Friday, December 31, 2010

On greenhouses and the Greenhouse Effect

From time to time, some people get excited about the fact that the greenhouse effect seems wrongly named. An extreme case is the paper of Gerlich and Tscheuschner, where they devote a whole section 2 (19 pages) to the question. It continues to arise; Judith Curry touched on it in a recent thread, where discussion turned to the old Wood's experiment

The standard response correctly says that none of this matters. The atmospheric  GHE is what it is, and is no less true if it turns out that the allusion to greenhouses is inaccurate. Wikipedia takes this approach, with a refreshing brevity and directness.

However, there is a little more to the story (which still says nothing about the truth of the GHE). Greenhouses do work mainly by blocking heat loss through convection. But IR flux blocking is not totally insignificant.

Wood's experiment

The description of this in the 1909 Philosophical Magazine is indeed brief. Some people give the impression that he built a greenhouse out of rock salt, but the geometry seems to be simpler and on a much smaller scale. Still, he concluded:
It seems to me very doubtful if the atmosphere is warmed to any great extent by absorbing the radiation from the ground, even under the most favourable conditions.

This final statement is little emphasised, but seems to be amply justified by the sketchiness of the note:
I do not pretend to have gone very deeply into the matter, and publish this note merely to draw attention to the fact that trapped radiation appears to play but a very small part in the actual cases with which we are familiar.

Indeed he didn't.

Modern Greenhouses

Ironically, the situation that forced Wood to use rock salt as an IR-permeable material is now reversed. In his day, greenhouses were always made of IR-blocking glass. Now they are often made of IR-permeable plastic. And, in support of Wood, they do work. They block convective heat transport, and also provide an insulating effect to limit conduction through the boundary.

However, a moderate amount of heat is still transferred by IR. Ironically, a reason why it isn't larger on balance is because of the countervailing downwelling IR, also a subject of occasional skepticism. Downwelling IR, from the atmosphere and especially clouds, is not that much less than upwelling. Glass blocks both upwelling and downwelling; the nett flux blocking is small, but if you are focussing on the fate of heat emanating from the surface, the IR fraction that is blocked us a larger fraction.

A real greenhouse issue

A covering that blocks IR actually makes a real improvement to greenhouse efficiency. It is more expensive, and people are willing to pay for it. Here are some industry observations:
NSW Ag Dept:
For example, films may be used to exclude ultra violet (UV) light for chemical free pest control or reflect long wave infra red (IR) radiation to improve heat retention at night. ... Long wave radiation (2500-40000 nm) absorbers reduce the loss of heat radiated from materials and objects (including plants) inside the greenhouse.

Or this advice
You can also buy a plastic film with an infrared inhibitor; it cuts heat loss inside the greenhouse by up to 20% on a cloudless night.

Actually, it would also cut loss on a clear day.

And here is a pamphlet which describes the practicality in some detail (for heated greenhouses, but the principle is the same).
The IR / anti-condensation treated films cost about $0.015 per square foot more than untreated films but reduce energy use by 15 to 20%.

Scale Issues

The main thing wrong with the Wood's style reasoning is scale. IR blocking is a minor effect on the scale of a greenhouse because convection is relatively much more effective. The temperature gradients are huge compared to what is maintained in the atmosphere. Wood compounded this by experimenting on a much smaller scale again.

In the atmosphere, IR transport is more important than convection, so the blocking effect of GHG's matters much more.



  1. I think it is pointless being pedantic about names and language.

    Most people named 'Smith' do not practice the profession of a blacksmith.
    Most people named Stokes probably don't live in or near a place called Stoke!
    Someone named 'Fletcher' probably doesn't make arrows.

    So why the fuss about naming something the 'Greenhouse Effect' many decades ago, then finding out later a different name would be appropriate?

    It's like metaphors, they rarely stand the test of time. But people don't change the name because of change.

    A 'window' on a computer is still called a window, even if the metaphor is tired and out of date.

    Even language in every day use is totally inappropriate. People don't put a mobile phone on the 'hook', but the term is still widely used throughout the telecoms industry for the signal that indicates a phone a call has been terminated.

    My advice is challenge the skeptics on their use of language. Basically they haven't a clue about language, if they come up with silly challenges like this.
    Maybe Gerlich and Tscheuschner should get a life if they wasted so much effort analysing the name of a process.

  2. You might also like, which has been around for a while.

    I'll copy in my section on his last para:

    Why is his second to last paragraph wrong?

    Firstly, note that unlike the experiments described earlier, this paragraph merely expresses his opinion.

    Second, although the troposphere is subject to convection, the stratosphere is not.

    Third, in contradiction to his assertion about "the very low radiating power of a gas", the troposphere is largely opaque to infra-red radiation, which is why convection is so important in moving heat up from the surface. Only in the higher (colder) atmosphere where there is less water vapour is the atmosphere simultaneously somewhat, but not totally, transparent to infra-red and thus permits radiation to play a part.

  3. This is a bit more modern experiment.

    Interestingly, the nighttime temperature in a greenhouse is typically less than the ambient air temperature outside of the greenhouse. Let's see if you can give an explanation as to why. ;-)

    IR blocking does help some, but the benefits aren't huge, as the data show.

  4. Pardon me if this is a double post... (My previous version appears to have been eaten.)

    This is a bit more modern experiment on greenhouses.

    Interestingly enough, the nocturnal temperature inside of the greenhouse is lower than the ambient temperature outside of it. I'd be interested in your explanation as to why. ;-)

    These data also show that IR blocking covering helps a bit, but not by much.

  5. Carrick,
    One thing that isn't clear is how effective the IR blocking actually was. I think the cover on these long tunnels is quite thin.

    The lower temp at night is presumably because IR loss from the ground means that the nett flow of heat is downward - the air is warmer than the ground. And the cover blocks that flow. The IR opacity should counter that, which presses the question - how opaque is it?

    Yes, the IR benefit is small, although Reid seems to think it is worthwhile.

    Sorry about the eating - yes, it was the spam filter. I've no idea why - the post that got through seems little different.

  6. Belette,
    Yes, Connolley is right. Actually what Wood says is contra Kirchhoff's Law - to the extent that the air can't lose the heat by radiation, it can't absorb it.

    Incidentally, here is praise for Wood from an unexpected source.

  7. I think you'll find that Belette would tend to agree with Connolley..

  8. Carrick,

    The covering, by blocking air movement, forces a temperature inversion inside the covering. Heat is lost from the ground by radiation faster than it can be replaced by convection and conduction from the air. Temperature inversions at night are common when the humidity is relatively low and there is no wind. It's the basic mechanism for the formation of dew and frost. The temperature profile time series follows the classic diffusion model with an initially steep gradient that over time becomes shallower and extends upward further.

  9. Eli has found something interesting along these lines basically the first systematic measurements of backradiation.

    FWIW the bunny did his bachelors at Hopkins when there were still a lot of folk about who remembered and revered Wood. It caught. The guy was elegant in much the same way Feynman was.

  10. Eli,
    That article was a great find. I hadn't realised there were so many Angstrom's. But as your second post noted, Anders K is the grandson of Anders J, who the (non-SI) unit commemorates. Knut, his father, made the instrument that he used. Knut was also the Angstrom who disputed Arrhenius' version of the greenhouse effect, with the saturation argument which held sway for many years.

    It's pretty amazing to read a 1980 interview with him too.

  11. Noel Barton emailed me to say that he had submitted a comment (twice) which did not appear. It didn't go to the spam filter either. It's a puzzle, and please let me know if anyone else has this experience. Here is his comment:

    Hopefully adding to the discussion and not completely off-topic …

    I’m interested in solar power generation and have been working on passive solar heat collection under a transparent insulated canopy. At the temperatures I’m interested in (80-140 deg C), infrared losses from the ground had to be included. My simulation model has molecular heat diffusion through the glass canopy, which includes an underside convection-suppression region (transparent slats or bubbles). Under the canopy I also include an IR reflective layer. Other features of the simulation model are variable insolation levels, absorption (both visible and IR) by the glass, convective heat transfer off the outside surface, and heat transport by airflow under the canopy to my heat engine.

    Details at N.G. Barton, “Annual output of a new solar heat engine”, Proc Australian Solar Energy Conf, Canberra 2010. Also

    If of interest, I’d be happy to run simulations of the behaviour of a greenhouse.

  12. A more recent update of the experiment mentioned by Carrick (above) to find the value of adding IR blockers to polyethylene is given below.

    In the conclusions (pg 2) the researchers found the greenhouse effect was tiny and inconsistent only occasionally evident.
    They concluded that the extra expense of adding IR lockers was not justified.

  13. Bryan,
    Well, I think the first point of interest is that the paper found a number of commercial films with IR blockers added, and noted that they were more expensive. So presumably someone thinks they do something.

    The striking thing to me about the actual temp measurements was that while the IR differential effect at night was small, so was the actual effectiveness of the tunnel (greenhouse) in Graphs 5 and 6. In fact, on the night of 3 Nov, featured in graph 7, all tunnels were colder than ambient. It may be that temps were measured closer to the ground inside than out.

    One thing to say about high tunnels - if the ground cools at night by radiating IR which is then blocked at the film, it may take a long time for the heat to be transferred back to the measurement point.

  14. Nick
    I think its quite likely that a number of purchasers believed in the "greenhouse theory" .
    Hence presumed that IR blockers would give an increased heating effect.
    Commercial suppliers as usual will cater for the market, even if the demand is irrational.
    I wonder how long it will take the agricultural sector to catch on?
    On your high tunnel point the theory says the blockers will radiate back to the Earth surface.
    It seems that the amount radiated back is so small as to make no practical difference.
    I think that the reason the polytunnel drops below outside ambient temperature on some cold winter nights is because the polytunnel is now cut off from circulating convection streams in the air outside.

  15. Nick: "In fact, on the night of 3 Nov, featured in graph 7, all tunnels were colder than ambient. It may be that temps were measured closer to the ground inside than out"

    I don't think that's the explanation:

    It's generally colder inside of a greenhouse at night (at least with typical meteorological conditions: hint) than it is outside, when measured at the same height above the ground. Here's an explanation I left on Jeff ID's blog:

    It is true that the interior of the greenhouse can warm more in the daytime, but paradoxically it can cool more rapidly at nighttime.

    The reason is that typically at nighttime a temperature inversion is set up, and in the “unprotected environment” as you get advection of air (“wind”) across the ground you get air exchange between the surface and higher altitudes (the wind moves faster higher up than it does near the surface due to surface friction)—paradoxically when you get a wind gust at night, you see a jump in ground temperature as a result of this.

    The greenhouse, because it is blocking advective air motion, cools to the temperature that would be present if there were no mechanical exchange of air between higher and lower altitudes…. only on a perfectly windless night would you expect the two cases (ground exterior and ground interior to the greenhouse) to converge. [This is a prediction, I've not seen data for that case.]

    The effect of the interior of a greenhouse getting cooler at night is shown in this document.

    It also shows that if you add IR blocking material, the temperature is higher than it would have been without the IR blocking material.

    Addendum: As you and Bryan point out, the effect is not substantive. IR blocking film looks to be another form of snake oil to me.

  16. Carrick.
    Yes, that explanation sounds right.

    I don't think the IR block gain is worthless. A few degrees is important when you're warding off frost.

    But high tunnels may not be the best test - there are other important greenhouse types. I note that they are passively vented - I don't know whether the venting can be blocked at night. That may be why they do a relatively poor job of preventing heat loss at night, with or without IR block.

  17. Nick Stokes "I don't think the IR block gain is worthless. A few degrees is important when you're warding off frost."

    Row cover is much more effective in protecting against frost that a shift of a few degrees in surface temperature.

    My wife owned a nursery, can you tell?

  18. I still have a small greenhouse (that is unvented. It might be interesting to instrument it with a few thermometers (one inside, one outside).

    I predict with a totally closed greenhouse, the nocturnal temperature will remain lower than the exterior temperature.

  19. Carrick,

    A glass covered greenhouse? Glass is opaque in the thermal IR. While the outer surface of the glass could become cooler than ambient on a clear night, are you really going to lose enough heat by conduction through the glass to make the inside cooler than the outside? If it is, I would suspect heat loss through the ground or some path other than the glass.

  20. Actually, it's plastic sheeting, and it's pretty thin so I suspect very little IR screening is occurring.

    Here's a photo of it...

    Remember that the nocturnal boundary layer has typically has a temperature inversion and a wind-sheer generated surface boundary layer (resulting from surface friction as the air advects over it). The apparent coiling of the greenhouse, IMO, is less due to the greenhouse than the lack of mixing of warmer upper air and the surface air due to the "protected" environment of the greenhouse.

    Recall that the ground mostly cools at night due to radiative heat loss, heat transfer into the contact layer of the atmosphere is generally unimportant, unless you have evaporative heat loss. I would guess conduction across the top of the greenhouse to the air immediately above it would be equally unimportant. I do know when you have a turbulent surface layer, there are two radiative balance equations that you look at, one for the ground, and one for the top of the turbulent boundary layer (shades of tropospheric physics.)

  21. If it's plastic, then I agree that you will get about the same temperature inversion inside the greenhouse that you would get outside on a calm night.