case any expected trend due to GHGs would be drowned out in the noise. But there is no positive evidence for this at all – as Hegerl et al point out, the estimates of multi-decadal variability in the models and observational records all overlap within their (substantial) uncertainties (arising from the shortness of the record, and the difficulty in estimating internal variability in the presence of multiple forcings).
So while it is conceivable be that there is a bias, it is currently undetectable, which implies it can’t be that large.”
http://www.realclimate.org/index.php/archives/2012/01/the-ar4-attribution-statement/#sthash.VQ8qhWiN.dpuf
To me the last sentence looks a lot like: “I turned my back on the grizzly bear. Since I can’t see it, it’s currently undetectable, which implies it can’t be that large.”
This is crazy. Look, if the thing you are measuring is so small that you end up in an argument as to whether or not it exists then it is clearly, blatantly, obvious that the thing is too small to worry about.