Ricochet is the best place on the internet to discuss the issues of the day, either through commenting on posts or writing your own for our active and dynamic community in a fully moderated environment. In addition, the Ricochet Audio Network offers over 50 original podcasts with new episodes released every day.
Adjusting for Publication Bias Reveals True Climate Sensitivity
One of my good friends (who I’ve unsuccessfully been prodding to join Ricochet) writes the underappreciated blog “Grok in Fullness” under the pseudonym Jubal Harshaw. Since he’s refused my brow-beatings, I’m forced to regurgitate his brilliance here.
His most recent post references two articles on climate science. The thesis of his article is that there is statistical bias in prestigious journals with regards to climate science (“I’m shocked, shocked to find that gambling is going on in here!”). Both “Publication Bias in Measuring Climate Sensitivity” and the counter article “No evidence of publication bias in climate change science” actually conclude the same thing, titles notwithstanding. Please go there to see all the lovely details complete with “funnel plots” and intellectual rigor.
But the bias is not the most interesting part for me. The most interesting part is the climate sensitivity conclusion, on which both articles agree. You see, CO2 has a mathematical contribution to the greenhouse effect that amounts to about 1.0 C for every doubling of carbon. It’s logarithmic, which already mitigates the effect of continued burning of fossil fuels. What it all comes down to is what the secondary “forcing” is (mainly the feedback loop of extra water vapor, a powerful greenhouse gas, released into the atmosphere due to increased temperature). Climate alarmists would have you believe the effect of all the other factors is 3x to 6x. It turns out both the bias adjusted factor and the “complete” factor (including the results in obscure journals) came out to about 1.6x.
This, to me, is awesome. Not because it comports with anyone’s particular bias on what they want climate change to be, but because their agreement makes it sound like the truth. Now we might have a solid idea what a doubling of CO2 will cause. Each doubling will cause around a 1.6 C increase in world temperature.
I’m going to leave alone if this is a good thing or a bad thing and just let it sink in with everyone that this is probably the closest to a concrete answer as we’ve ever had to this question. It also comports with the observed increase of 0.8 C with the 46% increase (280 to 410 ppm) since the start of the industrial revolution (a factor of 1.6 climate sensitivity actually predicts a 0.88 C increase).
Now that there’s enough data to have a ballpark idea of climate sensitivity, all of the debate should be able to flow from this probable fact. Use this value early and often (allowing for experimental uncertainty). It’s been pretty obvious for some time that the effect of CO2 is not zero or negative, and it’s also been obvious for some time that the effect isn’t an immediate catastrophe. This result is a good corroborator of common sense.
Published in Science & Technology
Yeah, I’m not budging, and not because I don’t like to respond to valid data and observations, but this field of study has been so thoroughly corrupted, and proven to be corrupted at the most respected and highest levels, that I will never trust any data they now or previously purported to offer for argument.
That is a principled stand and it’s important, because we should never allow this type of politicization and corruption of science.
So, no, they’re liars. Proven liars. Beyond that, their arguments and data don’t make sense. I’m sure that temperatures change from time to time and century to century, but we cannot claim to know within a tenth of a degree how much, when, nor that there is any benefit or danger from any such changes. We most certainly cannot trust them to guide policy changes for a few generations.
Mike,
This is the data when not falsified that shows no temperature increase at all in the last ~20 years. That is 20 years in which the amount of CO2 being dumped into the atmosphere was massively larger than before. As I said, to begin with, there isn’t any empirical corroboration of the whole effect of gross released CO2 on gross atmospheric temperature. In fact the best data that we have shows the hypothesis of global warming as disproved. All of the data that produced wild draconian restrictive measures has been shown to be useless. I sold parts per billion EPA certified analyzers and the calibration systems for them. You can’t imagine how difficult it is to attain a resolution that low and keep the instrument in calibration for continuous measurement. You are right the satellites help but the newer data disproves the hypothesis more than the old.
Jim Lovelock is the man that created the GAIA hypothesis. He doesn’t believe in it anymore and openly said so. If Bill Nye or any of the ‘science’ hustlers actually were responsible for producing science instead of just bloviating about it, they’d get the idea too.
Maybe if we beat the dead horse even harder it will get up and run around the track. Obama did that for the last 8 years.
Regards,
Jim
It has been warming at least since Napoleon lost at Waterloo. Issue is not whether there is warming but how much and whether and what portion can be attributed to human activity in some systematic way. The task is complicated by the fact that the science is grossly politicized and much of the temperature history for the 20th century when data collection began in earnest has been “adjusted” (almost uniformly in preferred directions) in order to make recent warming seem larger.
Attempts to put modern temps in a larger historical context through the use of proxies have generated considerable debate (e.g., the Hockey Stick) because of blatant bias in data set selection and methodology. Alarmists, for example, prefer methods that minimize/deny/localize the Medieval Warming or Roman Warming periods to try to sell the notion of “unprecedented” warming. I think they are losing that argument.
The radiative physics of CO2 forcing is undeniable. The climate effects of land use is also undeniable. However, the notion that CO2 emissions implicate a sensitivity of 2.0+ deg C is increasingly unlikely. The fanciful catalog of alleged horrors from warming (e.g., more hurricanes) was always demonstrably silly.
I think you and Mike H are in agreement, as am I. There has been some warming over past two centuries. How much is attributable to recovery from Little Ice Age and how much is attributable to basic greenhouse gas theory regarding impact of doubling CO2 remains uncertain. However, what we do know from trends so far is that the rate of increase is inconsistent with catastrophic warming theory postulating a series of positive feedbacks leading to a 5-8C increase in the 21st century.
There. Fixed that for you.
I’ve always been less than impressed with the phrase “greenhouse gas”, for the simple reason that CO2 and its “GG” peers behave quite a bit differently from the panes of glass in greenhouse. Glass passes short infra-red frequencies (solar) and reflects longer frequencies (re-radiated). It hardly absorbs any infra-red. Doubling the thickness of the glass in a greenhouse has no appreciable impact on how much infra-red is trapped, though it would slow down conduction losses.
“Greenhouse” gasses, on the other hand, don’t work by reflection. They absorb specific frequencies, whether sourced from the sun or by re-radiation. When you have enough of a particular gas to absorb nearly all of the incident infra-red (we do), whether inbound or outbound, adding more of that gas doesn’t really change how much infra-red you trap, except to extend the number of decimal places (0.9 => 0.99 => 0.999 => 0.9999, etc). What it is expected to do is to trap more inbound infra-red at a higher elevation and more re-radiated infra-red at a lower elevation. Note, a barely significant impact on the earth’s energy balance, but supposedly significant impact on the atmospheric distribution thereof.
Based on this mechanism, the IPCC models make very specific predictions what should be happening to the earth’s temperature profile from ground to stratosphere. The hypothesized positive feedbacks supposedly are due to side effects of the energy distribution, and those effects produce the catastrophic warming, especially the redistribution of water vapour. The US Navy’s weather balloon program, regularly capturing temperature profiles all over the world since the 1950’s, shows none of the changes predicted by the models.
Climate scientists cannot explain the lack of evidence for the expected atmospheric energy redistribution, and cannot explain any of the 20th century’s warming as being driven by CO2, and cannot explain the past 20 years of flat global temperatures in the face of continued exponential growth of global CO2 emissions.
Climate is certainly changing, but there’s zero actual, empirical evidence that climate scientists are right about CO2. The alarmism is a complete and utter fraud.
Yes, some of these comments give me the impression I’m being attributed beliefs that are not my own. The 1 C per doubling of CO2 is simple fact. The possible extra 60% warming due to other effects sounds reasonable. I wouldn’t have been surprised if the number was something else, even if it managed to decrease the effect of carbon.
No matter what the result ends up being it is moot from a policy standpoint. Almost no amount of warming (or cooling) would warrant government intervention. If you don’t believe that this result is meaningful or makes any sense in the first place I don’t really have a problem with that. But for me this is something I’ve been curious about for a while and it’s nice to see a couple different measures converging on the same result.
To be clear, the behavior of the molecular bonds in CO2 with respect to absorbing particular wavelengths of outgoing longwave radiation then radiating that energy out as heat energy is absolutely established. Even demonized “denialists” like Christy, Spencer and Lindzen have endorsed and patiently explained that fact while distinguishing it from the specious assumptions piled onto it by alarmists so that they can concoct a large putative sensitivity factor.
What is very poorly understood is how that heat energy may or may not be amplified or offset by much larger factors such as cloud formation and patterns of ocean heat trapping and release. But the raw physics of CO2 absorption of OLW is not “poorly understood.” Whether there are other factors that mitigate or interfere is not so clearly known which I assume is your main point with which I agree.
I agree that alarmism is a fraud, an ideological project masquerading as scientific fact but precisely because they call people like you and me are routinely dismissed by the ideologues as “science deniers” it is imperative that we never allow imprecision about known aspects of the science to ground such accusations.
Correct, I am not disputing the basic facts of infra-red absorption by CO2.
Partly. I am pointing out that the additional CO2 in the atmosphere is not changing the energy in/out balance of the world — there’s already enough CO2 in the atmosphere to capture virtually all of the long-wavelength IR that matches the CO2 spectrum. Adding more CO2 theoretically moves the absorption closer to the source, but virtually all is still captured. The logarithmic doubling yielding a linear increase in temperature is due to there being very little unabsorbed IR (in the CO2 spectrum). But my point is that the estimates of that factor, which includes the 2nd order effects, all come from models that are built on a 1st order effect that is not observed in real life. The models — based on the 1st order impact of CO2 — are missing something big. The lack of observed change in altitude vs. temperature profile — the first order effect of CO2 absorption of IR — completely invalidates the models built upon CO2 absorption assumptions.
The very term “greenhouse effect” conjures a contrasting image of elevated temperatures within a greenhouse versus normal temperatures outside that greenhouse. But the more honest comparison is between two greenhouses, one with 1/4″ glass panes, and the other with 1/2″ glass panes. The greenhouse with the thicker glass might be a fraction of a degree hotter, and not at all due to IR capture.
With current observations, I’m not convinced the doubling of CO2 in the atmosphere will produce 0.1C increase in global temperatures, much less the 1.0 minimum pitched by the IPCC or the 1.6 proposed by the OP. Observations have simply not matched the hypothesis — the extrapolation of CO2 longwave absorption to atmospheric impact is simply wrong, and the precise reasons remain unknown. The activists masquerading as scientists certainly don’t want to know those reasons, as their little charade will be crashing down around their ears. I’m hoping the Christys, Spencers, and Lindzens of this world keep pushing until we do understand it.
Phil,
Your objection seems very well founded. We have no empirical tested reason to believe the latest round of faulty theoretical fear mongering. Why would anyone assume that draconian massive government regulation is appropriate with such a report.
There should be a new children’s fairy tale. Chicken Little was walking one day and an acorn fell on his head. Chicken got some faulty data together, created a faulty theory about the threat of Oak Trees, and then convinced the local authorities to cut down all the Oak Trees. After Chicken had got half the Oak Trees cut down somebody got hold of his data and showed how stupid the whole thing was. Chicken started screaming denier and hired paid protesters to do violent demonstrations in support of his absurd theory…to be continued.
I hope Chicken finally gets what he really deserves because he’s done a huge amount of damage already and this garbage has got to stop.
Regards,
Jim
I saw some disagreement with your statement by others and I would like to add that maybe there theoretically IS a real global average temperature, but we are completely unable to measure it with a few scattered thermometers around the globe. In fact, I read that we only know the absolute “average temperature” to about plus or minus 3 degrees.
I am skeptical of any warming trend at all because the scientists who add up all these measurements violate one of the most basic laws of statistics – You cannot get a mathematical answer that has more decimal places than your instruments are capable of reading. This is a little hard to explain but I will try:
I take two temperature measurements on Friday and average them together. The first measurement is 2 degrees and the second measurement is 3 degrees. If you average them, the simple answer is 2.5 degrees. Now you would be tempted to make your answer exactly 2.5 degrees, and you would be scientifically WRONG. When a person reads the thermometer, he must make a judgement as to the closest whole number the mercury sits next to. This could have been anywhere between 1.6 and 2.4 actual degrees, but he writes down 2 degrees. The same with the reading of 3 degrees. The accuracy is only good to whole numbers and not to decimal places. The scientifically honest answer would be 2 and 1/2 degrees PLUS OR MINUS 1/2 degree. This is why a figure of 2.0 degrees is considered more accurate than simply 2 degrees. 2.0 implies an accuracy to tenths of a degree.
Now if I were to take my measurements with a thermometer that actually read in tenths of a degree and I came up with measurements of 2.0 and 3.0 then I could honestly average them to an answer of 2.5 degrees, because I was actually measuring in tenths of a degree.
Now I described all that boring stuff because Global Warming Scientists mostly ignore this law when averaging temperatures, and they seem to get away with it. The “plus or minus” zone of uncertainty I was talking about equals just about the entire proposed amount of Global Warming to date, which is around 6/10 of one degree Centigrade in the last 130 years. (Others put it around 8/10)
I’m late to this party but did anyone notice the following quote from the Reckova & Irsova paper?
This is certainly consistent with an eyeball fit to the data in Fig. 3. It would be nice to have the numbers and make a fit.
My impression is that the estimates of climate sensitivity have been decreasing with time, early estimates having been larger than recent ones. It takes a while for values to converge to the true one. This is reminiscent of the evolution of measurements of the elementary charge. Millikan’s value was too low; it took a long time for others to get the right answer when replicating Millikan’s experiment because of confirmation bias.
@skyler is onto something. This publication makes a good case that a mean global temperature is meaningless. Briefly, the argument is that temperature is a thermodynamic intensive variable, so averaging it is meaningless.
Edit: The article linked above was published in the Journal of Non-Equilibrium Thermodynamics 32, p.1 (2007).
The preceding analysis is incorrect for several reasons. If the 2° measurement is uniformly distributed between 1.5 and 2.5 (uniform probability distribution), the variance of that distribution is 1/12, hence the standard deviation is 1/√12 or about 0.29. Likewise, the 3° measurement also has standard deviation 0.29. When those two measurements are combined to make an average, the standard deviations are combined in quadrature, resulting in standard deviation 0.2. When combining many such measurements, one assumes the uncertainties have a Gaussian distribution. This is legitimate because of the Central Limit Theorem. Thus, it is perfectly reasonable to quote an average of many temperature measurements with a much higher precision than any individual measurement.
The Warmists are wrong about a lot of things but this ain’t one of them.
Dr. Lorentz, I want to thank you. I’ve known intuitively much of what is discussed in this publication, but I’m not a good enough mathematician or statistician to explain it in such depth. I find this article to be very persuasive. I’m saving this and I plan to take a lot of time studying it. I’m more committed than ever to not accepting the argument of a global average temperature.
I’ve seen the Scientists invoke the Central Limit Theorem in Global temperature measurements but I don’t think it applies there. To quote from Wikipedia “They all express the fact that a sum of many independent and identically distributed (i.i.d.) random variables, or alternatively, random variables with specific types of dependence, will tend to be distributed according to one of a small set of attractor distributions.”
I am not nearly as acquainted with the math as you seem to be, but my interpretation of this theorem is thus: When you keep measuring the same type of phenomenon over and over, you will eventually get closer to the truth by averaging the measurements, more so than by just taking a few measurements.
In that “scientese” quote from Wikipedia, the important part is “Identically distributed random variables with specific types of dependence.” What that refers to is that all the variables that affect the inaccuracy of the measurements must be constant and specific. This would be totally relevant when doing something very controlled like flipping a coin. If you flip a coin only five times and get four heads and one tail, you might conclude from this very small sample that flipping a coin produces heads 80% of the time. If you continue flipping thousands of times more you will find that it converges closer and closer to a 50-50 split. Using the same coin in the same atmosphere and falling on the same table are pretty constant and unchanging variables.
The trouble with Global Temperature Measurement is that there are thousands of different variables affecting the accuracy that are unique to each measuring station and are not constant. For instance; the vision of the person taking measurements (judging the gap in-between hash marks), the vast variability of different thermometers made by different companies over hundreds of years, how often the thermometers are calibrated or replaced, the robustness of the meteorological station to account for sunlight and man-made heat sources, slight shifts in the wind bringing different temperature air, even the reliability of the temperature reader to be there at the correct time of day. If one person read the same thermometer over and over for years, I think the Central Limit Theorem would then apply.
When you are talking about tenths and hundredths of one degree (and that is often the case when they say the Globe is “roasting, burning, sizzling etc…), almost anything is bound to affect the accuracy at that level. If not for the fact that all the temperature graphs exaggerate the y-axis by more than one hundred times, the temperature curve would be essentially a straight line.
I started reading it too. Am very intrigued, need time to digest. Thanks Dr. Lorentz! Aren’t you the guy that came up with the Lorentz Transformation??
Well, this is going to put some Ricocheteers in a pickle. They are either going to have to study the rebuttals to this paper and reject them, so as to agree that global temperature is meaningless. Either that, or they can continue to argue that mean global temperature is not changing, and/or that a higher mean global temperature would be better than the one we have now. Or, I suppose they could argue out of both sides of their mouths that the number is meaningless and that it needs to be higher. Or better yet, some could choose one side and some the other, and get into a big fight on the order of the Trump wars.
For the life of me, I don’t see how some people who know less about this field than even I do can say that it’s easier to argue the science and math of global warming than the politics of it. We all have experienced how people in positions of authority tend to self-serving and corruption. So why do some of us think it’s easier to argue the science than it is to argue against that?
It’s good that people arguing the science of it all, and that we all follow along as best we can, but to claim that that task is easier? Strange.
The Central Limit Theorem has been found to work very well for yield measurements distributed over agricultural fields. It’s central to the design of agricultural experiments. It’s not just for measuring the same thing over and over.
It is not required that the same person make the readings, only that the errors made are uniformly distributed. Of course, a systematic bias by a person would bias the result but that’s not the argument you are making. Indeed, you have modified your argument already. Your original claim was that it was not possible to average many measurements, each of which has a precision of roughly 1 degree, to arrive at a result with a higher precision. This is manifestly false, as I already demonstrated.
Here’s a simple example. Consider the following ten numbers, rounded to the nearest integer, as you hypothesized:
3 2 3 3 3 2 3 2 3 3
Their average is 2.7. It’s clearly distinguishable from 2 or 3 or 2.5 because there are more 3’s than there are 2’s.
It turns out that these numbers were, in fact, rounded versions of a series of data generated randomly between 1.5 and 3.5. Here are the original numbers:
2.9924 1.66935 3.13528 3.42187 3.32413 2.25803 3.23765 1.74698 2.66004 2.79046
The original numbers had an average of 2.72, while the rounded numbers averaged 2.7. You can see that averaging the rounded numbers is quite successful at estimating the average, even though each one is only known to the nearest integer. According to statistical theory, the integer estimate should be within about 0.1 of the exact average. So we got lucky and it was closer in this case. That’s statistics for ya.
I repeated this simulation with 1000 numbers. I won’t list them here. :) Here are the results:
average of the original numbers (not rounded): 2.52341
average of the rounded numbers: 2.529
Or we can continue to say that the global warming fraud industry is still a fraud and nothing any of them say can be trusted since they are proven frauds.
This vision of the alternatives is too narrow because it relies on a false premise. The measured global mean temperature is meaningful in one trivial sense: it can be compared to the predictions of the numerical simulations of the Earth’s climate (the general circulation models, GCMs). And there’s the rub, because the simulations have not been successful at predicting this statistic. This is important because the GCMs are the sole basis for all the predicted negative consequences of the addition of CO2 and CH4 to the air. The entire edifice of global climate change disaster predictions in the IPCC reports relies on the physical models (GCMs) and probably even weaker economic models.
The credibility of the GCMs is the crux of the matter. And one way to evaluate their credibility is to compare their predictions of this thermodynamically meaningless metric (mean global temperature) to the measured values. The GCMs make other predictions, which also have not fared too well against measurement but temperature is the headline value that everyone wants to talk about.
In case you missed it, the false premise mentioned above is that a thermodynamically meaningless variable can’t be useful in assessing the validity of a prediction methodology, in this case GCMs. If the GCMs are wrong about that, why should we believe them concerning anything else? To put it another way, what is the evidence that GCMs have any predictive value whatsoever?
dr,
Let me ask you the question in a very simple manner. Has any direct causal relationship between gross CO2 released into the atmosphere and global average temperature been conclusively demonstrated? We have been measuring with new generation after new generation of the most sophisticated equipment available for 40 years. Tens of thousands of sites collect & record data continuously around the world plus new satellite scanning capability. Yet any objective person must admit that the answer to my question is an unequivocal no. Millions of jobs lost, trillions of dollars of GNP destroyed, and billions of lives globally beggared all because of an intellectual fetish.
The only reason that this is being debated at all is to coddle the intellectual obsessives. Let’s try to keep the question open so they don’t feel bad. Meanwhile, every man, woman, and child on planet earth are paying for their safe space.
Regards,
Jim
It depends on what you mean by conclusively demonstrated. (Do I sound like Bill Clinton here?) How about this: adding CO2 to the air definitely changes how and where heat is absorbed in the atmosphere. By itself, CO2 has a modest effect. The hysteria comes about because of the much-ballyhooed feedbacks, which amplify the effects on heating the atmosphere through changes in the amount of water vapor in the air. Water vapor is the really important greenhouse gas that prevents Southern California from having the climate of the Antarctic.
This answer probably doesn’t satisfy but, remember, I already expressed my disdain for global average temperature as an especially meaningful statistic. Here’s a better question: Is there a sound basis for predictions made by IPCC Working Group II (effects of climate change on people and other elements of the biosphere) or Working Group III (assessment of mitigation strategies)? If the answer is no to this question then all the policy recommendations, not to mention all the greenie hysterics, are misguided. The answer to this question relies on GCMs, economic models, and biological predictions. GCMs have poor predictive power. Economic models are generally terrible (Black Swan, anyone?). I’m no biologist so I’ll leave that one alone. Among those three, the GCMs are the foundation; everything else is built on top of their predictions.
Not sure how conclusive it could ever be. Whatever warming for whatever recent period is selected from the existing record will not be significantly greater than what could occur as a result of natural variation which variation is itself the product of a large, complex chaotic system. The heavily-marketed notion that the “Consensus” (Peace Be Upon It) has achieved predictive analytic precision is a canard.
It is not unreasonable to assume that significant increases of CO2 will result in warming. It is not reasonable to assume that changes from increased CO2 will be unidirectional, clearly quantifiable, large and cause net economic harm.
It is when you’re trying to disrupt our national economy from making that assumption. It’s entirely unreasonable.
I think this is a clear NO!
Regards,
Jim
I think this is a clear NO!
Regards,
Jim
To everyone,
So the guy goes into his Doctor’s office. The Doctor takes a history. The guy describes a headache he had a couple months ago. He had a cold last month. This month his sinuses acted up a little. The Doctor says he thinks the guy needs open heart surgery.
The guy should:
Regards,
Jim
So my understand for around 38 years we have had accurate global temperature readings. That is when we started to use satellites to measure global temperature’s in the atmosphere. However I have not read the papers but I am not sure how they adjust for sun spot flares which appear to be the biggest variable to long term global temperature.
Brian,
You are absolutely right to bring up this variable and so many others that are unaccounted for which could have a major effect on the outcome. While ignoring variables like this we still get no confirming evidence of a relationship. How can we possibly base a policy on this? A policy that would profoundly damage the economy.
Regards,
Jim