Those who have examined the Wallace, Christy, and d’Aleo (WCD) 2016 and 2017 reports discussed in recent months on this blog may be understandably confused as to how they relate to the numerous climate models used by the UN IPCC and the USEPA over many years in support of their climate alarmism. The methodologies used are entirely different and the conclusions are largely if not totally opposite. The purpose of this post is to try to explain why and reach a judgment as to which one is scientifically valid. Only one is likely to be given the opposite conclusions.
The UN selected two parent organizations for their Intergovernmental Panel on Climate Change (IPCC) effort to show how modern civilization was bound to fail because of catastrophic global temperature increases. (Yes, the UN IPCC effort has never been a neutral one.) These parent organizations were the UN World Meteorological Organization (UNWMO), made up of the national meteorological organizations (such as NOAA in the US) and the UN Environment Programme (UNEP).
The UNWMO and its member national weather services in each country have had extensive experience with General Circulation Models (GCMs) because they have long used them for weather forecasting, which is what they do for a living. So this may have led to a decision to adapt them to the new purpose, climate forecasting, in response to the UN’s interest in this new problem in the late 20th Century. They have continued using this application over several decades until the present day. The GCM models may be the best approach to short-term weather forecasting but are well known to fail for longer-term (even as little as two weeks) weather forecasting.
They are particularly bad for climate forecasting, because it involves long periods of the operation of a coupled, non-linear chaotic system. Mike Jonas has argued that “It is clear that no matter how much more knowledge and effort is put into the climate models, and no matter how many more billions of dollars are poured into them, they can never be used for climate prediction while they retain the same basic structure and methodology.”
In my view the climate GCMs should never have been assumed to be sufficiently reliable for determining climate policy. But that is exactly the opposite of what has been done by the UN, climate alarmist groups, and even the Obama Administration. Even the IPCC itself admits (2007, WG1) that: “We should recognize that we are dealing with a coupled nonlinear chaotic system, and therefore that the long-term prediction of future climate states is not possible.”
So the UN IPCC attempted to undertake the gargantuan task of modelling the climate as a whole by aggregating data from a large number of grid cells covering the planet as part of their GCM methodology. Any new variable used requires additional data for each grid cell and an understanding of exactly how the new variable affects other variables. The results are extremely sensitive to the assumed initial conditions. In doing this Jonas argues that they left out some really important explanatory variables that explain global temperatures according to the results of the WCD reports.
The IPCC climate GCM effort quickly reached the limits of computer power, available budgets, and knowledge of how many important climate variables work. Because of this, lack of interest, or lack of knowledge, Mike Jonas believes major variables were left out of the GCMs, including oceanic oscillations, volcanic eruptions, and changes in the sun. Even a quick glance at global temperature charts over recent decades shows the great importance of oceanic oscillations, particularly the El Nino Southern Oscillations (ENSO) such as El Nino and La Nina. Yet the GCMs apparently left them out, possibly because they had little idea how they functioned or wanted the world to concentrate on their CO2-centric views.
For reasons explained by Jonas, the UN GCM effort was impossible from the start, has been much more complicated than necessary to answer the key questions, and is highly questionable as to their conclusions because of the IPCC’s decision to use GCM climate models. This will continue to be the case as long as these models are the principal tool used. The IPCC has been using a poor tool and not admitting that the tool has little if anything to contribute to solving the alleged problem.
The WCD Econometric Approach
Jonas recommends a top-down rather than the IPCC’s bottom-up GCM approach. The WCD econometric approach would appear to fit that description. Rather than trying to understand all aspects of climate in each grid cell in the world (but leaving out major explanatory variables), they picked out the variables that seemed most likely to affect the one they cared about (tropical and global temperatures), and used available data sets to determine the impact/contribution of each explanatory variable to the variable they cared about. They did not need to understand the physical interactions between every variable in each grid cell since all they wanted to know is how much of an effect each relevant explanatory variable had on tropical and global temperatures. They also used a completely different methodology and discipline, called econometrics, a subdiscipline of economics, not meteorology.
The computer effort required by the WCD approach is vastly smaller and the importance of each explanatory variable is determined by data, not the guesswork of fallible and all too often biased humans in the IPCC GCM efforts. Major explanatory variables are included rather than being left out. This approach does not show exactly how each variable operates, only how much of an effect it has on the critical dependent variable (global temperature). One potential complication is that a simultaneous parameter estimation process must be used, not just a direct least squares, on a single temperature equation. This process is somewhat more difficult to carry out, but in the end was not needed in this particular case because CO2 was shown to have no significant effect on temperature.
In summary, a much better and far simpler approach (if anything was needed at all) by the IPCC would have been that instead of turning to vastly complicated and ill-suited meteorological models and ignoring many of the obvious variables that might actually determine global temperatures (such as solar and volcanic activity, and ENSO oceanic oscillations), would have been to use the available data to determine the extent to which each of these and other variables affect global temperatures using statistical econometric techniques.
The IPCC models do the opposite by using current knowledge to try to determine the relationships between various somewhat known factors influencing climate and ignoring many of the variables of some importance, and then fiddling with the remaining unknowns to “tune” the models to fit historical observations but not necessarily the physics. This is nothing more than guesswork (and probably biased guesswork in this case) and is so unreliable as to be worthless despite the tens of billions of dollars spent on this approach. It has brought unexpected riches to the former meteorological modelers (now called climatologists) but has cost taxpayers dearly in terms of paying for the useless research and the faulty climate policies that are “justified” on the basis of useless and biased GCMs.
So a much more useful approach would have been (if anything, since climate alarmism is basically a non-problem) to turn to a completely different discipline: econometrics, as used in the WCD reports. Unfortunately this may have been precluded by the UN’s original choice of the WMO to pair with the UNEP.
An interesting question is whether the IPCC’s choice of using GCMs may also have been influenced by a guess that this would greatly complicate the task of critics by requiring the use of advanced expertise in many areas to understand the results. It may also have been the view of some climate researchers that this was a golden opportunity to get their climate GCM “play things” funded by taxpayers. Or maybe it was just ignorance on the part of the UNWMO and the IPCC as a whole concerning the usefulness of climate GCMs.
What the WCD Reports and IPCC Models Have Concluded
And what do the WCD reports done to date (on a pro bono basis) conclude: the IPCC GCM conclusions are wrong on the major issue of what effect CO2 has on temperatures. The IPCC models have been used to argue that CO2 is a virtual “control knob” for global temperatures. The WCD reports conclude that changes in CO2 have no statistically significant effects on temperatures. Global warming over the range of data used can be explained by natural factors when they are included as explanatory variables.
So which to believe? The IPCC modelers undertook a hopeless research effort and appear to have left out what turned out to be critical variables in their reports. WCD undertook a doable effort focused on the much needed answers to a few critical questions. The answer is very clear to me.
My Conclusions
I suggest a complete moratorium on climate alarmist-inspired spending until more detailed comparisons can be made and examined by all concerned. I personally believe we know enough to decide now to abandon the use of all GCM climate modeling and all efforts to reduce CO2 emissions.
Tens of trillions of dollars have been wasted in part because of the initial UN decision as to which disciplines should be involved in the research and the decision that the IPCC models were a sufficiently reliable basis on which to base climate policy. And all this should have been known before any work was ever started.
And as long as these misguided UN efforts are continued the world will probably continue to waste about $1.5 trillion per year for almost no benefits on the greatest scam in world history. The US should no longer participate in or fund the hopeless IPCC GCM effort.
[…] Their “evidence” are sophisticated global climate models that they claim can predict future temperatures. The models, of course, are built by the alarmists and, of course, yield the results they desire. They are basically worthless because they are trying to model a coupled non-linear chaotic system (climate) which cannot be usefully mode…. […]
As Dr. Gregg noted, those of us who have been involved in solving coupled integro-differential equations in other areas (electromagnetic theory for me) are well aware of the problems involved in modeling and solution. Without validation measurements or other data which can be used to test the theoretical predictions, the computed results are little more than guesses. To me, the main question for those using various GCM predictions is: Is it possible to prove the predictions wrong with actual data? It appears that, no matter what measurements are made, including the warming hiatus of the past several decades, modelers make excuses as to why the models are not wrong. Looks to me like the old problem of how does one prove a negative when positive proof might always be just around the next bend. Cheers.
[…] Alan Carlin | May 11, 2017 […]
Three types…….
People with some scientific training who realize climate change sounds like it has technical fatal flaws. They also have enough self confidence to think for themselves and not just accept the party line.
Then there are democrats and union members who stick with the party line, regardless of how absurd it is. Many may not have much technical training or they have a lot, but for various reasons are satisfied to stay within the liberal herd.
Third type: people who think climate change sounds like trendy nonsense from academics and other smug selfish people who want to tax or otherwise attack useful industries in the United States.
The first type is most represented here, though some are both 1 and 3.
Having done a significant amount of complex integro- differential modeling of Physics equations, we all realize that, given the linearization and approx techniques used, computational models are FAR from reality. We use them for getting ideas about solution behavior.
When it comes to climate modeling with cells of the order of kilometers and yet significant linearizations are needed, climate models are garbage in, garbage out. They have no real significance and are a detriment to telling the world about the chaotic real behavior of climate,
The current climate change movement has long been a leftist political movement in search of a means to exert global social control. Real science does not even get the back seat, it is dragged along behind covered with dirt. Hopefully people are waking-up and allowing President Trump the means of cleaning the climate gunk out of the Federal Government. We would save our freedoms, economy and trillions of dollars.
Modeling a nonlinear system which includes both known and unknown chaotic events may be fun, but can hardly be taken seriously. But the models have other strange assumptions (WAGs) Wild Ass Guesses, such as the assumption that water vapor feedback is the actual greenhouse gas culprit, creating 2 to 3 times the temperature increase as caused by the corresponding increase n co2 level.
Then there is the unsatisfied cavea which accompanies the ghg hypothesis – the NECESSARY condition that there by a warmer region about 10k above the tropics. Not there. Testimony for that is data from millions of radiosondes taken over the past decade or two. It ain’t there.
Then there is the basic problem. There is NO evidence that co2 level has EVER had any impact on the planet’s global temperature. The ONLY correlation tracking both up and down trends in variation by co2 and temperature show temperature variation happening FIRST, and only hundreds of yearsLATER similar variation in co2 level – the OPPOSITE of what alarmist need. Even if it had been found, it would not have been SUFFICIENT. there still needs to be evidence.
No evidence, not even a correlation (which does not imply causation.) This is probably why Lindzen stated that to be a skeptic the proposition in question must at least be plausible.
This seems like a profound and obvious point once made.
I am no modeler, but weather forecasting is a short term prediction.
It was important in the early history of computing btw. Pioneering computers were justified on grounds of improved weather forecasting. And today if the weather service produces a five day forecast, I am hopeful some of it will be proved right.
Can such a model of weather do much for climate over centuries? Seems very unlikely.
I should look up the meaning of econometrics, bu would venture a guess that it integrates probability into economics. Economists are going to be interested in long term changes in economies, not five day forecasts. The great Milton Friedman began life as a statistician, btw. So it makes plausible sense that econometric models, created by worldly realists, would be deeply respectful of knowns, unknowns, and unknown unknowns, to borrow from Donald Rumsfeld, RAND vet.
It has been wisely said, all models are wrong, some are useful. Models aim to simplify and thus must be wrong to some degree. But some can still be very helpful for specific circumstances, as are short term weather forecasts.
Decades ago, many used to say, garbage in, garbage out. it was a platitude, but true. Today, maybe kids grow up with the faux realities of computer games and are vulnerable to believing Fake News pumped out by politians, Mainstream media, and universities that should actually know better and champion real science. It should not all have to default to Doc. Jeepers!
This kind of basic point about models should have long ago been made by self respecting members of the National Academy of Science. its a national disgrace more if it’s members have not spoken out.
There is a typo mention of “lease” squares. This suggests that you write this stuff yourself without a lot of editors helping out. I like that.
Direct logical writing and thinking. The occasional typo only to be expected. Impressive.
Comment by Alan Carlin: Many thanks for pointing out my typo, which I just fixed. Yes, I write almost everything myself.