By Rud Istvan, edited by Charles Rotter
WUWT reader Max alerted us to a 1994 Naomi Oreskes et. al. paper published in the prestigious journal Science. Her paper was a critical analysis of Earth Science numerical models.
I asked Rud to take a look, since he had previously written on climate models both here and in the ebook Blowing Smoke. What follows is an edited version of what Rud sent us, approved for publication by him.
After a quick read of Oreskes’s paper, I felt a double whammy was in order:
1. Explain Oreskes ‘science’ per se.
2. And then explain her later duplicitous conversion to rabid climate alarmist.
This is evidenced by her books ‘Merchants of Doubt’ and ‘Why Trust Scientists’.
This post is also another opportunity to restate, yet again, (using Oreskes own early explanations) some of the key problems with IPCC climate models, specifically CMIP5 for AR5.
For those unfamiliar with Oreskes, she received a degree in geology and subsequently became a practicing geologist.
Later, she returned to Cal Berkley for a PhD in history of science. After this she then became a rabid climate change alarmist, as evidenced by her books noted above. She became famous for her Warmunism. I noted this in footnote 24 to essay Climatastrosophistry in my ebook Blowing Smoke, which I based on former Czech president Vaclav Klaus’ 2007 book “Blue Planet in Green Chains”.
Oreskes’s work led to a tenured Harvard professorship. Her intellectual abandonment of her previous work and conclusions about earth science models, as encapsulated in her earlier paper in Science, is indicative of her career/financial turn to the dark side.
Her 1994 Science paper on earth systems numerical models used hydrology and geochemistry examples. We shall quote her reasoning, but substitute climate model examples. This is fair, since her 1994 paper explicitly also included meteorology and oceanography, implicitly including climate models. For extra fun, this guest post 36uses her exactly worded paper major subheadings, albeit in a slightly altered sequence for exposition purposes.
“Verification and validation of numerical models of natural systems is impossible. … Models can only be evaluated in relative terms, and their predictive value is always open to question. The primary value of models is heuristic.”
Uh Oh! Not good for IPCC reliance on climate model projections to prescribe drastic climate policies such as in the Paris Climate Accords. How did Oreskes later writings ‘forget/disavow’ her earlier writings? While she has not explained her transformation; the old adage ‘follow the money’ might.
Verification: The Problem of ‘Truth’
“To say that a model is verified is to say that its truth has been demonstrated, which implies its reliability as a basis for decision-making. However, it is impossible to demonstrate the truth of any proposition, except in a closed system. … Numerical models may include closed mathematical components that may be verifiable. … However, the models that use these components are NEVER closed systems.”
Earth orbits around the Sun, receiving its sunlight energy and reflecting back some 30% proportion based on albedo, This is the the climate model cloud problem, from AR5 WG1 chapter 7. Some further proportion is lost back to space via long wave infrared outbound warming caused by the incoming sunlight energy, modulated by greenhouse gasses including CO2 and H2O. This postulated electromagnetic radiation imbalance is the so-called greenhouse effect, or anthropogenic global warming (AGW). All Earth systems are open systems to space. Not good for ‘sophisticated’ IPCC climate models according to early Oreskes.
“In contrast to the term verification, the term validation does not necessarily denote an establishment of ‘truth’. … Rather, it denotes the establishment of legitimacy. … For all the reasons discussed above, the establishment that a model accurately represents ‘actual processes occurring in a real system’ is not even a theoretical possibility.”
That conclusion is another big problem for Warmunists. IPCC climate models may not represent reality, and Oreskes said there is no way of finding out if they might.
Calibration of Numerical Models
“In the earth sciences, the modeler is commonly faced with the inverse problem. The distribution of the dependent variables is the well-known aspect of the system (guest post comment, e.g. GAST or SLR). The process of tuning the model–that is, the manipulation of the independent variables to obtain a match between the observed and simulated distribution or distributions of a dependent or distributions of dependent variables—is known as calibration.”
This early Oreskes explanation goes to the heart of the climate model ‘parameter tuning’ problem. The typical CMIP5 grid is 280km x 280 km at the equator. But important processes like convection cells (thunderstorms) happen on a 2km to 4km grid. This problem is illustrated by observed/modeled Arizona Thunderstorms.
The large Arizona grid ‘smeared’ meteorological models are effectively useless.
The present climate model problem is that the typical CMIP5 grid square was about 280km x 280km at the equator. The CFL constraint on numerical solutions to partial differential equations means, according to an NCAR rule of thumb, that halving grid size requires a 10X, (one order of magnitude) increase in computational intensity.
An about 7 orders of magnitude computational intractability constraint means such crucial processes as convection cells and Thunderstorms must be parameterized. The official AR5 ‘Experimental Plan’ for CMIP5 required a model initialization at year-end 2006, followed by a mandatory temperature hindcast of three decades to 1976. The problem is that this hindcast covers the entire temperature rise from about 1975 to about 2000. That period’s rise is virtually indistinguishable (both visually and statistically) from an equivalent rise about 1920 to 1945, as MIT Prof. Emeritus Lindzen pointed out.
The Warmunists’ model problem is that the former period rise cannot be attributed to AGW; there simply was not enough increase in CO2. It must be mostly natural, not anthropogenic. The latter period climate model parameter tuning issue necessarily drags in the attribution problem, i.e how much of the latter period is natural instead of AGW. The CMIP5’s attribution to AGW necessarily means all climate models run hot.
“If the predicted distribution of dependent data in a numerical model matches observational data, either in the field or laboratory, then the modeler may be tempted to claim that the model was verified. To do so would be to commit a logical fallacy.”
It is worse than a logical fallacy if the climate models do not match observations. There are two very salient examples: the tropical troposphere, and ECS.
CMIP5 models produce a tropical troposphere hot spot where none exists in reality. This is best illustrated by Dr. John Christy’s 2017 Congressional testimony, oft reproduced in various forms here at WUWT
Several recent papers have covered the ECS observational/model discrepancy, the most rigorous being two from Lewis and Curry. Their first set out methodology and results, their second slightly modified their first based on several insignificant criticisms.
From this sad saga come two ineluctable conclusions:
First, Naomi Oreskes sold her scientific soul to the Devil by joining the Warmunism academic movement after first publishing the opposite in Science.
Second, climate models simply cannot deliver their promised Warmunist goods. Never could, never will.
Credit: Source link