Patrice Poyet has just published a new 431-page eBook entitled, The Rational Climate e-Book, it is free to download here. Dr. Poyet studied geochemistry, remote sensing, and computer science at Ecole des Mines de Paris / Nice University. He received his doctorate in 1986. As an expert computer modeler, he spends much of the book evaluating climate computer models and uncovers their often-unstated underlying assumptions.
English is not Poyet’s first language and some of his phrasing is awkward, but even so, the narrative is compelling and interesting. You stop noticing the odd sentence structure very quickly.
Poyet declares that climate science is now the religion of our time. He then notes that cooling is far more hazardous than warming and quotes a passage from Trevelyan’s A Shortened History of England:
“The last half dozen years of Williams’s reign (i.e., the 1690s) had been the ‘dear years’ of Scottish memory, six consecutive seasons of disastrous weather when the harvest would not ripen. The country had not the means to buy food from abroad, so the people had laid themselves down and died. Many parishes had been reduced to a half or a third of their inhabitants.” (Trevelyan, 1942, p. 432).
This was the coldest period of the Little Ice Age. During this time people not only died due to cold and drought in Scotland, but also in China and many other parts of the world (May, 2020c, pp. 26-27). Poyet, notes as many others have, that the climate models used to supposedly “prove” humans are controlling the global climate have not successfully predicted anything. The lack of predictive skill invalidates them.
Chapter 2 reviews atmospheric physics and the carbon cycle. Here he does an interesting calculation that shows that only 6% of the CO2 entering the atmosphere is from fossil fuels. The additional CO2 from both nature and fossil fuels has encouraged more plant growth. Net primary productivity of plants has increased 33% since 1900 and continues to increase with additional CO2. We put about 5 ppm/year (molecules per million) into the atmosphere using fossil fuels. Recent warming has increased out-gassing of CO2 from nature to about 80 ppm/year, for a total add of 85 ppm per year. The net increase in the Mauna Loa records each year is less than 2 ppm. We cannot tell which CO2 molecules are taken up by plants and which are not, but plants prefer the carbon isotope 12C over 13C and fossil fuels have more 12C per unit volume than the atmosphere, so fossil fuel CO2 is preferred by plants.
The 13C to 12C average ratio in coal, oil, and natural gas relative to the atmosphere is about -29 ‰ (parts per thousand). Monitoring the change in the isotopes in the atmosphere through time allows us to show that the average residence time of a CO2 molecule in the atmosphere is four to five years. This is dramatically different than what the IPCC reports in AR5, where they say:
“About half of the [fossil fuel] emissions remained in the atmosphere (240 ± 10 PgC) since 1750.” (IPCC, 2013, p. 467).
They also say:
The removal of human-emitted CO2 from the atmosphere by natural processes will take a few hundred thousand years (high confidence). (IPCC, 2013, p. 469).
Neither of these statements fit the Mauna Loa CO2 observations. Using Poyet’s calculations, the ratio of 13C to 12C has declined by 7 ‰. The above statements would require a drop of 13 ‰ or more. Hermann Harde came to the same conclusion in 2019 (Harde, 2019). He examined the components of the IPCC calculations and concluded that humans are not the primary cause of the recent increase in CO2, the dominant cause is recent warming. Harde shows that human fossil fuel emissions contributed no more than 17 ppm (15%) of the estimated CO2 increase of 113 ppm since 1750. As Poyet says, the IPCC calculation is both inaccurate and deceptive.
The IPCC focusses on the amount of CO2 that is stored in rocks, especially carbonates, over geological time and ignores intermediate storage. The atmosphere only contains about 2% of the total surface CO2. We define surface as from the ocean floor to the top of the atmosphere. Most of the rest is in the oceans and in the muddy sediments below the ocean water, which we include in the surface if the mud is in full communication with the ocean water above it.
Poyet shows that his calculation of a 5-year CO2 atmospheric residence time can be integrated easily to show that, over the period from 1959 to 2018, where we have good data, only 52 Gt-C (gigatons of carbon) or 6% of total atmospheric CO2 and 14% of the CO2 emissions are left at the end of 2018, using emissions according to OurWorldinData.org.
A note on units. Poyet uses Gt-C which is gigatonnes of carbon. OurWorldInData.org uses millions of tonnes of CO2, divide their values by 3,664 to get values comparable to Poyet’s.
Poyet explains why the IPCC models of human CO2 emissions make no sense. The IPCC Bern function assumes an initial static equilibrium between the four main reservoir of CO2, the oceans (39,000 Gt-C), soils (3,700 Gt-C, including permafrost), atmosphere (869 Gt-C), and terrestrial vegetation (600 Gt-C). This assumption completely ignores the huge and variable flow of CO2 from the upwelling deep-ocean waters in the Southern, Pacific and Indian Oceans. See this post for a description and map of the upwelling regions.
In this context, humans emitted ~10 Gt-C in 2018. In the same year, upwelling deep water in the Southern Hemisphere oceans emitted ~275 Gt-C, this value is not constant, it depends upon climate over 1,000 years ago when the water was taken to the deep ocean, the climate, animal, and plant activity as the water was transported around the world to where it ultimately was brought to the surface in the southern oceans, and on the surface temperature and water composition when it reaches the surface. Human emissions are less than 4% of surfacing deep-water emissions, well within the error of the estimated deep-water emissions. The surface ocean emits 100 Gt-C per year, another number that is highly dependent upon animal and plant activity as well as temperature and salinity. Even terrestrial plants and soils emit 75 Gt-C per year, over seven times human emissions. The soil emissions depend upon precipitation, temperature, and soil composition. Ignoring these variable emissions and assuming they all cancel to an equilibrium state prior to the technology age is an unwarranted over-simplification. As Poyet makes clear the Earth is never in a steady state, it is always adapting.
Poyet’s book goes on to discuss many other climate model calculations and assumptions. The author is a former computer modeler and appreciates Poyet’s perspective and the way he shows how each calculation must be programmed, which highlights the implicit assumptions in the calculation. These are assumptions that are not always obvious at first glance. But, when they are properly explained, the reader has an immediate “facepalm” moment and utters “OMG” or worse under his breath.
Section 4 of Chapter 2 is an overview of the idea that a dominant factor in the Earth’s surface temperature is the air pressure at the surface. In this discussion, he believes that convection carries most of the thermal energy absorbed by the surface to higher altitudes, where it can be radiated to space more easily. The point in the atmosphere that determines the surface temperature is the point where radiation emitted primarily by water vapor is more likely than not to make it to space in one hop. This is the height where Earth’s cooling begins, Poyet labels this height TOA, or top of atmosphere. The warming begins here and increases as we approach the surface with what Poyet calls the gravitational lapse rate. Water vapor is the major greenhouse gas and the height where it mostly disappears is critical. Poyet believes that CO2 is an insignificant contributor to surface temperature, and the water vapor content and pressure are far more important.
Poyet is probably correct. Most emissions of infrared radiation (OLR) to space from the troposphere are from water vapor, this is apparent from the fact that Earth’s OLR is linear with surface temperature. If the air were dry, the response would be nonlinear according to the Stefan-Boltzmann law, as explained by Daniel Koll and Timothy Cronin (Koll & Cronin, 2019). In the stratosphere, where the air is dry, most OLR emissions are from CO2.
This is an interesting and controversial section. Many people (skeptics and alarmists alike) often recoil at the idea that surface air pressure and convection dominate the Earth’s surface temperature. They ridicule the so-called “sky dragons” and their ideas. This writer is undecided, their arguments, while not fully developed, perfect or proven, are grounded in physics and must be seriously considered, at least with regard to the popular CO2 portion of the greenhouse effect. The CO2 warming effect, man-made or natural, has never been measured in nature, some have used observations to estimate the maximum CO2 effect (Lewis & Curry, 2018), but that only brackets the effect on the high side, it could be less. In the meantime, no one has claimed that the gas laws do not apply to atmospheres.
The height and density of the atmosphere, especially the height to the OLR emission layer at roughly 300 mb (~30,000 ft or ~9 km on average, but the altitude varies a lot) and the surface air pressure must affect the surface temperature. The speed and efficiency of latent heat transport (evaporation), as well as cloud cover are also important. As Poyet makes clear, it is hard to see how the CO2 concentration matters very much in the lower atmosphere. Downwelling infrared radiation from greenhouse gases exists, especially from clouds, but how much is due to CO2 and how much is emitted by water vapor and condensing water vapor? For more discussion of this idea, with reference to Venus, see here.
Section 6 deals with climate sensitivity, Poyet is in line with Lindzen and Choi (Lindzen & Choi, 2011) and Ferenc Miskolczi [(Miskolczi, 2014) and (Miskolczi, 2010)] that any effect from doubling the CO2 concentration will be small and might even be negative when all feedbacks are factored in. He also found, as have many others, that high-altitude specific humidity has been falling since the 1950s, especially at the critical altitude where net infrared radiation emissions (OLR) to space begin, above 20,000 feet (6,100 meters or 400 mb). If true, it means higher altitude atmospheric water vapor content is not increasing with surface temperature. This negates a critical assumption in the current climate models. Changes in water vapor at 500 mb (18,000 feet) to 300 mb (30,000 feet) altitude have 29 times the effect on OLR as the same change would have on the surface. As noted above, OLR is consistently linear with surface temperature, so the model assumption is probably incorrect.
Section 7 is appropriately entitled “The Greenhouse Mess.” As Poyet and others (Gerlich & Tscheuschner, 2009) have noted, the “greenhouse effect” seems to be whatever the author of the current book or article wants it to be. It seems, with only a very slight exaggeration, that every author has a different definition. Some assume no convection and no scattering of radiation, some assume that the atmosphere or large portions of it are in thermodynamic equilibrium, and still others assume that emissivity equals absorptivity, none of these assumptions are even close to reality. None of the models consider clouds, which have a climate forcing ten times larger than a doubling of CO2. They also mostly ignore the effect of the huge amount of circulating thermal energy in the oceans.
He discusses the importance of water vapor also. The surface of the Earth is nearly opaque to infrared radiation, if it weren’t for water evaporating from the surface, we would all cook. Only 17 meters of tropical air is required to block 80% of the IR emitted from the ground or ocean surface. The escape route for this thermal energy is evaporation and the resulting convection. The convection starts spontaneously because water vapor has a lower density than dry air and rises. Water vapor and condensing water vapor are the primary radiators of thermal energy, they emit some 90% of the Earth’s OLR to outer space.
The next section of the book discusses climate history over geological time. As Poyet makes clear we are living in one of the Earth’s colder times, one of four major ice ages in the past 600 million years. He later discusses sea level, ocean oscillations, like ENSO, glaciers, extreme weather and the myth of ocean acidification. He wraps up Chapter 2 with a discussion of the effect of volcanic eruptions and tectonics on climate.
Early in Chapter 3, the chapter on computer climate models Poyet writes:
“The only certainty I have is: don’t take your computer program for reality and this applies to ‘climate science’ as well.” (Poyet, 2020, p. 222).
As a former computer modeler, this author agrees completely. After working on a computer model for many years, it is easy to trust its results, it is your “baby” after all. Then, it fails you, it comes as a crushing blow, you doubt the observations, you are convinced those telling you that you failed are lying. But, no, it is the model that failed. You must accept this fact, examine the model and try to fix it. The data and observations are the reality and cannot be fixed. This is the lesson that all computer modelers must learn. Poyet provides numerous examples of failed climate modelers claiming the observations are wrong and their models are right. This sort of juvenile behavior never worked for me and it will not work for them.
In this chapter we learn that understanding the movement of water and water vapor are the keys to understanding climate. Unfortunately, climate models cannot calculate the vertical movement of water and water vapor, nor can they predict cloud formation or cloud cover. They also cannot predict or model thunderstorms, which move enormous amounts of thermal energy from the surface to the stratosphere. These unpredictable events are crucial to modeling both climate and weather (Poyet, 2020, p. 223).
It is well known that weather and climate are chaotic systems, which means either weather and climate are chaotic or that the equations we used to describe them are inappropriate or both. Therefore, weather forecasts beyond two weeks are useless. Averaging multiple forecasts does not extend this time limit, the same is true of climate forecasts. The IPCC disagrees with this principle and uses averages of poor models to predict the climate decades into the future, but averaging garbage creates average garbage, not better predictions. Kip Hansen wrote very much the same thing in 2016 (Poyet, 2020, p. 225):
“Averaging 30 results [of models of] chaotic behavior … does not do anything even resembling averaging out natural variability. Averaging 30 chaotic results produces only the average of those particular 30 chaotic results.” (Hansen, 2016)
The IPCC tries to claim that climate predictions are somehow different than the integral of weather over time, but they are not, climate is the integral of weather over time – averaging does nothing to change that. The lack of predictive skill in modern climate models is shown by the UK MET office which has tried to make seasonal forecasts with their model and failed miserably (Poyet, 2020, p. 226). The IPCC, in AR5, claims that the models are correct, even though recent predictions have all failed:
“Projections from previous IPCC assessments can also be directly compared to observations, with the caveat that these projections were not intended to be predictions over the short time scales for which observations are available to date. Unlike shorter lead forecasts, longer-term climate change projections push models into conditions outside the range observed in the historical period used for evaluation.” (IPCC, 2013, p. 825).
In other words, we failed to predict the near-term change in climate, but our predictions for 100 years from now are accurate. Poyet rightly ridicules this nonsense.
Poyet shows us, as have many others, that the “discretization” done to allow the climate models to run on the computer using a global set of “cells” or grid boxes is problematic. These boxes are 100 kilometers on a side and can be over a kilometer thick. Properties and weather in these huge boxes are constant, yet one or two thunderstorms could easily fit inside one of them. Many critics have noted that these boxes are a form of boundary condition on the differential equations being solved and when solving differential equations, the boundary conditions matter much more than the equations themselves (Poyet, 2020, p. 245). Anyone who has walked from the shade of a tree into sunshine, near a lake, knows that the meteorological properties are not uniform across one of these boxes.
Poyet, as well as others, have noted the importance of clouds and water vapor in determining our climate. But neither are well characterized in climate models, as Dr. Mototaka Nakamura (climate modeler, PhD MIT), has written:
“Accurate simulation of cloud is simply impossible in climate models, since it requires calculations of processes at scales smaller than 1 mm. So, clouds are represented with parametric methods in climate models. Are those methods reasonably accurate? No. If one seriously studies the properties of clouds and processes involved in cloud formation and dissipation and compares them with the cloud treatment in climate models, one would most likely be flabbergasted by the perfunctory treatment of clouds in the models. The parametric representations of clouds are ad hoc and are tuned to produce the average cloud cover that somewhat resembles that seen in the current climate. Can we, or should we, expect them to simulate the cloud coverage and properties in the “doubled atmospheric carbon dioxide” scenario with reasonable accuracy? No.” (Nakamura, 2018) the original is in Japanese, this is from an English version also by Nakamura.
Dr. Nakamura has correctly written that climate “models are Mickey Mouse mockeries” of the real world.
Politicians are interested in climate models since they can use them to frighten the public into giving up their individual freedoms and prosperity. Politicians are not interested in the truth, only persuading or intimidating people into supporting their views. Unlike in science, consensus matters to a politician. Poyet quotes Mahatma Gandhi near the end of his book:
“Many people, especially ignorant people, want to punish you for speaking the truth, for being correct, for being you. Never apologize for being correct, or for being years ahead of your time. If you’re right and you know it, speak your mind. Speak your mind. Even if you are a minority of one, the truth is still the truth.” Mahatma Gandhi
This is science. One scientist, with the right data and proper analysis will prevail, even if everyone else disagrees. Science is not consensus and consensus is not science.
Despite the odd sentence structure and occasional awkward phrasing, we recommend the book, and it is free. Add it to your digital library as a reference on climate model interpretation and basic atmospheric physics. I did not check every calculation in the book by any means, but I did thoroughly check the sections discussed in this post and found no problems. Download the book here (Poyet, 2020).
Writing about climate science, especially the “greenhouse effect,” is a minefield. The term is misnamed, the atmosphere does not work like a greenhouse, and the phrase is poorly defined as discussed by Gerhard Gerlich and Ralf Tscheuschner in 2009 (Gerlich & Tscheuschner, 2009). Poyet walks the minefield carefully, and accurately in my opinion.
Download the bibliography here.