Chronos said:
I dissolve 1 gram of salt into 1 liter of distilled water. If I add more water to the solution the question becomes - "is salt lost or merely diluted"?
My point was that if
rhoradiation ~ R-3 then the energy of the CMB would be merely diluted, however in fact
rhoradiation ~ R-4 so it is
more than diluted. If you integrate the radiation density over the volume of the universe the total
decreases with time. GR is quite happy with this as it doesn't set out to conserve energy in the first place, only energy-momentum, i.e. the 'rest' masses of individual atoms.
The measurement of gravitational/cosmological red shift is a measurement of the energy of the photon compared with the mass-energy of the atom it interacts with. A red shift is observed when the cosmological photon is compared with an equivalent laboratory photon. Each has been emitted/absorbed by the same, identified, atom, say sodium, and so either the energy of the photon has decreased or the mass-energy of the atom has increased.
GR understands the observation by the former interpretation, and so the photon has mysteriously lost energy, over and above the 'dilution' factor.
GR is a type of 'tired light' theory! It interprets the red shift as a doppler effect, and that is self consistent, however it is not the only possible interpretation.
That then raises the question of "Where does the energy of a doppler shifted photon go?" The answer lies in the definition of energy levels by different observers in mutual motion.
Note that without relativistic effects being taken into account that in the classical doppler shift there is an interesting conundrum. If we treat light as a stream of particles, photons, energy seems to be lost in classical doppler shift as photon number is conserved while the energy of individual photons has decreased. Treat light as a stream of radiation, however, and the energy is not lost, the red shift being compensated by the extra time taken for the stream to be received!
Garth