A New Mechanism to Explain Observations Incompatible with the Big Bang
| Scientific Paper | |
|---|---|
| Title | A New Mechanism to Explain Observations Incompatible with the Big Bang |
| Read in full | Link to paper |
| Author(s) | Paul Marmet |
| Keywords | Big Bang, observations, redshift, Doppler phenomenon |
| Published | 1991 |
| Journal | Apeiron |
| Volume | 1 |
| Number | 9-10 |
| No. of pages | 21 |
| Pages | 119-139 |
Read the full paper here
Abstract
The Big Bang model describing the origin of the Universe has been accepted mainly on account of the lack of alternatives to explain certain observations. This model, in which the redshift of remote galaxies is interpreted as a Doppler phenomenon, runs into impossible difficulties with the cosmological background at 3 K because this background is too homogeneous. Many observations, such as the redshift on the solar limb, the redshift of hot binary stars, the K-effect and a plethora of other observations, are not compatible with current theories. An alternate mechanism is described which yields a redshift without Doppler effect. This mechanism is already confirmed by several observations, and leads to an unlimited Universe model. Results are compared with proposals made by Halton Arp.
Overview
Written for Apeiron in 1991 by Paul Marmet, then in the Department of Physics at the University of Ottawa, this paper does two things at once: it assembles a case that the Big Bang model fails against several classes of observation, and it offers a specific physical mechanism — slightly inelastic, non-dispersive forward scattering of photons by atoms and molecules — to produce the cosmological redshift without any recession of galaxies. The mechanism is Marmet's own, published in more detail in 1988, and is a tired light proposal with an unusually concrete quantum-electrodynamic basis: he insists that "no new basic physics is required," only the well-known fact that photon–atom collisions are never perfectly elastic.
The cosmological consequence is an unlimited Universe, infinite in space and time, satisfying the Perfect Cosmological Principle. Marmet argues this model accounts more naturally for the extreme smoothness of the 3 K background than expansion does, and he closes by comparing his account with Halton Arp's intrinsic-redshift proposal, identifying one observation — the redshift at the solar limb — where the two diverge and where he takes his own to succeed.
The argument
Why the Big Bang fails
Marmet opens with a point of historiography: Misner, Thorne and Wheeler's Gravitation states that "Hubble discovered the expansion of the Universe," whereas Edwin Hubble's own 1937 book and papers show that "he spent most of his life trying to disprove the hypothesis that the cosmological redshift was due to the Doppler effect. Hubble neither discovered nor even believed in the expansion of the Universe: he discovered the cosmological redshift."
He then presses an internal inconsistency. The primeval atom — all the mass of the Universe in near-zero volume — "represents the most extreme example of a black hole that we can think of. Since it is known that nothing can be emitted from black holes, how can the primeval atom expand?" The usual repair, a gravitational constant that grows in after creation, he calls "too subjective to be acceptable" and no part of general relativity. A second objection is philosophical: because time itself is said to begin at creation, no cause of the Big Bang can be sought, "even quantum fluctuations could not produce the Universe since, at that instant, time did not exist," so the model "leads to the rejection of the principles of causality and rationality" and differs from other creationist accounts only in the number of years since creation.
He then takes the four standard arguments in turn. On redshift, he cites Reboul's Untrivial Redshifts catalogue of 780 references and Arp's and Narlikar's books, noting that quasars often show different redshifts in absorption and emission and sometimes more than one redshift for the same object, and that the Sun's chromosphere shows different redshifts in different lines from a single location. He also finds it suspicious that quasars all appear beyond a certain distance from us, which on the Doppler reading places us at the centre of the Universe — "the same belief people had in Galileo's time." On light-element abundances he cites Lerner's argument that helium-4 and other light isotopes form in massive stars, concluding with Lerner that "either the blackbody spectrum or the light element predictions of the Big Bang is clearly wrong." On the 3 K cosmic microwave background, he makes his sharpest claim: because radiation is coupled to matter, and matter is "extremely lumpy in the form of galaxies and clusters," the Big Bang should not produce a homogeneous background, whereas the COBE and Lange results show none of the expected inhomogeneity. In an unlimited Universe, by contrast, all interstellar and intergalactic matter far from stellar light must sit in equilibrium and emit a Planck spectrum at its own temperature — smooth by construction. He adds the observational difficulties: "mature galaxies" at z = 3.4 reported by Lilly, whose appearance "so soon after the Big Bang poses a serious threat," and the Great Wall and Great Attractor, structures too large to have assembled in the available time.
The redshift mechanism
The core of the paper is an argument that photons interact with atoms far more often than is usually appreciated, and that these interactions are slightly inelastic. Marmet's demonstration is deliberately elementary. From the refractive index of air (n = 1.0003), light traversing 100 metres is delayed by 3 cm relative to vacuum — a delay "about one billion times the size of the atom," which requires of order a billion photon–molecule collisions. Yet an object 100 metres away is not fuzzy, even through a telescope. Therefore "photon–molecule collisions without angular dispersion are an everyday experience that is completely ignored." In extragalactic space, where the gas density is more than twenty orders of magnitude lower, he estimates about one such interaction per week, with Rayleigh scattering (which does diffuse light in all directions) far rarer still.
Each such interaction involves a finite absorption–re-emission delay. The passing electromagnetic wave polarises the atom, displacing its electron from its normal quantum distribution; the transferred momentum accelerates the electron; and by Maxwell's Equations an accelerated electron radiates. So at least one very soft secondary photon is emitted by bremsstrahlung, and the forward-scattered photon comes out with slightly less energy. Marmet gives the loss per collision as about 10−13 of the incoming photon's energy — a fractional loss, so that Δλ/λ is constant and the effect is, as he puts it, "almost indistinguishable from the Doppler redshift." The secondary photon carries a wavelength some 1013 times longer, thousands of kilometres, far beyond the longest wavelength (144 m) reached by radio astronomy, so it is undetectable. He grounds the inelasticity in standard references — Jauch and Rohrlich, Bethe and Salpeter — rather than in new physics.
Tests and consequences
Two confirmations are offered. The first is the solar limb effect: spectra taken at the limb are redshifted relative to disk centre beyond what the Sun's rotation accounts for, an anomaly reported since Halm in 1907. Because a limb line traverses more chromospheric material, the collision mechanism predicts a greater shift, and Marmet reports "accurate agreement with the observed curves," better than the Schatzman–Magnan and Finlay–Freundlich alternatives plotted alongside. He adds that the model explains the absence of redshift in lines formed in very high chromospheric layers and the stronger shift of the iron line at 5250 Å, which forms deeper. The second is hot binary stars: the hotter component should show the larger redshift, because a hotter blackbody has a shorter coherence length (by Fourier transform), and this is what is observed — illustrated with HD193576, whose components' centre-of-mass "velocities" differ by 90 km/s.
Quantitatively, the mechanism reproduces the Hubble Constant if the mean cosmic density is about 0.01 atom/cm3. Marmet devotes several pages to arguing this matter exists but is invisible: cold molecular hydrogen has no permanent electric dipole, its first rotational state is effectively forbidden, the second-state transition takes some 30 billion seconds, and only near UV stars is H2 detectable at all. He points to the flat rotation curves measured by Rubin, which imply 90–99% of galactic mass is unseen, to Burbidge's finding that clusters have far more kinetic than gravitational energy, and to the Great Attractor's ~1016 solar masses, concluding that "most of the galactic and intergalactic gas responsible for the non-Doppler redshift is molecular hydrogen." Finally he compares with Halton Arp, whose intrinsic redshifts depend on the epoch at which matter was created. Both accounts fit the data that young, high-redshift galaxies are also hot, since Marmet's hotter stars are redshifted for the coherence-length reason. But the solar limb discriminates: "it is highly unlikely that matter on the sun's limb ... is younger than matter everywhere else." He is also careful to say the Magellanic Cloud data are too imprecise (given a ~1 km/s effect and the intrinsic velocity dispersion) either to confirm or to invalidate his mechanism.
Assessment
The paper's strongest feature is that its mechanism is stated concretely enough to be argued with. Unlike most tired-light proposals, Marmet does not simply postulate an energy loss proportional to path length; he names the interaction, derives its fractional character (which is what makes Δλ/λ constant and reproduces a Hubble-like law), puts a number on the loss per collision, and predicts where the lost energy goes. The refractive-index argument is genuinely instructive: a great many photon–molecule interactions do occur in clear air without blurring an image, and that fact deserves the attention he gives it. His historical correction about Hubble is accurate and worth repeating. And the treatment of Arp is a model of how a dissenting author should handle a rival dissenting theory — identifying the one observation on which they part company rather than claiming universal superiority. His refusal to overclaim from the Magellanic Cloud data is similarly creditable.
The difficulties are severe, and the most damaging is one Marmet raises himself and then does not resolve. If a photon in extragalactic space interacts about once a week and each interaction involves a finite absorption–re-emission delay with a momentum transfer to a bound electron, then over billions of years the accumulated transverse momentum kicks must blur distant images. The 100-metre air argument shows blurring is small over a billion collisions at optical densities; it does not show it is zero, and the extrapolation to 109-light-year paths amplifies whatever residual there is by an enormous factor. Distant galaxies and quasars are observed sharp to the diffraction limit of the Hubble Space Telescope, and gravitationally lensed quasar images remain point-like. Marmet asserts non-dispersion but never bounds the scattering angle. The claim that the energy goes into secondary photons "a few thousand kilometres long" also removes the mechanism from test by construction, and such a soft-photon bath, integrated over cosmic path lengths, would carry an enormous total energy that is not accounted for anywhere in the paper.
Three further conflicts with measurement should be named. First, the redshift here arises from intervening material, so it must depend on the amount of intervening material; but the Hubble relation is observed to be remarkably uniform across lines of sight of very different gas content, and quasars seen through galaxy haloes do not show extra reddening-independent redshift. Second, and decisively, a scattering redshift shifts wavelengths but does not stretch time: the light curves of Type Ia supernovae are observed to be broadened by exactly (1+z), as expansion requires and as no tired-light process reproduces. That test was not available in 1991, so this is not a fault of the paper, but it is now the sharpest evidence against its central claim. Third, Marmet's headline argument — that the 3 K background is "too homogeneous" for the Big Bang — was overtaken almost immediately: COBE's DMR announced ΔT/T anisotropies of order 10−5 in 1992, one year after publication, and WMAP and Planck have since resolved the acoustic peak structure in detail. The homogeneity he takes as fatal turned out to be smoothness with precisely the small fluctuations the model needed. Conversely, his insistence that vast quantities of cold molecular hydrogen must lurk unseen was a reasonable inference from the rotation curves in 1991; subsequent baryon accounting from primordial deuterium and from the CMB acoustic peaks constrains the total baryon density well below what his 0.01 atom/cm3 requires, which is where a modern reply to the paper would have to begin.