Arguments
Answers to the top ten global warming 'skeptic' arguments
Posted on 6 May 2014 by dana1981
Roy Spencer is one of the less than 3% of climate scientists whose research suggests that humans are playing a relatively minimal role in global warming. As one of those rare contrarian climate experts, he's often asked to testify before US Congress and interviewed by media outlets that want to present a 'skeptical' or false balance climate narrative. He's also a rather controversial figure, having made remarks about "global warming Nazis" and said,
"I view my job a little like a legislator, supported by the taxpayer, to protect the interests of the taxpayer and to minimize the role of government."
In any case, as one of those rare contrarian climate scientists, Spencer is in a good position to present the best arguments against the global warming consensus. Conveniently, he recently did just that on his blog, listing what he considers the "Top Ten Good Skeptical Arguments," throwing in an 11th for good measure. He also conveniently posed each of these arguments as questions; it turns out they're all easy to answer.
1) No Recent Warming. If global warming science is so "settled", why did global warming stop 15 years ago, contrary to all "consensus" predictions?
Quite simply, it hasn't. Even global surface temperatures (which is how Spencer is likely measuring 'global warming', although they only account for about 2% of the Earth's warming), have warmed about 0.2°C over the past 15 years, according to the best available measurements. More importantly, the planet has continued to accumulate heat at a rate equivalent to 4 Hiroshima atomic bomb detonations per second over the past 15 years.
2) Natural or Manmade? If we don't know how much of recent warming is natural, then how can we know how much is manmade?
We do.
Net human and natural percent contributions to the observed global surface warming over the past 50-65 years according to Tett et al. 2000 (T00, dark blue), Meehl et al. 2004 (M04, red), Stone et al. 2007 (S07, light green), Lean and Rind 2008 (LR08, purple), Huber and Knutti 2011 (HK11, light blue), Gillett et al. 2012 (G12, orange), Wigley and Santer 2012 (WS12, dark green), and Jones et al. 2013 (J13, pink).
The IPCC stated with 95% confidence that most of the global warming since 1950 is human-caused, with a best estimate that 100% is due to humans over the past 60 years. The IPCC was able to draw this conclusion with such high confidence because that's what the scientific evidence and research clearly and consistently concludes.
3) IPCC Politics and Beliefs. Why does it take a political body (the IPCC) to tell us what scientists "believe"? And when did scientists' "beliefs" translate into proof? And when was scientific truth determined by a vote…especially when those allowed to vote are from the Global Warming Believers Party?
The IPCC merely brings the world's top climate scientists together every 5 to 7 years. It's those scientists who summarize the up-to-date status of the scientific research in their respective fields of expertise. The IPCC report and the 97% expert consensus on human-caused global warming are themselves not proof of anything. They summarize and reflect the scientific evidence – that vast body of evidence is the reason the consensus exists.
4) Climate Models Can't Even Hindcast. How did climate modelers, who already knew the answer, still fail to explain the lack of a significant temperature rise over the last 30+ years? In other words, how to you botch a hindcast?
Global surface temperatures have risen more than 0.5°C over the past 30 years. That rise is significant, both in the statistical sense and the subjective sense. Climate models have accurately reproduced that rise.
5) …But We Should Believe Model Forecasts? Why should we believe model predictions of the future, when they can't even explain the past?
Climate models have accurately reproduced the past, but let's put them aside for a moment. We don't need climate models to project future global warming. We know the planet will warm between about 1.5 and 4.5°C in response to the increased greenhouse effect from a doubling of atmospheric carbon dioxide (the 'climate sensitivity').
In a business-as-usual scenario, atmospheric carbon dioxide levels are expected to surpass 900 ppm by 2100 – that's close to two doublings from the pre-industrial level of 280 ppm. Hence we know that business-as-usual will cause between 2.5 and 7.5°C (most likely 5°C) warming if we stop carbon dioxide levels from rising beyond about 900 ppm. This is based on simple math and what we know about the physics of the climate – no fancy models needed.
6) Modelers Lie About Their "Physics". Why do modelers insist their models are based upon established physics, but then hide the fact that the strong warming their models produce is actually based upon very uncertain "fudge factor" tuning?
Putting aside the accusation that hundreds of climate modelers are all liars – the answer is that their models are indeed based upon well-established physics. Spencer's question likely refers to the uncertain size of the cooling influence of aerosols. However, that is a physical uncertainty. We don't have very good measurements of this effect; unfortunately the rocket carrying NASA's Glory satellite that had instruments to measure the climate effect of aerosols crashed two years ago. Nevertheless, climate models use the available data to account for their influence, and their projections include the associated uncertainties.
7) Is Warming Even Bad? Who decided that a small amount of warming is necessarily a bad thing?
We're headed for about 5°C global surface warming above pre-industrial temperatures by 2100 if we continue on a business-as-usual path. 5°C is the difference between average temperatures now and those during the last ice age. That's not "small" by any stretch of the imagination. As for who decided that amount warming is a bad thing – climate scientists researching the impacts of climate change.
These people labor in such a shabby edifice, like a casino; a structure of misdirection engineered to inspire misplaced confidence and entice gullible commitment but so obviously grubby, shopworn and fraudulent when the lights are turned up.
I would agree with Dr. Spencer on this. These are clearly the best arguments the denialist community has, and they are profoundly weak.
I was even more amused by his list of 10 bad skeptical arguments, in part because I still read all of them being used all over the internet.
How is this man still employed at the university level?
Dr Spencer's list is pretty much the reason why I maintain that self-styled climate science "skeptics" are anything but.
If those are the top ten, climate skepticism is a complete and total bust.
I am always looking into Spencer's site this time of the monh, because UAH are the first to publish a temperature anomaly.
Thanks for this response to his post.
Today, he is pouring scorn on the White House Climate Report announced earlier. But for the first time, I thought his attempts at humourous dissent were so forced that he sounded ... pathetic, even pitiful. Like a Flat Earther who cannot figure how why no one is convinced by his invincible logic.
Tomorrow I might go back to feeling irritated again.
I find it a fascinating comparison to look at Spencers 10 worst 'skeptic' arguments and 10 best 'skeptic' arguments. And yes, I'm using quotes because they are really pseudoskeptical, grasping at any straw to support their views rather than following the evidence.
Quite a difference there. And emblematic of climate denail in general.
Coming from an active climate scientist, arguments like "Is CO2 bad?" are disappointing.
He even accuses fellow scientists of liars that use fudge factors - which in itself is below the behavior expected from a serious scientist. But if he himself resorts to the things he tries to impute to others, then it's downright sarcasm.
Can someone provide a link Spencers 10 worst 'skeptic' arguments, for completeness of the discussion herein? Thanks.
Dr Spencer's "top 10 worst 'skeptic' arguments".
okay ... so before I get my head chewed off, I want to say that if your only reply to me will be how stupid I am, or how it should be clear to me that the models are right, please address that to someone else.
From what I've gathered several prominent folks point to the model are significantly off in their projections. This is not just something that sceptics bring up. Folks who support AGW do so as well. Here is short article published on Nature that says the same thing.
LINK
What do you all think? and when I ask this question I mean thoughtful commentary/criticism
[Dikran Marsupial] Please just ask the question in a calm rational manner, and you will find plenty of people happy to have a rational discussion about the answer, however your defensive tone is unlikely to result in productive discussion. Rather than just point to an article, ask a specific question about it. Link to the paper activated.
[RH] Shorten URL that was breaking page format.
My main reservation about the Fyfe paper is that the statistical test that it uses estimates the plausible variability due to internal climate variability (things like ENSO) from the spread of the model runs around the mean for that model. If the models are unable to predict the forced response of the climate (i.e. the climate change caused by changes in forcings such as CO2) then I don't see how they can be expected to accurately model the unforced response (internal climate variability). IIRC the details of this are in the supplementary material, rather than the paper. Thus the statistical test is not a particularly reliable one, and the authors have not really taken this into account. I prefer the more basic test of seeing whether the observations lie within the spread of the model runs, which seems a little more robust. Essentially either the models over-estimate the warming, or they underestimate the internal variability, or a bit of both.
The fact that mainstream scientists discuss this is of no surprise to me, I work with climatologists every now and again and they spend lots of time criticising models because making them better is what they do for a living, and you can't improve something without being aware of the deficciencies. However, at the same time you need to keep some perspective and look into other reasons for the apparent difference, e.g. by controlling for the effects of ENSO. The mainstream position on this varies from "its the models" or "its internal variability" and everywhere in between, with I suspect "somewhere in between" being considered the most plausible (at least according to the climatologists I have discussed this paper with).
"Thus we employ here an established
technique to estimate the impact of
ENSO on global mean temperature, and
to incorporate the effects of dynamically
induced atmospheric variability and major
explosive volcanic eruptions5,"
are you saying this is not as "established" as they claim it is. Are they publishing such a report on Nature Mag claiming such vasts differences on models vs observation which are just statistical tricks? Is there another method that would make the model right?
Is it not possible to take simple position on the matter? Something like "The models are correct given blah blah blah" or "Yes, the models are off by a significant margin for a 20 year span".
Fyfe and company are claiming model predictions are off by roughly 50%!! 0.30 ± 0.02 °C vs 0.14 ± 0.06 °C .... This is not a small margin!
skeptical101 you need to read the supplementary matrial in detail. I am a statistician, so the point was easier for me to spot than it would be for those with other backgrounds. Using the spread of the models *is* an established method, but that is pretty much because there are no real alternatives as we can't estimate the plausible variability of the climate from the single realisation that we can actually observe (i.e. the Earth's actual climate). However, that doesn't mean that it is robust.
Fyfe et al. is just one paper on this topic, there have been many other groups working to understand the cause of the apparent model-observation difference. IMHO Fyfe's paper overstates the conclusions. The best approach is to look at every paper in the context of the other published on the same topic.
It isn't really possible to take a simple position on the matter without leaving out important caveats. We fundamentally don't know whether the models are underestimating variability or overestimating climate sensitivity or both. We will get more confidence on this as the amount of data increases, but the uncertainty won't go away completely.
It is worth noting that the 1998/99 El-Nino event pushed the observations about as high into the upper tail of the model projections as they are into the lower tail now. Does that mean that the models underestimated climate sensitivity then and overestimate it now?
It is also important to note that the models are not designed to project climate on a decadal basis, but on a centennial basis, where the effects of internal variability can more reasonably be expected to average out. This means that just because the models are struggling to explain recent climate, that does not imply that their centennial projections are unreliable.
In my opinion, it is a case of "a bit of both".
I'm not a statistician so I definitely missed the finer point. From an everyday laymen perspective, where this battle is being waged, evidence, for better or worse, is debated based on the end result.
The models , regardless of what it is they are doing or attempting to account for (be it "variability" or "climate sensitivity"), in the end make a projection. That end number, that projection, from what is being presented to me clearly "seems to be" at odds with observational data.
(Rob P) See this SKS post: Climate Models Show Remarkable Agreement with Recent Surface Warming for a discussion of the CMIP5 climate model projections. And contrary to your claim, note the agreement between observed warming and the models when the models are fed the climate forcing that actually occurred:
If that end projection is in fact anywhere near that far off from observational data, we should not use it as an argument to support AGW.
That is not to say that scientist should not continue to work on them. I recently listened to a debate between Kerry Emanuel and John Christy. When John brought up the model discreptancy issue, Kerry essentially said "the models suck right now, I don't like them but we still have all this other evidence that in context says we are definitely under AGW and it is serious." [my words not Kerrys]
Now, if Kerry is conceiding that point, one which is being made by other respectable scientists (as seen in this paper), that should clearly make one think about bringing up the models as evidence for AGW. From the previous comments on this thread, one would think the assertions by Christy are insane. 20 years is not an insignificant number! According to Kerry, the models also undershoot temperature before that time.
skeptical101 @12:
1) Fyfe et al show the difference between model predictions and observations, but then note that there are several distorting factors in that comparison. The try to characterize those factors, and adjust the results accordingly, with the results shown as the residuals in figure 3:
As can be clearly seen, the adjusted temperature data lies clearly within the 95% confidence interval of the adjusted model data. Hence, based on Fyfe et al, the models have not been falsified. Suggestions to the contrary merely misrepresent their paper.
As a side note, it is worthwhile noting that the only adjustment required for the model data is the volcanic adjustment, which brings the CMIP5 results inline with the CMIP3 results. That shows the difference between them is largely a function of the relative date of the data runs to historical volcanic erruptions. "Skeptical" claims that that discrepancy can only arise from a difference in climate sensitivity are shown, therefore, to be far from skeptical, and to merely read into data the conclusions they wish to find rather than analyze the data on its merits.
2) The adjustment algorithm uses in Fyfe et al was one developed by the CRU team, and so is a reasonable choice of algorithm. However, it is not the only algorithm used for that purpose, and differs substantially in its resulst from other algorithms. This may be, in part, due to its use of successive singular regressions instead of a multiple regression, with the former method being fraught with perils. Fyfe et al also use a method which determines a "forcing" from ENSO from the tropical east Pacific rather than from temperatures. This indirect method may introduce further errors. The end effect is that the adjusted observed trend from Fyfe et al are significantly less than the approx 0.17 C per decade obtained by multiple regression by Foster and Rahmstorf, or the 0.16 C per decade from the intuitive method by Nielsen-Gammon:
3) One way to address which regression gives the most accurate result would be to compare their adjusted trends across all trend intervals to see if they come out the same. If they varied widely the adjustment is not reliable. Unfortunately Fife et al saw fit to only compare across to trend intervals. However, comparing the mean of all trend intervals in a given period should largely average out short term factors such as ENSO and volcanic adjustments. The result will be a fair estimate of the underlying trend. When we do so with CMIP5 and with the three main temperature records we fins an underlying trend for CMIP5 matching that for Fyfe et al, and for CMIP4, and a discrepancy between observations and models far closer to that obtained by Foster and Rahmstorf than that obtained by Fyfe et al. That give greater confidence in the former than the later:
4) All of these methods still leave a small discrepancy between models and observations, most likely in the range of 15-20%. It is probable, however, that at least part of that discrepancy is due to an under reporting of observed trends as shown by Cowtan and Wray, and by the recent Best global (land plus ocean) tempertature record. In all, it is likely that models slightly over estimate the underlying trend, but the over estimate is far less than that indicated by Fyfe et al.
@Skeptical101,
I would like to know if you accept the following fundamental points about this issue:
If you accept these points then the only issue becomes the magnitude of their influence. Without statistical analiyis it is possible to see the possible relationship between ENSO and temperature. The following site shows the measured magitude of El Nino and La Nina.
http://www.cpc.ncep.noaa.gov/products/analysis_monitoring/ensostuff/ensoyears.shtml
There is a pretty clear correlation between the ENSO changes and global average surface temperatures in all temperature sets. Even Dr. Roy Spencer's chart of the temperature values he has interpretted from satellite data points out the 1998 El Nino at that bump.
@Skeptical101,
As a follow-up to my previous questions: After reviewing the pattern of ENSO magintude do you understand that it is a rather random pattern that would not be easily predicted? And if so, do you acknowledge that any modeling into the future would be best done based on an ENSO Neutral condition rather than trying to speculate about the pattern of the ENSO fluctuations?
Another consideration is that the explanation NOAA provides for how they establish the baseline for ONI values points out that the ONI neutral ocean surface in the region the ONI is monitored has been increasing. This is the reason the ONI values table is headed as being based on the latest baseline, with a link to the previous baseline table of values.
Hope this helps you better understand what is going on.
Skeptical101 #14 My interpretation and synopsis of the considerable technical detail and references provided by Tom Curtis #15 & One Planet #16, #17 is that your "...not use it as an argument to support AGW" is correct if used over periods in which short term natural variability influences the trend strongly (<30 years was mentioned sometimes) and, in particular, the models are not able to predict the ENSO conditions at all well. However, over the somewhat longer term of several decades or more it'll all average out and leave the trend. I recall some article about South Pacific Easterly winds strengthening the last "couple decades or so" (?) and pushing more heat down into the South Pacific than might have been expected. I find that graph that's often on SKS per Tom #15(2) to be quite illuminating and I keep trying not to mention that it makes me wonder whether El Nino years are pulling away from La Nina & neutral since ~1991 with El Nino at 0.23 Celsius / decade because I know the data since then is too sparse and varied and definitely <30 years. If I'm right in 30 years hold a seance and tell me about it.
Skeptical101, a recap to some responses above: ... so short trends are currently deemed very unpredictable and the modellers know this. They aim at the more predictable long-term trends (even in hindcasting), where internal variability tends to wash out. As evidence, if you add in the internal variability details to past forcasts (volcanoes, ENSO, PDO, etc), you get decent short-term results. We also note that currently all analyses/studies of the models necessarily are in the short term range, making easy to create strawman arguments (eg, since models don't claim to be accurate short-term) especially those arguments that also ignore the law of large numbers. And Spencer and others further push their "luck" by cherry-picking years like '98 and by ignoring the best (BEST) data sets to focus on rose-colored data sets.