Tuesday, September 13, 2022

 



CLIMATE SENSITIVITY: AN UNSETTLED ISSUE

Posted by:  on: May 10, 2021

NOTE#1: ECS=EQUILIBRIUM CLIMATE SENSITIVITY, RCS=RELATIVE CLIMATE SENSITIVITY

NOTE#2: THIS POST IS A SUMMARY OF PAPERS ON CLIMATE SENSITIVITY FOUND IN THE LITERATURE

LINK: RCS: https://scholar.google.com/scholar?hl=en&as_sdt=0%2C5&q=RELATIVE+CLIMATE+SENSITIVITY&btnG=

LINK: ECS: https://scholar.google.com/scholar?hl=en&as_sdt=0%2C5&q=CLIMATE+SENSITIVITY&btnG=

EXAMPLE PAPER: ROE AND BAKER

THE SUMMARY OF THE LITERATURE BEGINS HERE

Svante Arrhenius carried out some 10,000 spectral-line calculations by hand and concluded that ECS was about 5 C°. However, he had relied upon what turned out to be defective lunar spectral data. Realizing this, he recalculated ten years later and, in 1906, published a second paper, in which he did what climate “scientists” refuse to do today: he recanted his first paper and published a revised estimate which true-believing Thermageddonites seldom cite. His corrected calculation, published in the then newly-founded Journal of the Royal Nobel Institute, suggested 1.6 C° ECS, including the water-vapor feedback. Guy Stewart Callendar published his own calculation in 1938 and predicted 1.6 C° ECS.

Then there was a sudden jump in predicted ECS. Plass (1956) 3.6 C° ECS. Möller (1963) found ECS=1.5 C° to 9.6 C with the effect on clouds being the creation of uncertainty. Manabe & Wetherald (1975) 2.3 C° ECS. Hansen (1981) 2.8 C° ECS. Hansen (1984) 1.2 C° RCS and 4 C° ECS, implying a feedback fraction of 0.7. In the Hansen tstimony before the U.S. Congress he reported ECS of 3.2 C° on a centennial time scale “for a business-as-usual scenario” but he was way off with observed warming of 1C. In 1988 Michael Schlesinger with general circulation model and feedback fraction of 0.71 found ECS =3.6 C°. Gregory et al. (2002) used a simplified energy-balance method and found the probability distribution, strongly right-skewed. Gregory et al 2002: ECS= 2 C°. Subsequent papers using the energy-balance method yielded ECS to be 1.5-2 C°, much less than the 3.7C to 4C° in the current general-circulation models predict. In Lacis 2010 we find ECS=1.2C but with the expected large system gain factor, effective ECS=4C.

The climate science assumption is that non-condensing greenhouse gases (25% of the total terrestrial greenhouse effect) provide the stable temperature structure that sustains the current levels of atmospheric water vapor and clouds via feedback processes that account for the remaining 75% of the greenhouse effect. This assumption introduces an error. This error is the source of the large warming forecasts of climate scientists and their models.

Unfortunately, the above passage explicitly perpetrates an extraordinary error, universal throughout climatology, which is the reason why modelers expect – and hence their models predict – far larger warming than the direct and simple energy-balance method would suggest. Let us walk round the feedback loop for the preindustrial era. We examine the preindustrial era because when modelers were first trying to estimate the influence of the water-vapor and other feedbacks, none of which can be directly and reliably quantified by measurement or observation, they began with the preindustrial era. For instance, Hansen (1984) says: “… this requirement of energy balance yields [emission temperature] of about 255 K. … the surface temperature is about 288 K, 33 K warmer than emission temperature. … The equilibrium global mean warming of the surface air is about 4 C° … This corresponds to a system-gain factor of 3 to 4, since the no-feedback temperature change required to restore radiative equilibrium with space is 1.2-1.3 C°.

First, let us loop the loop climatology’s way. The direct warming by preindustrial noncondensing greenhouse gases (the condensing gas water vapor is treated as a feedback) is about 8 K, but the total natural greenhouse effect, the difference between the 255 K emission temperature and the 287 K equilibrium global mean surface temperature in 1850 is 32 K. Therefore, climatology’s system-gain factor is 32 / 8, or 4, so that its imagined feedback fraction is 1 – 1/4, or 0.75 – again absurdly high. Thus, 1 K RCS would become 4 K ECS. Now let us loop the loop control theory’s way, first proven by Black (1934) at Bell Labs in New York, and long and conclusively verified in practice. One must not only input the preindustrial reference sensitivity to noncondensing greenhouse gases into the loop via the summative input/output node at the apex of the loop: one must also input the 255 K emission temperature (yellow), which is known as the input signal (the clue is in the name). Then the output from the loop is no longer merely the 32 K natural greenhouse effect: it is the 287 K equilibrium global mean surface temperature in 1850. The system-gain factor is then 287 / (255 + 32), or 1.09, less than a third of climatology’s estimate. The feedback fraction is 1 – 1 / 1.09, or 0.08, less by an order of magnitude than climatology’s estimate. Therefore, contrary to what Hansen, Schlesinger, Lacis and many others imagine, there is no good reason in the preindustrial data to expect that feedback on Earth is unique in the solar system for its magnitude, or that ECS will be anything like the imagined three or four times RCS.

As can be seen in the quotation from Lacis et al that CO2 is the control knob, climatology in fact assumes that the system-gain factor in the industrial era will be about the same as that for the preindustrial era. Therefore, the usual argument against the corrected preindustrial calculation – that it does not allow for inconstancy of the unit feedback response with temperature – is not relevant. Furthermore, a simple energy-balance calculation of ECS using current mainstream industrial-era data in a method entirely distinct from the preindustrial analysis and not dependent upon it in any way comes to the same answer as the corrected preindustrial method: a negligible contribution from feedback response. Accordingly, unit feedback response is approximately constant with temperature, and ECS is little more than the 1.05 K RCS. Why, then, do the models get their predictions so wrong? In the medium term (top of the diagram below), midrange projected anthropogenic medium-term warming per century equivalent was 3.4 K as predicted by IPCC in 1990, but observed warming was only 1.65 K, of which only 70% (Wu et al. 2019), or 1.15 K, was anthropogenic. IPCC’s prediction was thus about thrice subsequently-observation, in line with the error but not with reality.

Since the currently-estimated doubled-CO2 radiative forcing is about the same as predicted radiative forcing from all anthropogenic sources over the 21st century, one can observe in the latest generation of models much the same threefold exaggeration compared with the 1.1 K ECS derivable from current climate data (bottom half of the above diagram), including real-world warming and radiative imbalance, via the energy-balance method. The error of neglecting the large feedback response to emission temperature, and of thus effectively adding it to, and miscounting it as though it were part of, the actually minuscule feedback response to direct greenhouse-gas warming, is elementary and grave. Yet it seems to be universal throughout climatology. Here are just a few statements of it: The American Meteorological Society (AMS, 2000) uses a definition of feedback that likewise overlooks feedback response to the initial state – “A sequence of interactions that determines the response of a system to an initial perturbation”. Soden & Held (2006) also talk of feedbacks responding solely to perturbations, but not also to emission temperature–“Climate models exhibit a large range of sensitivities in response to increased greenhouse gases due to differences in feedback processes that amplify or dampen the initial radiative perturbation.”

Sir John Houghton (2006), then chairman of IPCC’s climate-science working group, was asked why IPCC expected a large anthropogenic warming. Sir John replied that, since preindustrial feedback response accounted for three-quarters of the natural greenhouse effect, so that the preindustrial system-gain factor was , and one would thus expect a system-gain factor of or today. IPCC (2007, ch. 6.1, p. 354) overlooks the large feedback response to the 255 K emission temperature: “For different types of perturbations, the relative magnitudes of the feedbacks can vary substantially.” Roe (2009), like Schlesinger (1988), shows a feedback block diagram with a perturbation ∆R as the only input, and no contribution to feedback response by emission temperature – Yoshimori et al. (2009) say: “The conceptually simplest definition of climate feedback is the processes that result from surface temperature changes, and that result in net radiation changes at the top of the atmosphere (TOA) and consequent surface temperature changes.”

Lacis et al. (2010) repeat the error and explicitly quantify its effect, defining temperature feedback as responding only to changes in the concentration of the preindustrial noncondensing greenhouse gases, but not also to emission temperature itself. This allows an empirical determination of the climate feedback factor [the system-gain factor] as the ratio of the total global flux changeto the flux change that is attributable to the radiative forcing due to the noncondensing greenhouse gases. This empirical determination … implies that Earth’s climate system operates with strong positive feedback that arises from the forcing-induced changes of the condensable species. … noncondensing greenhouse gases constitute the key 25% of the radiative forcing that supports and sustains the entire terrestrial greenhouse effect, the remaining 75% coming as fast feedback contributions from the water vapor and clouds. Schmidt et al. (2010) find the equilibrium doubled-CO2 radiative forcing to be five times the direct forcing: “At the doubled-CO2 equilibrium, the global mean increase in … the total greenhouse effect is ~20 W m-2, significantly larger than the ≥ 3initial forcing and demonstrating the overall effect of the long-wave feedbacks is positive (in this model).”

IPCC (2013, p. 1450) defines what Bates (2016) calls “sensitivity-altering feedback” as responding solely to perturbations, which are mentioned five times, but not also to the input signal, emission temperature: “Climate feedback: An interaction in which a perturbation in one climate quantity causes a change in a second, and the change in the second quantity ultimately leads to an additional change in the first. A negative feedback is one in which the initial perturbation is weakened by the changes it causes; a positive feedback is one in which the initial perturbation is enhanced … the climate quantity that is perturbed is the global mean surface temperature, which in turn causes changes in the global radiation budget. … the initial perturbation can … be externally forced or arise as part of internal variability.” Knutti & Rugenstein (2015) likewise make no mention of base feedback response: “The degree of imbalance at some time following a perturbation can be ascribed to the temperature response itself and changes induced by the temperature response, called feedbacks.”

Dufresne & St.-Lu (2015) say: “The response of the various climatic processes to climate change can amplify (positive feedback) or damp (negative feedback) the initial temperature perturbation.” Heinze et al. (2019) say: “The climate system reacts to changes in forcing through a response. This response can be amplified or damped through positive or negative feedbacks.” Sherwood et al. 2020 also neglect emission temperature as the primary driver of feedback response – “The responses of these [climate system] constituents to warming are termed feedback. The constituents, including atmospheric temperature, water vapor, clouds, and surface ice and snow, are controlled by processes such as radiation, turbulence, condensation, and others. The CO2 radiative forcing and climate feedback may also depend on chemical and biological processes.” The effect of the error is drastic indeed. The system-gain factor and thus ECS is overstated threefold to fourfold; the feedback fraction is overestimated tenfold; and the unit feedback response (i.e., the feedback response per degree of direct warming before accounting for feedback) is overstated 30-fold at midrange and 100-fold at the upper bound of the models’ predictions. The error can be very simply understood by looking at how climatology and control theory would calculate the system-gain factor based on preindustrial data:

Since RCS is little more than 1 K, ECS once the sunshine temperature of 255 K has been added to climatology’s numerator and denominator to calm things down, is little more than the system-gain factor. And that is the end of the “climate emergency”. It was all a mistake. Of course, the models do not incorporate feedback formulism directly. Feedbacks are diagnosed ex post facto from their outputs. Recently an eminent skeptical climatologist, looking at our result, said we ought to have realized from the discrepancy between the models’ estimates of ECS and our own that we must be wrong, because the models were a perfect representation of the climate. It is certainly proving no less difficult to explain the control-theory error to skeptics than it is to the totalitarian faction that profiteers so mightily by the error. Here, then, is how our distinguished co-author, a leading professor of control theory, puts it: Natural quantities are what they are. To define a quantity as the sum of a base signal and its perturbation is a model created by the observer. If the base signal – analogous to the input signal in an electronic circuit – is chosen arbitrarily, the perturbation (the difference between the arbitrarily-chosen baseline and the quantity that is the sum of the baseline and the perturbation) ceases to be a real, physical quantity: it is merely an artefact of a construct that randomly divides a physical quantity into multiple components. However, the real system does not care about the models created by its observer. This can easily be demonstrated by the most important feedback loop of all, the water vapour feedback, where warming causes water to evaporate and the resulting water vapour, a greenhouse gas, forces additional warming.

Climatology defines feedback in such a way that only the perturbation – but not also the base signal, emission temperature – triggers feedback response. The implication is that in climatologists’ view of the climate the sunshine does not evaporate any water. In their models, the 1363.5 W m-2 total solar irradiance does not evaporate a single molecule of water, while the warming caused by just 25 W m-2 of preindustrial forcing by noncondensing greenhouse gases is solely responsible for all the naturally-occurring evaporation of water on earth. This is obvious nonsense. Water cares neither about the source of the heat that evaporates it nor about climatology’s erroneous definitions of feedback. In climatology’s model, the water vapour feedback would cease to work if all the greenhouse gases were removed from the atmosphere. The Sun, through emission temperature, would not evaporate a single molecule of water, because by climatologists’ definition sunshine does not evaporate water. Heat is the same physical quantity, no matter what the source of the heat is. The state of a system can be described by the heat energy it contains, no matter what the source of the heat is. Temperature-induced feedbacks are triggered by various sources of heat. The Sun is the largest such source. Heat originating from solar irradiance follows precisely the same natural laws as heat originating from the greenhouse effect does. All that counts in analysing the behaviour of a physical system is the total heat content, not its original source or sources. Climatology’s models fail to reflect this fact. A model of a natural system must reflect that system’s inner workings, which may not be defined away by any “consensus”. The benchmark for a good model of a real system is not “consensus” but objective reality. The operation of a feedback amplifier in a dynamical system such as the climate (a dynamical system being one that changes its state over time) is long proven theoretically and repeatedly demonstrated in real-world applications, such as control systems for power stations, space shuttles, the flies on the scape-shafts of church-tower clocks, the governors on steam engines, and increased specific humidity with warmer weather in the climate, and the systems that got us to the Moon.

Every control theorist to whom we have shown our results has gotten the point at once. Every climatologist – skeptical as well as Thermageddonite – has wriggled uncomfortably. For control theory is right outside climatology’s skill-set and comfort zone. So let us end with an examination of why it is that the “perfect” models are in reality, and formally, altogether incapable of telling us anything useful whatsoever about how much global warming our industries and enterprises may cause. The models attempt to solve the Navier-Stokes equations using computational fluid dynamics for cells typically 100 km x 100 km x 1 km, in a series of time-steps. Given the surface area of of the Earth and the depth of the troposphere, the equations must be solved over and over again, time-step after time-step, for each of about half a million such cells – in each of which many of the relevant processes, such as Svensmark nucleation, take place at sub-grid scale and are not captured by the models at all. Now the Navier-Stokes equations are notoriously refractory partial differential equations: so intractable, in fact, that no solutions in closed form have yet been found. They can only be solved numerically and, precisely because no closed-form solutions are available, one cannot be sure that the numerical solutions do not contain errors. Here are the Navier-Stokes equations: So troublesome are these equations, and so useful would it be if they could be made more tractable, that the Clay Mathematics Institute is offering a $1 million Millennium Prize to the first person to demonstrate the existence and smoothness (i.e., continuity) of real Navier-Stokes solutions in three dimensions. There is a further grave difficulty with models that proceed in a series of time-steps. As Pat Frank first pointed out in a landmark paper of great ingenuity and perception two years ago – a paper, incidentally, that has not yet met with any peer-reviewed refutation – propagation of uncertainty through the models’ time-steps renders them formally incapable of telling us anything whatsoever about how much or how little global warming we may cause. Whatever other uses the models may have, their global-warming predictions are mere guesswork, and are wholly valueless.

THE UNCERTAINTY ISSUE: The problem is that the uncertainties in key variables are so much larger than the tiny mean anthropogenic signal of less than 0.04 Watts per square meter per year. For instance, the low-cloud fraction is subject to an annual uncertainty of 4 Watts per square meter (derived by averaging over 20 years). Since propagation of uncertainty proceeds in quadrature, this one uncertainty propagates so as to establish on its own an uncertainty envelope of ±15 to ±20 C° over a century. And there are many, many such uncertainties. Therefore, any centennial-scale prediction falling within that envelope of uncertainty is nothing more than a guess plucked out of the air. Here is what the uncertainty propagation of this one variable in just one model looks like. The entire interval of CMIP6 ECS projections falls well within the uncertainty envelope and, therefore, tells us nothing – nothing whatsoever – about how much warming we may cause. It is difficult to convince skeptics and Thermageddonites alike that Pat Frank is right. The reviewers of Dr Frank’s paper did not distinguish between accuracy and precision and did not understand that a temperature uncertainty is not a physical temperature interval but simply a matter of not knowing. Climate science is obsessed with interpreting uncertainty as confidence intervals without the understanding that uncertainty means not knowing. This odd issue in climate science is discussed in a related issue on this site: LINK: https://tambonthongchai.com/2020/04/22/climate-science-uncertainty/

RELATED POST ON CLIMATE SENSITIVITY UNCERTAINTY: https://tambonthongchai.com/2019/05/02/a-history-of-climate-sensitivity/ : WE KNOW THAT THERE IS A CLIMATE SENSITIVITY PARAMETER BUT WE HAVE NO IDEA WHAT ITS VALUE IS AND YET WE CAN FORECAST CLIMATE ARMAGEDDON BASED ON THE VLUE OF THE CLIMATE SENSITIVITY PARAMETER. THIS IS THE MAGIC OF CLIMATE SCIENCE.

climatesensitivityvalues
SENSITIVITY2
climatesensitivityvalues3

CONCLUSION: (1) THE LESS WE KNOW THE MORE THE OPTIONS AND THE GREATER THE VARIBILITY. (2) CONVERSELY, THE GREATER THE VARIABILITY THE LESS WE KNOW. (3) THE FAILURE OF CLIMATE SCIENCE TO ACCEPT THE UNCERTAINTY ISSUE IN ECS DERIVES FROM THEIR FAILURE TO UNDERSTAND UNCERTAIINTY. THIS ISSUE IS DISCUSSED IN A RELATED POST ON THIS SITE: LINK: https://tambonthongchai.com/2020/04/22/climate-science-uncertainty/

Edit this entry.

Leave a Reply

Search


Categories

Archives

Blog at WordPress.com.

:)
Posted by at

No comments:

Post a Comment

Subscribe to: Post Comments (Atom)