- JoNova - https://www.joannenova.com.au -

Why greenhouse gas warming doesn’t break the second law of thermodynamics

This is generating many comments, see below for an update!

Behind the scenes some skeptics are suggesting that CO2 can’t warm us because the atmosphere is colder than the planet, and  it would break the 2nd Law of Thermodynamics (see Postma*, for example, p 6 – 7). I disagree. The 2nd Law of Thermodynamics applies to net flows of heat, not to each individual photon, and it does not prevent some heat flowing from a cooler body to a warm one.

Imagine three blocks of metal side by side. They are 11°C, 10°C, and 9°C. Think about what happens to the photons coming off the atoms in the middle of the medium temperature block between the other two. If heat never flows from cooler blocks to warmer blocks, all those photons have to go “right“, and not ever go “left”, because they “know” that way is towards a cooler block? (How would they?!)

The photons go both ways (actually every way, in 3D). There are more coming from the 11°C block to the 10°C block, sure, but the the 10°C block is sending ’em back to the 11°C block too. So heat is flowing from cold to hot. It happens all the time. Net heat is flowing always hot to cold. But some heat is going the other way, every day, everywhere, bar possibly a black hole.

People are being caught by semantics. Technically, strictly, greenhouse gases don’t “warm” the planet (as in, they don’t supply additional heat energy), but they slow the cooling, which for all pragmatic purposes leaves the planet warmer that it would have been without them. It’s a bit like saying a blanket doesn’t warm you in bed. Sure, it’s got no internal heat source, and it won’t add any heat energy that you didn’t already have, but you sure feel cold without one. —  Jo


Guest Post by Michael Hammer


I have lost count of the number of people claiming that global warming is impossible because the atmosphere is colder than the surface and thus cannot return heat to the surface since that would contravene the second law of thermodynamics. This is wrong and is based on an incorrect interpretation of the second law. The second law does not say a cold object cannot pass heat to a warmer object, it states that NET heat flow is always from warmer to colder.

As stated in the previous section, any object above absolute zero radiates energy. This energy is radiated in all directions. If such radiated energy strikes another object some or all of it is absorbed depending on the absorptivity of the object struck. The absorption does not depend on whether the object struck is warmer or colder than the object that emitted the energy, it only depends on the absorptivity of the struck object. However that object also emits energy some of which will radiate back to the first object and again be absorbed. Because the warmer object emits more energy there will be more traveling from warmer to cooler than vice versa and hence the NET heat flow will be from warmer to cooler.

Net heat flows from warmer to a cooler body, but some heat flows from a cooler body to a warm one.

Net heat flows from warmer to a cooler body, but some heat still flows from a cooler body to a warm one.


This apparent paradox is again based on a misunderstanding. Imagine you are standing outside on a cold winters night. It’s really cold and you are soon chilled to the bone so you step inside. Inside it’s a pleasant 20°C and almost immediately you feel warmer. But you are at 37°C and the room is 17°C cooler at 20°C how can it warm you, the second law of thermodynamics forbids it! No it doesn’t. When you were outside, you body was radiating energy to space but because the environment was so cold there was very little radiating back to you so the net loss was substantial. When you step inside your body is still radiating exactly the same amount of energy (remember the amount radiated depends only on the temperature and emissivity) however now the warmer walls of the room radiate more energy back to you than did the cold outside. Since the walls are colder than you are you still radiate more energy than you receive (heat flow is still from you to the room) but the difference between what you radiate and what you receive is less. You lose less net energy when inside than when outside so you feel warmer inside the room and it is easy to feel the room is warming you. In fact it is more accurate to say the room cools you less than did the outside.

Exactly the same situation exists with respect to Earth’s surface. Without the green house gases in the atmosphere the surface would be radiating directly to outer space which is extremely cold (-269°C). The green house gases prevent some of that radiation to space and thus keep the surface warmer than it would otherwise be. They do not do this by reducing the amount of energy the surface emits – doing that would entail changing the surface emissivity. Instead they radiate energy back onto the surface so that the net energy loss is reduced.


Michael Hammer is an electrical engineer who has spent over 30 years conducting research for a major international spectroscopy company.  In the course of this work he generated around 20 patents which have been registered in multiple countries.  Patents are rarer and more rigorous than peer reviewed papers, only available for economically valuable work, and costing thousands of dollars to process and maintain. Spectroscopy deals with the interaction between electromagnetic energy (light) and matter and it is this interaction which forms the basis of the so called “green house effect” in the atmosphere.


PS from Jo: Just so we’re clear here, I think CO2 molecules absorb Infra Red and have some warming effect, but I think feedbacks from clouds or humidity keep those effects so small that the total effect of adding more CO2 is minor, and not worth taking any action over. (There’s more info on feedbacks in these posts).

*The Postma example correctly shows that two ice cubes at 0°C will not heat one cube above zero, but it’s all about context. If the surrounding air is even colder, being next to an ice cube will keep you warmer longer. Even ice can be a “blanket”.
Being wrapped in ice  would slow heat loss if you happen to be on a rock in outer space.  Eskimoes stay warmer in an igloo. (Yes there are lots of reasons why, but the point remains… if you reduce your heat loss you stay warmer.)


Joseph Postma himself has replied — see comment #97 (it was caught in the spam filter and delayed)

From Michael Hammer #144
Let me try to put it another way. If you lie in bed without a blanket you lose a lot of energy and feel cold. If you now cover yourself with a blanket, the blanket reduces your energy loss. With a reduced energy loss you get less cold than you otherwise would have. Whether you consider “get less cold” to be semantically equivalent to saying “you will be warmer than you otherwise would have been” is up to you but I point out that general usage would say the blanket warms you. This is despite the fact that the blanket is colder than you are. No the blanket being colder does not transfer NET heat from itself to you, it merely reduces the energy loss allowing your internal heat generation to raise your temperature more. If you put a blanket over a piece of cold steel it does not make the steel warmer.

An exactly analogous situation exists with respect to Earth. There is an external energy input notably the sun. An opaque atmosphere reduces the energy loss from the surface to space which allows the energy input from the sun to raise the temperature slightly. The effect of more CO2 is to very slightly increase the range of wavelengths around 15 microns at which the atmosphere is opaque.

The mechanism by which this energy loss is reduced cannot be by reducing the heat radiated by the surface because the atmosphere cannot influence the emissivity of the surface. Rather it acts by returning some of the energy radiated back to the surface. This is the back radiation.

If you want the analogy with a blanket to be more accurate consider the survival blankets which are simply a silvered sheet of thin plastic. Clearly the thin plastic has negligible impact on conduction. It could act by reducing convection but then again it does not need to be silvered to do that. A transparent sheet of plastic would do that just as well yet a transparent sheet of plastic does not work anywhere near as well as a silvered sheet. The silvered sheet works so well because the shiny surface has very low absorptivity and emissivity so it loses very little energy by radiation. There is still a difference in that the silvered surface reflects the energy back onto your body rather than via an absorption and then emission process but the overall impact is very similar.



From Jo
Here’s why talk about whether its convection or conduction vs radiative cooling is irrelevant
The blanket analogy is perfect because we are discussing whether it’s possible for a cooler item to induce (somehow) an increase in temperature of a warmer item. NOTE: The cooler item has no internal heat source, but the warmer item (Earth or body) does have energy added in. The method of heat transfer is irrelevant. (Talk of two ice cubes misses the point unless one cube is heated by the sun, or burns fat. )

The point is that YES, obviously in the real world, blankets keep us warm. Pink batts “lift the temperature of your home in cold weather”. They don’t do it by supplying energy, they do it by blocking energy loss. The cooler item is not supplying a single new joule of energy, but there another mechanism of increasing an objects temperature. It’s called insulation. It’s a reality we all know and use every single day. Why deny it?

Can commenters move on from repeating the truism that a colder object can’t make a warmer one even warmer without supplying energy? We all know that, but it applies to a closed theoretical system with no extra source of energy. In the systems with blankets/people and the sun/earth there IS an extra source of energy at least until you’re dead or the sun burns out.


Comments have overflowed. The first 500 are on the next page.

Comments that refer to other comments will still have the original numbers (up to comment #300)

ie “comment #767” now means comment #267

7.6 out of 10 based on 21 ratings