Why Radiation from the Atmosphere
  Has Absolutely No Effect on Climate    
As the land and water surfaces of the Earth cool they lose thermal energy by conduction, diffusion, evaporation, radiation and chemical processes.  Most of this energy is absorbed by the atmosphere which then radiates it away, though some of the radiated energy gets straight through to space.  The atmosphere also absorbs more than a quarter of the incident radiation from the Sun (and carbon dioxide assists with this) so the atmosphere keeps the surface cooler in daylight hours.  Details here.
The proponents of the so-called Greenhouse Effect conjecture that some of the radiation that emanates from the atmosphere goes back and further warms the Earth.  However, radiated energy is not thermal energy as such:  it has to be absorbed and converted to thermal energy.  If it is reflected or transmitted through glass, for example, it is not absorbed.  But even if not reflected or transmitted, it can still be "rejected" by a surface which is warmer than the source of the radiation.  In fact it has to be, and always is.  We cannot have a case of a cooler atmosphere warming a surface which was already warmer and possibly getting hotter any sunny morning. This would violate the Second Law of Thermodynamics if radiation from a cooler atmosphere were converted to thermal energy in the surface.  Furthermore, such radiation cannot affect diffusion and evaporation.  These processes accelerate when the rate of radiative transfer reduces as explained my Radiated Energy and the Second Law of Thermodynamics.  Note also the conclusions made by Prof. Claes Johnson in his "Computational Blackbody Radiation."
Electromagnetic radiation has been postulated as having wave-particle duality, meaning it can have the nature of both particle and wave motion.  However, to consider it purely as a string of "photon" particles that, when it hits a "blackbody" will inevitably warm it is simply incorrect.  This is the fundamental fallacy of the arguments about how "back radiation" supposedly warms the Earth. 
In order to convert to internal thermal energy in any "blackbody" the incoming radiation must have an energy level above a "cut off" value which depends on the temperature of that body.  Put simply, most of the high energy associated with incoming solar radiation (direct from the Sun) is above the cut off and so does warm the surface.  But there is some low energy infra-red radiation in the solar spectrum (some of which gets absorbed by carbon dioxide) and some other even lower energy radiation from the atmosphere.  The radiation from the atmosphere is below the cut off and will be immediately re-emitted in the same frequency range.  As we all know, high energy UV radiation can warm and burn the skin, just as it warms the sand and rocks at the beach.  But we feel no warming effect from the assumed large quantity of low energy "back radiation" supposedly received from the atmosphere at night.
According to Wien's Displacement Law, the peak frequency, υmax = αkT/h  where h is Planck's constant, k is Boltzmann's constant and  α = 2.821439..)  Hence the peak (or "cut off") frequency is directly proportional to the absolute temperature, T.  
Now the frequency distributions for a few
different temperatures are shown at the
right.  Note that the frequencies for cooler
temperatures are a sub-set of those for
any higher temperature.  When radiation
from a cooler source strikes a warmer
target all its frequencies will resonate and
not be absorbed. 
However, radiation from a warmer source
has some higher frequencies than any
cooler target, and the energy in that
radiation with those frequencies is converted to thermal energy.  If this were not the way things work in nature, then the Second Law of Thermodynamics would not operate in all situations.  For example, consider what happens when the Sun is already warming the surface on a clear day at around 11am. The radiation and other heat loss from the surface to the atmosphere has already been taken into account, and yet there is still more radiation into the surface than out of it.  Now add radiation from the cooler atmosphere.  If that could transfer thermal energy to the warmer surface then such would clearly violate the Second Law.  Yet the models assume this energy transfer takes place 24 hours a day.
This false conjecture came about because absorptivity measurements done with much higher frequency visible light were wrongly assumed to apply to much lower frequency radiation from a cold atmosphere.  This is totally wrong, because absorptivity must reduce to zero (0) when the source becomes cooler than the target.
Johnson makes a highly significant "breakthrough" point here in that he finds that we can separate the two-way propagation of waves from the one-way propagation of "heat" energy by waves.  This is not possible with particle theory, and that is the very reason why everything has gone wrong in the models which assume that particles of heat energy (photons) hitting the surface will warm it regardless of the temperature of that surface.  The alarmists are basically claiming that you could use a mirror to reflect infra-red radiation back onto its source and thus further warm the source.  This would be creating energy and it simply does not happen.  Thermal energy is transferred when it is first converted to radiated energy in a warmer source and then converted back to thermal energy when it is absorbed by a cooler target.
I find that it helps to think of the night-time situation without incoming solar radiation.  We have the surface still radiating infra-red as it cools after the warmth of the day.  But would a mirror surface (polished aluminium) reflecting that radiation back again cause the surface to start getting warmer again?  Hardly!  Yet that, in essence, is what the alarmists are saying.  Indeed, Professor Nahle has been constructing experiments which show what does happen in reality.  Watch for his new one to be published in 2012, the results of which are progressing well I understand.
Professor Johnson sets out a detailed mathematical model which is consistent with (and in fact derives) the observed physical laws.  It centres around the concept of resonance (and near resonance) of the wave nature of radiation and essentially disposes of any need to imagine "particles" having no mass.
Johnson's work is as revolutionary as that of any previous scientist, Einstein, Planck or whoever.  "Planck's statistical mechanics is replaced by deterministic mechanics," he writes, explaining that this is "viewing physics as a form of analogue computation with finite precision with a certain dissipative diffusive effect, which we model by digital computational mechanics associated with a certain numerical dissipation."
His model and the computations explain why the spectrum of the incoming radiation from the Sun has hardly any overlap with that of the outgoing spectrum of infra-red radiation.  The high frequency coherent radiation above the cut off temperature converts, in a one-way process, to incoherent thermal ("heat") energy, whereas any radiation (mostly from the atmosphere) which has a lower frequency than the cut off frequency will be immediately re-emitted (or turned around by multiple deflections*) and will not be converted to thermal energy.
Thus “back radiation” can never warm the surface and the "greenhouse effect" is a physical impossibility.  Just as is the case with conduction, heat transfer is always from a warmer body to a cooler one even with radiation.  The mechanism which ensures this is twofold: (a) the immediate re-emission or scattering* of any radiation from a cooler source and (b) the conversion to thermal energy of any radiation from a warmer source.
Now, the word "never" might be a little strong because there can be rare situations when the surface is actually cooler than the air immediately above it.   But such situations would have no significant effect on overall climate.
It has been shown with spectroscopy that a gas does not absorb radiation from a cooler emitter, only that from a warmer one.  This can only be explained by Professor Claes Johnson's result.  But what about liquids as in the ocean?  Well water molecules certainly ought to absorb IR radiation.  But it turns out (here) that frost can remain on the ground all day (shaded from the Sun) without being melted by all that backradiation which is supposed to have about a quarter the power of the Sun.
Does the backradiation even exist?  Well there is certainly some IR radiation which can be detected by an IR camera.  But the original IR camera measures frequency (which can be converted to wavelength) as explained here.  Then the wavelength is converted to a temperature, or just given an artificial colour in order to form an image.  But it is incorrect to use the Stefan-Boltmann Law (assuming that some "packet" of the atmosphere is a blackbody) and from that equation to infer radiation intensity.  Instead (to see if backradiation has any effect) you would need to measure any warming effect on some object like a metal plate.  I am currently working on a series of such experiments, but I can tell you already that, even after a whole night, there is no more than 0.1 degree C difference between the shielded and the unshielded test samples of sand or soil which are in wide necked vacuum flasks.
This is what true science is all about - deriving a theoretical model and then proving that it conforms with empirical evidence.  This has never been done, and never could be done, with the back radiation hypothesis - because it does not conform with reality.  It is pseudo-science.
We are all aware of seasonal variations in each hemisphere.  Clearly these are caused by variations in the angle and duration of sunlight.  In general, on a worldwide basis, temperatures return to about the same level at the same time the following year, but there can be small natural variations, and indeed natural cycles are observed, both short term and long term.  These are discussed on my other site www.earth-climate.com and from these I deduce that temperatures will now start to decline until about 2028, as can be seen to some extent from the curved trend below.  A long-term (about 1,000 year) cyclic trend is still rising at the rate of about half a degree C per century, but even that trend should pass its maximum within 200 years. 
For more detail click here to read my paper.
Copyright Douglas Cotton   22 December 2011  -  Last updated 12 March 2012