A common type of experiment to demonstrate the greenhouse effect is essentially to direct heat lamps at the bottom of two closed jars, one with regular air in it, and one with a higher concentration of $\rm CO_2$. The jar with more $\rm CO_2$ in it heats to a higher temperature, and this is attributed to its infrared-absorbing and back-radiating qualities, i.e. to be a demonstration of the atmospheric greenhouse effect.


However I came across a critique of this class of experiments that the effect is actually due to the fact that CO2 has a worse thermal conductivity than air (from this source, at 300K air has 26.2 milliwatts per meter kelvin of conductivity while carbon dioxide has 16.8).

Link to the article: http and pdf. (Unfortunately it's only in German but DeepL can translate pretty well.)

The authors claim to have performed the experiment with both $\rm CO_2$ and Argon and demonstrated that Argon provided a similar effect as the $\rm CO_2$ -- Argon having a similar conductivity to $\rm CO_2$ but, of course, being non IR-interacting.

These are their results for replicating the effect with CO2:

enter image description here

At minute ~60 they pump $\rm CO_2$ into the container, with a resultant heating of the air inside, while at minute ~260 they pump it back out.

This is their graph demonstrating a comparison of the effect between $\rm CO_2$ and Argon:

enter image description here

That is, they observe a similar effect with Argon.

Their (machine-translated) conclusion:

The noble gas argon is an IR-inactive gas that can neither absorb nor emit thermal radiation. If $\rm CO_2$ and argon show the same warming effect, one must look for the cause outside of thermal radiation. Heavy gases have a lower specific thermal conductivity λ than air (the table in Fig. 4). When these gases are introduced into the tube, they reduce the heat flow within the apparatus. The heavy gases act like an insulating layer. Thus it can be determined:

The Ditfurth experiment does not show the greenhouse effect, but is a phenomenon of heavy gases.

If they did perform the experiment correctly with Argon it seems it definitively demonstrates the point.

Note that the critique is published by EIKE which according to desmog.com is a "climate change denial organization". This doesn't automatically make the critique wrong, but, it must perhaps be taken with a grain of salt. And of course even if the critique is correct, it doesn't invalidate any of the other evidence for the greenhouse effect, it just means this type of experiment is not a valid way to demonstrate it.

  • Is the critique indeed valid? On the face of it it seems sensible, in that it's at least possibly true, but I lack the expertise to critically evaluate the claim.
  • If it's not valid then where does it go wrong?
  • $\begingroup$ It seems to me that you are basically reposting the same question. That is, as far as I know, not welcome on SE. What has "triggered" people on your first attempt is that you were showing a really poorly executed experiment that demonstrates nothing of value. $\endgroup$ Apr 15, 2023 at 22:43
  • $\begingroup$ @FlatterMann: sorry maybe I shouldn’t have posted the trigger warning! In what way is this the same question though? This is about a critique of a different type of experiment entirely than the other question linked to. I edited out the last bullet point to separate them more cleanly. $\endgroup$
    – Cloudyman
    Apr 15, 2023 at 22:46
  • $\begingroup$ I would also remark that whatever you see on YouTube about science is typically very low quality content. That's why we have the peer review concept in science to begin with, to keep the quality up. YouTube is basically just entertainment for the masses. Anybody can post anything for any reason or no reason at all. Personally I wouldn't waste my time on it and I certainly won't do it as part of a SE question, either. Just my two cents. $\endgroup$ Apr 15, 2023 at 22:48
  • 1
    $\begingroup$ Yeah, don't worry about "trigger warnings" as most professional scientists, especially physicists, are only concerned with where the evidence leads. Even if it flies in the face of the public consensus and politics. $\endgroup$
    – joseph h
    Apr 15, 2023 at 22:50
  • 1
    $\begingroup$ Maybe as a general comment: not every system can be scaled to benchtop size. A planetary atmosphere is one that can not be scaled easily, if at all. We can and are conducting a planetary greenhouse gas experiment as we speak, of course. The most reliable readout will be rising sea levels. The icecaps of planets are excellent calorimeters, even if they have a slightly long time constant. $\endgroup$ Apr 15, 2023 at 22:54

3 Answers 3


Yes, the critique is valid in that in most of these jamjar experiments the change is caused by other properties of the gas. Heat is usually lost fastest via convection, which is affected by the density, heat capacity, viscosity, and thermal conductivity through the boundary layer. A different gas has different physical properties, and transfers heat by convection or conduction differently. They also often fail to distinguish the IR radiating from the gas and the IR radiating from the walls of the jar, or the walls of the lab.

However, more importantly, this isn't how the greenhouse effect works in a convective atmosphere anyway! The jamjar experiments are not demonstrating the actual mechanism of the Earth's greenhouse effect, so refuting them doesn't do anything to refute the real mechanism. (The serious scientific disputes are about the feedbacks, not the mechanism.)

See this previous answer here on a similar question: https://physics.stackexchange.com/q/757941

Or read "Water Vapour Feedback and Global Warming", Soden and Held, Annu. Rev. Energy Environ. 2000. 25:441–75. The discussion just before/after figure 1 is the most relevant. https://www.annualreviews.org/doi/pdf/10.1146/annurev.energy.25.1.441


I don't know of any way to demonstrate the planetary greenhouse effect on a laboratory scale. The large difference in atmospheric density between the surface and the stratosphere is an essential part, and you can't reproduce that without a column many kilometers high.

The actual demonstrations are mostly astronomical: almost all of the light we see from the sky originates in some sort of atmosphere, so we have lots of atmospheres to study. Venus is a fine example of a runaway greenhouse effect on a planet similar to Earth. The surfaces of stars are too cool for thermonuclear reactions, but the blanketing effect of their atmospheres keeps their cores much hotter: it's a version of the greenhouse effect.


The way to experimentally answer this question is to combine the calorimetry (the technical name for a heat-capturing experiment) with a measurement of the transmission and absorption of the light through the chamber. Consider that air is transparent, but kilometers of air are not equally transparent at all wavelengths. For example, the USA’s Blue Ridge Mountains turn from blue to green as you approach them from a great distance.

The experiment would be relatively straightforward to do with a well-calibrated spectrometer. You would measure the power per unit wavelength out of the lamp, and measure the transmitted power spectrum through the air- and CO$_2$-containing calorimeters. Then you can compute how much energy per unit wavelength is being absorbed by each vessel, and integrate over the spectrum to find the expected energy transfer.

Without calculating, I don’t know how precise such a spectrometer would need to be. Keep in mind that infrared light was discovered by Newton, by putting thermometers in the output of a prism and noticing that the thermometers below the red end of the visible spectrum still recorded temperature increases. But some types of glass are substantially opaque to infrared light — the greenhouse effect is, after all, named for its similarity to the behavior of a glass greenhouse.

My instinct is that it’s unlikely for a benchtop-sized container of carbon dioxide at atmospheric pressure to be substantially infrared-opaque. That is, again, without calculating.

  • 2
    $\begingroup$ The average thickness of the atmosphere is approx. 8000m. CO2 concentration is currently around 400ppm. That's a total CO2 "thickness" of 8000*400e-6=3.2m. Not quite benchtop size but quite doable. It just doesn't help. The current excess energy that is being deposited at the surface is on the order of 1W/m^2 with an average solar flux of 1kW/m^2 on the surface on a sunny day. Good luck with precision calorimetry, especially since you are neglecting the actual problem: climate feedbacks due to the changes in water vapor distribution and cloud cover, albedo etc.. $\endgroup$ Apr 16, 2023 at 5:53

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.