A little while ago, I proved conclusively :-) that the recent temperature increase isn't from waste heat. In essence, its obvious: the energy from waste heat is about 1/100th of the radiative forcing from GHG's.
When I wrote that, I had forgotten about Nordell 2003 (subscription required, but you can see the abstracts, and the people commenting on it), which has recently resurfaced amongst the comments at RC.
Thomas Palm succinctly summarises it thus:
I've read the paper by Bo Nordell and it is junk. Here is as summary I wrote in sci.environment after having read it:
The difference between the surface of the Earth and space is 33 degrees.
Geothermal heat produces about 0.07 W/m^2 heat. And if this causes 33
degrees heating the climate sensitivity must be 470 K/W/m^2. Thermal
pollution is 0.02 W/m^2 and will thus give a heating of 0.02*470= 9 K.
The mathematics he uses reduces this by a factor of 3, but essentially
this is his argument.
If Nordell *were* right, we'd be seeing huge fluctuations in global temperature as a result of the small variations in solar forcing.
Only just recently, two letters have been sent to the journal Nordell published in (Global and Planetary Change) saying pretty much the same, but more politely, as befits a scientific journal: the first by Curt Coveya, Ken Caldeiraa, Martin Hoffertb, Michael MacCrackena, Stephen H. Schneiderc, Tom Wigley; and the second by J. Gumbel and H. Rodhe (the latter are pretty fierce by academic standards, saying This paper is seriously flawed and contradicts basic principles of physics. The paper contributes wrong and misleading arguments to the important discussion about human impact on the Earth’s climate).
Both of these point out the obvious, as I did: that the thermal pollution term is tiny by comparison with the GHG term. They also point out that Nordells model is far too simplistic to be any use: As these basic assumptions by the author are in error, all subsequent conclusions are also in error. We would also like to comment on the atmospheric model used by the author to obtain his radiative results. This model is based on simplistic assumptions that make it inappropriate for any quantitative calculations (G+R again).
The original Nordell paper, in its intro, makes a couple of characterisic septic-type mistakes: repeating the debunked wv-is-98% claim (OK, so *we* hadn't debunked it when Nordell published, but it was just as dodgy then); quoting Lindzen *1992* on models (how they got away with ref'ing that in an alleged scientific journal I don't know).
Nordell then responds to the comments, but to my mind unconvincingly.