2.3 The Expansion of the Universe
In the relativistic Friedmann-Lemaître cosmological model the wavelength of a freely propagating photon is stretched in proportion to the expansion factor from the epoch of emission to detection:
The first expression defines the redshift z in terms of the
ratio of observed wavelength to wavelength at emission. The
cosmological expansion parameter a (t) is proportional to the
mean distance between conserved particles.
The most direct evidence that the redshift is a
result of expansion is the thermal spectrum of the
CBR [26].
In a tired light model in a static
universe the photons suffer a redshift that is proportional to
the distance travelled, but in the absence of absorption or
emission the photon number density remains constant. In this case a
significant redshift makes an initially thermal spectrum distinctly
not thermal and inconsistent with the measured CBR spectrum.
One could avoid this by assuming the mean free path for
absorption and emission of CBR photons is much shorter than the
Hubble length, so relaxation to thermal equilibrium is much
faster than the rate of distortion of the spectrum by the
redshift. But this opaque universe is quite inconsistent
with the observation of radio galaxies at redshifts z ~ 3 at
CBR wavelengths. That is, the universe cannot have an optical
depth large enough to preserve a thermal CBR spectrum in a tired
light model. In the standard world model the expansion has two
effects: it redshifts the photons, as a (t), and
it dilutes the photon number density, as n a (t)-3.
The result is to cool the CBR while keeping its spectrum thermal.
Thus the expanding universe allows a self-consistent picture: the
CBR was thermalized in the past, at a time when when the universe
was denser, hotter, and optically thick.
I have not encountered any serious objection to this argument;
the issue is the expansion factor. In the relativistic
Friedmann-Lemaître model the expansion of the universe traces
back at least as far as redshift z ~ 1010, when the light
elements formed in observationally reasonable amounts
[18].
In the model of Arp et al.
[27] the expansion
and cooling traces back to a redshift only moderately
greater than the largest observed values, z ~ 5, when there
would have been a burst of creation of matter and radiation
followed by rapid clearing of the dust that thermalized the
radiation. The Arp et al. picture
for the origin of the light elements has not been
widely debated. If it were agreed that it is viable then a
choice between this and the Friedmann-Lemaître model would
depend on other tests, such as the angular fluctuations in the
CBR, as discussed next.