I am on the case with my own theory. Two other theories are:
- The late Paul
Marmet's
theory involving photons and
low-density H and H2 gas: http://www.newtonphysics.on.ca/HUBBLE/Hubble.html
On 25 November 2005 Hilton Ratcliffe wrote to the altcosmology list ( http://groups.yahoo.com/group/altcosmology/ ) that Prof. Paul Marmet died in May as a result of cancer.
This was first developed, or first published at least, in the late 1980s. I don't understand how this works - why the photons only lose energy to hydrogen atoms or molecules. By email, Paul Marmet told me that his theory is not about coronal heating - he told me he believes the conventional theory about it being heated by magnetic waves. Paul's theory inspired me to think some more, and come up with another theory about free electrons, which I now think is incorrect. But further thinking and Googling lead me to another theory. Googling "plasma redshift" in late 2002 lead only to the following theory . . .
- Ari Brynjolfsson's theory: The Plasma Redshift of Photons. Like Paul Marmet's theory, this involves low densities of particles where the particle spacing is comparable to or greater than the coherence length of the light. However this theory is for plasmas, so it might be able to explain coronal heating. An abstract from 2000 is here: http://www.eps.org/aps/meet/APR00/baps/abs/G2750005.html . As noted above, the final paper is at http://arxiv.org/abs/astro-ph/0401420 .
When a photon with a short coherence length (as is the black-body radiation from the Sun) passes through a low density plasma (or arguably a neutral, partially ionised/ionized gas, or even molecules to very small grains of dust) it is not passing through an homogenous medium. It travels at full light-speed in the vacuum between the particles, and then is slowed down when it encounters one or more particles. If we imagine the wavefront being a parallel front travelling through empty space, with a single particle (say an electron, a proton, an ion, a nucleus, a molecule etc.) in its path, the wavefront is very slightly slowed down in the vicinity of that particle. This puts a slight *dimple* in the otherwise planar (actually its the surface of a large sphere) wavefront.
A thought experiment indicates that this, or something like it, must occur. A plasma has a refractive index above 1.0 - it slows down light which passes through it. No matter how low density the plasma, as long as there are some particles, the refractive index will still be above 1.0. Therefore, even when the density is so low that the inter-particle distances are far greater than both the wavelength and coherence length, the wavefront must still be slowed slightly - and this must occur in the vicinity of the individual particles, rather than the large spaces of pure vacuum between them.
Here is a rough attempt at an illustration, showing the wavefront travelling from the left to right, with red and blue for positive and negative electrical polarity (in the vertical direction, or at right angles to the plane of the picture) of the wave:
Update 2006-09-11: I have calculated an impulse which has the same spectrum as blackbody light, and it is much shorter than the multiple up and down waves I have illustrated above. Please see: ../simmering/#impulse.
As the wavefront is slowed by the particle (indicated by the dot, but with concentric circles to show what we might think of as contours of light-slowing influence for light of about this wavelength), it temporarily couples some of its momentum to the particle. This can be shown via the following arguments / thought experiments:
A beam of light being absorbed, reflected or deflected transfers momentum to the object which absorbs it or changes its direction. We know that such energy is deposited in quanta, each one resulting from what we later can describe as a "photon". It is reasonable to assume that each such deposition carries with it a small quantity of inertia. (A square metre of sunlight at the Earth - 1356 watts - conveys about 0.44 milligrams of momentum. This is the mass of a drop of water 0.65mm in diameter.)A wavefront of light passing into a glass block couples some momentum to its front surface, according to how much its speed is slowed. On re-emerging at the other side the same force is made against the rear surface as the photon regains is normal speed. We would expect a temporary coupling of inertia from the wavefront to any object which slows it, including a thin piece of glass, which is thinner than the wavefront, or even a soapy water film. Since we know that overall, any plasma slows the light (and therefore temporarily receives some of its inertia), we expect that the edge of a dense plasma will receive some fraction of the inertia, just as the front edge of a block of glass does. When the plasma is so low in density that the interparticle spacing is greater than the wavefront thickness (AKA "coherence length"), each particle is like a separate, independent, obstacle to the wavefront, which would otherwise be travelling at full vacuum light-speed. So we expect the wavefront to couple some momentum to the particle as it approaches it, and to kick back against the particle as it proceeds onwards, just as it would a small block of glass.
The wavefront contains the momentum - we can absorb electromagnetic radiation at all other places but wherever the wavefront is travelling, and the wavefront still exists and will couple its momentum when the "photon" collapses and delivers its energy. For instance, as the wavefront of what we later observe as a single photon propagates through space, we can make the space in front of it and behind it opaque to light, without upsetting the wavefront itself. To the extent we can do this whilst not altering appreciably the probability of its energy and momentum being deposited as a "photon" in the same locations as it would if we had not made space opaque, then we can reasonably say that our moving band of transparency in an otherwise opaque space contained the wavefront as it travelled.
So the momentum must exist in the wavefront, even if the source no longer exists and the destination is not yet known or does not yet exist.
A second thought experiment, considering how a prism bends light, shows that when the light enters a medium which slows it, that some of the momentum is physically coupled to the surface of the object where the slowing occurred. Likewise, when the light emerges back to full vacuum light speed, it pushes back against the surface it emerges from by a similar amount, to restore its vacuum speed full momentum. The proof is that the two surfaces of the prism deflect the light beam to a different path, and so give it some momentum at right angles to its initial trajectory, whilst also reducing its momentum in the initial direction. All the action in this occurs at the surfaces of the prism, so its here that the momentum must be partially coupled to the glass.
Because the deformation in the wavefront (due to it being slowed in the vicinity of the particle) is always behind the rest of the surrounding (full light-speed) wavefront, I believe this temporary coupling of momentum to the particle does two things:
- It
drags
the particle in the direction of the propagation of the
wavefront. This is a net conveyance of momentum - the particle
never gives back its momentum fully to the wavefront, whereas if the
wavefront travelled through a block of glass then on emerging, the
block gives up all the fraction of momentum it absorbed when the
wavefront entered it. This is a gut-feeling aspect of the
theory - highly qualitative and something I want to think about a lot
more. (For instance, if a solitary particle in space is bathed in
photons coming from one direction, the particle is continually
distorting the wavefront of the photons, which I think is doing work on
something which contains momentum. Therefore, I figure the
particle
must be tangling with the wavefronts and must, overall, gain some
momentum in the direction of the travel of the light. The
asymmetrical
deformation of the wavefront is what makes me think the particle is
exerting forces on the wavefront which are not entirely returned.)
- The
photon continues with less energy - a lower frequency and longer
wavelength. It is redshifted.
Such an effect would be hard to measure in the lab, because it is a very slight effect, and it can't generally be observed with coherent light, such as from a laser, since the coherence length (and therefore the thickness of the wavefront) is so long that we can't make a sufficiently sparse plasma for the inter-particle distances to be larger than the coherence length of the light. Doing the test with short coherence length light is easy - but that is of many wavelengths, so would be hard (surely impossible in the lab) to measure the redshift. Maybe using a solar furnace and an artificially created low-density plasma of a density such as that found in the transition region or low solar corona, it would be possible to measure some energy absorption in the test plasma by chopping the light on and off at tens to thousands of Hertz - and by measuring energy and/or momentum deposition in the plasma which is in phase with the chopped light. See the section below on experiments: #experiments .
For a given wavelength of light, with a given coherence length, the relationship between plasma density and redshift follows the pattern shown below. At zero density there is zero redshift. As the density (number of particles per cubic volume, which is proportional to the number of particles encountered by a "photon" per unit distance) increases, redshift rises linearly - until a limit is approached, after which the trend is to less increases in redshift, and then to less redshift and ultimately (at high enough densities) effectively no redshift whatsoever.
That limit is where the increased plasma density means the decreased the inter-particle spacing leads to the wavefront, on average, engaging with two ore more particles inside whatever volume of space (say roughly a sphere with diameter equal to the coherence length, or perhaps a three-dimensional ellipse, as long as the coherence length and about 1 or 2 wavelengths in diameter) which results in a generally continuous slowing of the wavefront. In other words, as the density approaches that required to provide an homogenous medium which slows the wavefront evenly rather than just close to each particle, the redshift declines towards zero.
Sorry about the lousy graph. I am yet to figure out how the redshift changes with wavelength. For instance, do X-rays generally have less redshift than visible light, because these "photons" seem to pass between most matter, even in solid objects? What of the CMB?
Perhaps if protons are greatly more effective at slowing light compared to electrons, then the plasma-redshift process would be relatively insensitive to the locations of the electrons. All that would matter is the wavefront encountering protons far enough away from other protons that it is like, to a significant extent, the wavefront enountering the proton on its own in vacuum.
The process could be more complex than I have imagined. Maybe it only happens when a proton (or other nuclei or ion) is close to an electron, but not close to any other protons. Perhaps the electron is needed for whatever mechanism by which the wavefront delivers a little of its energy. That would make sense, since without another object to push against, the wavefront can't give one particle any kinetic energy, except by delivering momentum - yet that momentum is the mass-energy of the energy it delivers.
If that is the case, then perhaps the process is an operation on an electron which is near a proton, but not as near or in orbit as an electron would be in an atom. This sounds reminiscent of Paul Marmet's theory about hydrogen atoms - just with the electrons further away, but still close enough that the wavefront can do something simultaneously to both particles. Maybe then it is a special case of free-free absorption, operating on a pair of particles, as free-free absorption does, but without the quanta being absorbed or scattered - just with it being redshifted a little. Maybe there's some wacky quantum mechanical effect which only manifests with two particles close to each other. However, any such two-particle theory seems predict that the strength of the effect would be proportional to density squared, since it can only happen when two particles are within some distance of each other. For instance, halving the number of protons and electrons in a cubic metre means that there is only 1/4 of them on average spending time within a certain distance of another particle. This would make the effect vastly less effective in low-density plasmas such as the solar wind or the Void IGM.
There are all sorts of implications of this redshift mechanism. For instance it would affect different wavelengths to different degrees for a given plasma, since such wavelengths would (typically) have different coherence lengths and so be more or less affected according to how this compared with the distribution of particle spacings. An initial, perhaps naive approach to this is as follows:
Emission lines would be less redshifted than continuum light (black-body, synchrotron etc.) because they are the product of resonant systems, which produce photons of long coherence length. The situation with absorption lines is trickier still - what would happen? Would the surrounding continuum wavelengths be redshifted a lot and so shift the whole line to the red? Or is the finely tuned absence of certain wavelengths itself a highly resonant, and therefore long coherence length, property of the light - making it only subject to redshift in a plasma with sufficiently large interparticle spacings? If so, then wouldn't the surrounding continuum wavelengths be shifted into the wavelength of the absorption line and fill it up? This is where we need to consider the entire system, potentially spanning the visible extent of the Universe, as a single quantum mechanical system.
Plasma redshift such as this could provide explanations for the cosmological redshift and the redshift of quasars - by one or more shells of plasma around the core which is sparse enough to redshift the light, but which is denser than the inter-cluster medium, or the Void IGM, and therefore provides more redshift per parsec than the Void IGM. (More on this below in the section on the Transverse Proximity Effect.) It could also explain the differences in redshifts between emission lines, which typically form near the core of the quasar, and absorption lines (formed by cooler gas further from the core) - an intervening layer of redshifting plasma makes the near-core emission lines become redshifted by the time they pass through the outer layer which does the absorption, with the distance between the two being very much less than that normally assumed based on the expanding Universe Doppler=distance theory. See footnote on different redshifts of QSO emission and absorption lines.
The "Lyman forest" can be explained by multiple clouds of H I (atomic hydrogen) fairly close (AUs to light years?) to the quasar core, but separated by bands of redshifting plasma. (Why there are such clouds containing neutral H, near a quasar and amongst all this plasma, is another question . . . ) If the jets of a quasar punch through such layers, then this would explain why we don't (usually) see Lyman forest absorption lines in the broad spectrum light of BL Lac objects. (I assume that what we call a "BL Lac object" is us looking at a quasar straight down one of its jets - but maybe there could, in some cases, be clouds of redshifting plasma and/or Lyman absorbing neutral H I beyond the lobes of the jets, which would give rise to Lyman forest lines and other absorption lines at one or more parts of the increasingly redshifted light's travel.) Likewise, Lyman forest lines could be explained by H I clouds well away from the quasar, in the midst of the IGM - with the IGM doing the redshifting between clouds. But I suspect that most of these clouds and/or shells of absorbing H I and redshifting plasma are local to the quasar's general vicinity - products of collapsing material, being heated and perhaps at times driven away by radiative losses coupling momentum to them. (Below I discuss how these Lyman Alpha forest clouds can be investigated with the Transverse Proximity Effect: #TPE .)
(This raises all
sorts of interesting questions. A quasar which is a black-hole
driven by accretion may starve itself if it is only feeding on sparse
IGM plasma / gas if the light it puts out is absorbed sufficiently to
radiatively accelerate the IGM away from the quasar. Then it
would fade, and the accretion would begin again . . . but it only takes
a star or dust cloud or the like to wander into the vicinity for there
to be
plenty of feedstock to power the quasar and keep up its heating and
repulsion of redshifting plasma in a shell around it.
Furthermore, do black-holes eventually (to us - but fairly quickly,
perhaps, according to the time experience of the infalling matter),
split due to some process driven by the angular momentum they
contain? If so, then this
could explain Halton Arp's theory that quasars are emitted by large
black holes / quasars in disturbed galaxies. The new-born
quasars,
or quasar pairs, would need to be initially surrounded by a substantial
shell of redshifting plasma to explain his observation / hypothesis
that
they are, at first, high redshift and low luminosity. Even if all
this is true, I wish I could think of a model to explain Halton Arp's
engaging thesis that certain types of galaxies pop out pairs of
quasars, in opposite directions, with those quasars maturing, becoming
lower redshift and higher luminosity, and then evolving into BL Lac
objects and ultimately clusters of
galaxies! This is such a fabulous, distributed, feminine,
story compared to the boy's own Big Bang - but I suspect that reality
is not quite like this charming theory.)
Assuming that the coherence length (AKA wavefront thickness) is generally proportional to wavelength, this redshift theory predicts less redshift for microwaves than for visible light - except where the plasma, such as the Void IGM, is of a sufficiently low density that the interparticle spacing is greater than the microwave coherence lengths. But perhaps, for shorter wavelengths, such as X-rays, the photons hardly ever encounter matter, so they are not redshifted very much at all. One possible arrangement of my plasma redshift theory is that only when a photon's path (this is fictitious - just one way of looking at the situation) comes within a wavelength or so of a particle, it is subject to redshift. Since X-rays are such short wavelengths, they tend to pass through plasma for longer distances on average than a photon of visible light before coming close enough to a particle to be affected by it.
I wonder about the redshift of 21cm and other emission lines of microwaves in the jets and lobes of radio galaxies, compared to the redshift of light from their optically visible core. Showing that these two redshifts were identical would be serious challenge to, or disproof of, my plasma redshift theory, since I guess that even in the Void IGM, the interparticle distance is not so long as to be greater than the coherence length of the 21 cm waves - so I would expect these waves to be less redshifted than visible light.
> this means "tilde greater than" by which I
mean "greater-than or about equal to".
