[madaxe_again]

There's an interesting implication here for redshift observations and Hubble expansion, insofar as it provides a mechanism for photons to lose energy while in a highly sparse medium, populated largely only by other photons.

I'm keeping my money on expansion and dark matter being hooey brought about by incomplete of incorrect understanding of light.

Only recently there was that rather interesting piece about simulated momentum transfer from photons in media.

It's exciting - we'll potentially be lopping off a huge branch of dead wood from the tree of science, from which new ideas can grow.

[m1el]

> I'm keeping my money on expansion and dark matter being hooey brought about by incomplete of incorrect understanding of light.

That's a very interesting hypothesis. Unfortunately, it's easy to verify that photon-photon scattering doesn't explain the expansion of the universe.

1) photon-photon scattering is elastic, so no energy is lost, and no redshift is occuring

2) if it isn't elastic, the scattering is a random process. So different photons will lose a different amount of energy, which means that measured spectra are going to be blurred.

3) rather than seeing redshift, due to photon-photon scattering, you'd see fog, which gets cloudier and cloudier with distance.

[madaxe_again]

Those are all solid points, so I retract - but the momentum paradox bit stands!

I wish I had someone else to chat physics with - isolation leads to screwy notions which can easily be quashed by the right counterpoints.

Edit: a thought. Please (genuinely!) tell me where I'm wrong. As it's elastic, could we not end up with groups of lower energy photons with the same vector as an original high energy photon, which would similarly explain redshift? Doesn't address your point re: fog, however, unless they're universally tightly grouped.

[triclops200]

It looks like from what we know, 2 high energy (it's much less common with low energy) photons collide and change direction: from what I understand, they don't split into low energy photons. It'd be basically impossible for all the high energy photons to scatter directly away from earth (why would the earth be special?) leaving only low energy ones towards us.

[Coding_Cat]

I doubt that is the case. As the article nores photon-photon scattering has been a predicted for a long time by QED. Even if it wasn't detected before it's impact on redshift could be calculated.

If you look at the abstract on the linked nature page the interaction has a cross-section of 70nb, which corresponds to a circle of radius 1.5e-9 nanometer[1]. It might occur a few times near a supernova or the swirl of a an accretion disk but I doubt it has a meaningful effect in deep-space.

A more exciting result would have been if they wouldn't have seen this effect. That would mean the theory was wrong and we had a new datapoint to look at.

[1]: these measurements are only for a specific energy range, they might vary dramatically with different energy levels but the key point is that this result agrees with theoretical predictions meaning that is should also be possible to calculate the contribution to redshift due to photon-photon scattering.

[madaxe_again]

but I doubt it has a meaningful effect in deep-space.

Don't underestimate how often unlikely events can happen in a big enough space...

As to photon-photon scattering, I might be wrong, but I don't believe it's considered in any models for expansion - similarly, if you do build light momentum transfer into your model, then expansion goes away - but because we "know" expansion to be true, those results aren't considered or published.

I think our givens are wrong. It'd hardly be the first time. I think I might be wrong. That'd hardly be the first time either.

But, as a betting physicist, my money is on expansion (at the very least acceleration) being bunkum.

[Coding_Cat]

>Don't underestimate how often unlikely events can happen in a big enough space...

That's true but the density of photons (from a given source) also drops of cubically as you move away from the source.

> imilarly, if you do build light momentum transfer into your model, then expansion goes away

I'm not really a cosmologist, but I know some physics (QFT in particular). If you have a derivation for a formula giving the impact of YY-scattering on redshift I'd love to read it.

> but because we "know" expansion to be true, those results aren't considered or published.

We know redshift to be true, and we have evidence for expansion based on measurements. Something that simply denies this will have a hard time being published. A theory that explains those findings without using expansion however would certainly make waves.

As and aside, since the cross-section of photon-photon scattering is energy dependent (and, as far as I can tell cosmological redshift is not), wouldn't that be a way to distinguish them? Scattering should occur more often at higher energies meaning that after a significant amount of scattering a bundle's spectrum should clump up more into the red.

[jpttsn]

The way I've always heard it, as a layman, is that red shift could be explained only by an expanding universe.

I'm sure there's more rigor to it than that, but it always seem s like a pretty big conclusion. A mildly surprising claim explained by a wildly surprising theory.

[madaxe_again]

As a physics grad, red shift can only be explained by an expanding universe with our understanding of light physics in the mid 20th century.

If you look at the world only through green glasses, you could be forgiven for thinking that everything was green. Your interpretation of observations is only as good as the theory you use to interpret. Think of epicycles, which for millenia seemed obvious and correct, until heliocentricity (for all bodies, not just earth) became the dominant model due to improved theory - but no new observations.

If light behaves even slightly differently to how we currently believe, everything from galactic rotation to expansion goes out of the window.

Here's the bit on momentum transfer - they even explicitly cite the implication for Hubble expansion.

https://www.sciencedaily.com/releases/2017/06/170630085627.h...

[contact_fusion]

The article referenced in that piece is available here, from the arxiv: https://arxiv.org/pdf/1603.07224.pdf The article has been up for some time, but has just now been published by Physical Review A; it is common practice for physicists to post a preprint on arxiv prior to publication, which happens after peer-review.

The momentum paradox you refer to is (possibly?) the Abraham-Minkowski controversy ( https://en.wikipedia.org/wiki/Abraham-Minkowski_controversy ) about electromagnetic momentum in dielectric media. I'm not an expert on this subject, but I would doubt that this new work definitively settles the controversy. Of course this is, no doubt, work towards settling the issue. My (limited) understanding is that the controversy is really about interpreting certain quantities that behave like momentum in certain contexts, and which contexts apply in certain experiments to measure them. I do not believe it constitutes a crisis in our understanding of light; this is a very technical detail.

Of course, using the mechanism of momentum transfer to the transmitting medium as an explanation of redshift - and by doing so, refuting the expansion of the universe - is just yet another "tired light" explanation. (This refers to the idea of explaining redshift through a path-dependent loss of energy for photons traveling from great distances.) This is not a new notion, dating at least to Zwicky in the fifties. This article ( https://arxiv.org/abs/astro-ph/0106566 ) details efforts to demonstrate the reality of this expansion. These efforts do not assume anything about the exact mechanism responsible for tired light, merely the notion that light loses energy as it travels. They refute this to better than 10 sigma.

>with our understanding of light physics in the mid 20th century. >If light behaves even slightly differently to how we currently believe, everything from galactic rotation to expansion goes out of the window.

I... suppose I have to admit that if the current theory of light is wrong, then there may be changes to how we interpret these results. You should know though that this would be extremely unlikely. Generally revolutions in physics tend to subsume the effective results of the theories that are replaced. Quantum mechanics provides a good example: if you take the limit h -> 0 (making Planck's constant zero), you recover classical mechanics. Relativity provides another: if you take the limit c -> infinity in relativity, you also recover classical mechanics. Our understanding of light is quite good.

Forgive me, but I'm detecting a little bit of an "international scientific conspiracy" vibe here. Unpopular work is published all the time, provided that it withstands scientific scrutiny. I know that sounds like I'm dodging the issue, but really, you can apply a cui bono here: what do scientists stand to gain by propping up "wrong" science? We don't get paid a lot, you know.

>Think of epicycles, which for millenia seemed obvious and correct, until heliocentricity (for all bodies, not just earth) became the dominant model due to improved theory - but no new observations.

I don't mean to pick nits, but it was improved observations (Kepler, Brahe, etc.) that drove the acceptance of the heliocentric model. Kepler was famous partly due to his unprecedentedly accurate measurements. Theory was not necessarily rigorous at the time, often referencing theological arguments; heliocentrism was hotly debated, but not novel. Later it was realized that epicycles form a basis set for any trajectory on the surface of a sphere; any trajectory can be reproduced using a sufficiently large number of them. This was a pitfall that astronomers of the time could never have known. Remember that they did not have Newton's insights yet.

On a final note, its great to hear that ATLAS has some good evidence for gamma-gamma interaction!

edit: fix link

[yread]

do you mean interaction between high energy gamma rays from distant quasars? That could change background radiation density, right?