Can We Really Model Climate Change?

September 25, 2021 Updated: September 29, 2021


Hardly a day goes by when we aren’t assailed by claims of doom and gloom over “climate change,” with some well-known politicians warning us that our world will cease to exist as we know it by 2035 (a short 14 years from now!).

Just a few weeks ago, representatives of the government blamed the catastrophic flooding in New York and New Jersey in the aftermath of Tropical Storm Ida on “climate change,” as did another member who earlier blamed another natural weather phenomenon on “climate change.” Apparently, neither understands the difference between “weather” (the cause of fires in California, hurricanes/tropical storms such Ida, tornadoes, etc.) and “climate” (longer-term swings in temperature, precipitation, etc.).

The fact is that, statistically, such “weather” events (e.g., tornadoes) are no more frequent or intense than they have been in the past. Accordingly, you should be excused for asking about the basis of these extravagant claims.

Do the claimants have a magical crystal ball with which they can see the future more clearly than the rest of us, or is it simply the blind leading the blind with some of the blind being more vocal than others? After all, paleontologists and earth scientists will tell you that Earth’s climate has been changing since time immemorial and it has changed drastically and cyclically during the approximately 2.5-million-year history of the human species.

However, at issue is a more subtle phenomenon related to changes induced by the human species itself. Humankind has had that capacity for less than 300 years (since the start of the Industrial Revolution, from the mid-1700s).

But, you may say, 300 years is just a “blink of an eye” within typically climatic (not weather) cycles that occur over thousands, tens of thousands, or even hundreds of thousands of years. For example, 300 years of CO2-producing industrialization represents just 0.3 percent of a 100,000-year ice-age cycle.

To provide the reader with some idea of the complexity of the climate issue, it’s necessary to review a few basic facts upon which we can all agree. The major natural climate cycle is the ice ages (Milankovitch cycles) that are due to variations in the eccentricity of Earth’s orbit about the sun and which now appear every 100,000 years. Earth’s eccentricity, currently, is near its least elliptic (most circular) and is very slowly decreasing, in a cycle that spans about 100,000 years.

However, variation in the eccentricity of Earth’s orbit is considered to be a relatively minor contributor to long-term climate change. There are other factors that need to be considered, including the obliquity (the angle Earth’s axis is tilted with respect to Earth’s orbital plane) and the direction that Earth’s axis of rotation is pointed (“axial precession”).

For about a million years, the obliquity has varied between 22.1 and 24.5 degrees perpendicular to Earth’s orbital plane. That is important because it determines the seasons, and those seasonal swings in temperature can be large (e.g., 80 degrees Fahrenheit in Winnipeg, Manitoba). Currently, the angle (23.4 degrees) is about midway between the extremes and is very slowly decreasing in a cycle of about 41,000 years.

As the angle decreases, the seasons become milder with increasingly warmer winters and cooler summers. Clearly, obliquity is an important contributor to “climate change,” but it’s natural and isn’t induced by human activity.

Earth wobbles on its axis as it rotates (“axial precession”) due to the gravitational influences of the sun and the moon, resulting in Earth bulging at the equator, which affects the ocean tides, for example. The axial precession has a cycle of about 25,772 years. This phenomenon also impacts the long-term cycles of the climate by making seasonal contrasts more extreme or less extreme in opposite hemispheres.

While a discussion of this topic is fascinating in its own right, my purpose in introducing it is to show that Earth’s climate is a superposition of at least three natural cycles related to the mechanics of the solar system. Because these cycles have different phases, it’s possible that they could constructively or destructively interfere, resulting in extremes of temperature that have nothing to do with humans.

Now, superimposed upon these mechanical cycles, we have the impact of the biosphere (sans humans) with its own impact on the climate, including the CO2/O2 cycle of photosynthesis, water respiration, solar energy reflectance, turbulent atmosphere, ocean currents, variations in solar output, and a myriad of other factors.

My point is that the climate is an exceedingly complex physicochemical system, and we must ask the question: Do climate models faithfully include each of these phenomena, in a deterministic manner with sufficient detail that they can be described by the relevant constitutive equations and natural law constraints in a form that renders the predictions reliable (see below)? Or are we again being led by the blind?

To be clear, my goal isn’t to pass judgment on any specific climate model, for that would entail a much deeper analysis than that which I present here. I simply wish to make the reader aware of the stringent conditions that must be met when modeling complex physicochemical systems, such as our climate, the results of which may impact how future multitrillion-dollar investments are made.

Hopefully, this discourse will prompt people to ask the right questions before approving such expenditures.

At the outset, it’s important for the reader to note that science doesn’t advance by consensus (universal agreement). If everyone agrees on something, it may simply mean that they are all wrong. Such is the case before all revolutionary changes in science (e.g., Einstein’s relativity revolution and Planck’s quantum revolution).

Accordingly, when I hear climate change proponents claim that 97 percent of scientists agree that human-induced climate change is real, I cannot help but think back to Albert Einstein and the other scientific revolutionaries and wonder what they might have thought of that pronouncement!

Likewise, science is based upon proof at the minute level of detail and not upon “belief,” which has no place in the scientific lexicon. When I hear someone proclaim “I believe in climate change,” I shudder and want to respond: “Well, prove it to me.”

Two great philosophies exist with respect to prediction: empiricism, which is the philosophy that everything that we can ever know we must have experienced; and determinism, which posits that we can predict the future from the past upon the basis of the known physical laws (“Laws of Nature”). Thus, all scientists collect data that are converted into knowledge, and that knowledge is eventually used to formulate the Laws of Nature that, unlike the Laws of Man, are inviolate and are true under all circumstances everywhere in the universe.

Indeed, I like to define “science” as the process of transitioning from empiricism (what we observe) to determinism (what we know and can predict) upon the formulation of the Natural Laws. Thus, the Natural Laws represent the condensation of all scientific experience so that when we invoke such a law it contains knowledge extending back thousands of years, to before Aristotle and Archimedes.

Impeding the transition from empiricism to determinism is “complexity.” Volumes have been written on complexity, and space doesn’t allow even a cursory review of the subject here.

“Complexity” is like driving on a highway on a foggy night. The fog obscures your vision and allows you to see just a short distance along your path. Now, you switch on your high beams, and lo and behold, you can see much farther. Thus, you have used an instrument (the high beams of your automobile) to allow you to see farther and with greater clarity.

So it is in science; in fact, it is fair to say that the digital computer (our “high beams”) has allowed us to advance science more over the past four decades than science had advanced throughout previous history. In other words, the computer has greatly extended our intellects, which is the role of models!

The development of models in human intellectual pursuits is a very complex subject that extends well beyond this op-ed, but an excellent, somewhat technical review is given by Frigg and Hartmann. I will focus on deterministic models, as their predictive powers are so much greater than those of empirical models since “prediction” is the single most important attribute that climate models claim to possess.

All deterministic models have a common structure, either explicitly or implicitly. All deterministic models must have a theoretical basis that is, itself, based upon observation. These observations may be presented as postulates or assumptions, with postulates being based directly on observation.

Thus, it’s important to note that a theory can be no more valid than the postulates and assumptions upon which it’s based. Also, the postulates must not presuppose the output of the model; that is, the postulates must not accept as an empirical fact that human-induced global warming is occurring.

Thus, if one starts with a postulate that states that the climate is changing and that humans are responsible for that change, the chances are that the model will predict exactly that, but the prediction will be invalid because of the input of even unrecognized bias.

The prediction environment is also problematic, because what is sought is a reliable difference between two large, fluctuating numbers: the climate as we now know it (including human impact) and the unknown climate that might have existed were there no human impact. We can measure how the climate is changing in current time using a variety of techniques, but how do we measure the climate as it might have existed over the same period sans human impact? The short answer is that we cannot!

Unfortunately, only the former is commonly reported in newspapers, giving the impression to non-experts that it is all due to human impact. It isn’t, and the human impact component is normally only minor, but nevertheless, it’s an important component. The headlines wouldn’t be as dramatic or fear-inducing if the readers were reminded that the human component is the difference between what we observe and what we model in the absence of human impact, and that is normally small.

Thus, we are concerned with changes in temperature due to human impact of a few tenths of a degree Celsius on a background that fluctuates several tens of degrees Celsius daily to monthly (due to the weather) and even more (typically 30 to 45 degrees C) over the seasons of the year.

That isn’t to dismiss the possible seriousness of human-induced climate change if, in fact, it is actually occurring at the rate claimed by the doomsayers. That is the still-unanswered challenge.

There is no question that climate change modeling exists at the very edge of feasible modeling, and it’s for this reason that the nature of the models employed must be critically examined. You see, part of the problem is that the art and science of modeling is seldom taught in universities; somehow, students are expected to know how to model complex physicochemical systems as though it was part of the human genome.

The author became so concerned with the lack of modeling skills among graduate students that he, while a distinguished professor of materials science and engineering at the Pennsylvania State University, taught a course titled “Theories and Models in Science and Engineering.” I don’t recall anyone involved in climate change taking the course.

I began this op-ed with the question “Can We Really Model Climate Change?” and I will end with a considered opinion. The answer is a qualified “yes,” but only, in my humble opinion, if the modelers adhere to certain rules. The modelers must:

  • Carefully describe the theoretical bases of the model, clearly state all postulates and assumptions, and demonstrate that they don’t reflect any preconceived bias toward a certain result.
  • List and describe the constitutive equations and constraints, demonstrate that all equations are independent, and demonstrate that there are a sufficient number of equations to cover all unknowns.
  • Refrain from introducing “ad hoc” factors (“crook’s constant” in my student days) to make the model “work.”
  • Ensure that any calibrating data are known from independent experiments and are known to be true within well-established bounds.
  • Carefully define “success,” and ensure that the scientific method of prediction and assessment is strictly adhered to, including the rejection of the model if only one incorrect prediction is made that cannot be corrected by valid reassessment of model parameters and input data.

I emphasize that the “rules” of modeling outlined above have been established by numerous scientific philosophers over thousands of years of collective work and have been instilled into the transition from empiricism to determinism that I call “science.”

As the saying goes, “The devil is in the details.” The emphasis on the natural laws renders these rules in compliance with previous scientific experience and must be recognized in any attempt to model complex physicochemical systems in a deterministic manner, including (and especially) Earth’s climate.

To ignore them is to do so at your (and our) peril.

Views expressed in this article are the opinions of the author and do not necessarily reflect the views of The Epoch Times.

Digby D. Macdonald is a native of New Zealand, a naturalized US citizen, and is a Professor in Residence (semi-retired) in the Departments of Nuclear Engineering and Materials Science and Engineering at the University of California at Berkeley. He holds B.Sc. and M.Sc. degrees from the University of Auckland and a Ph. D. from the University of Calgary (1969), all in Chemistry. Prof. Macdonald has published more than 1100 papers in peer-reviewed journals and conference proceedings and has published four books. He is a Fellow of the Royal Society of Canada, the Royal Society of New Zealand (the “National Academies” of those countries and is a Member of the EU Academy of Sciences. He enjoys a H-index of 79 and his papers have been cited over 27,846 times.