Gravitational Bending vs. Refraction

[Gort]

I'd like to call attention to this paper: https://arxiv.org/ftp/physics/papers/0409/0409124.pdf
I'm not trying to question the validity of GR in bending of light near a star (such as the sun), as this paper apparently does. But surely Newtonian refraction of light passing through the sun's upper atmosphere must play a role. How does one separate the contributions due to Newtonian refraction and GR in computing the total observed effect in the bending of a light path?

 

[Ibix]

I would expect that you propose a model of the composition and density of the medium, deduce its refractive index and absorption spectrum as a function of distance from the star, and then calculate the effect on light paths, presumably using something fun like Maxwell's equations in a medium in curved spacetime (or maybe you can get away with Huygens' Principle). Then you go and track the apparent position and brightness of a star in a range of wavelengths as it approaches the star.

Significant chromatic aberration would be a simple dead giveaway for this. Light in a vacuum in GR will have the same path independent of frequency. Light in a medium will not (unless you tune your medium's properties super carefully, I suppose).

 

[Gort]

Although I agree that can be done, was that actually done during initial tests of GR? I seem to recall that the stellar light deflection measuring by Eddington during the 1919 eclipse was "precisely" what GR predicted. No mention of refraction through the solar atmosphere that I recall. Did I miss something?

 

[Ibix]

I don't know if Eddington worried about it or not. It's worth noting that the corona is not stable, so its effects (if at all significant - I don't know) would be like a mirage - i.e. would show random variation. Eddington is far from the only person to have studied gravitational lensing, and I know it's been done in multiple wavelengths and with multiple different bodies (not just the Sun). Anything is possible, of course, but I find it difficult to imagine that circumstances have conspired so well to make regular refraction look like GR under such a range of circumstances (your paper's author obviously disagrees, but you don't seem to be asking my thoughts on that).

It's worth noting that "precise agreement" is a rather optimistic description of Eddington's results. He was certainly a lot closer to Einstein than the semi-Newtonian model he was comparing against, but there was an element of showmanship in how Eddington (an Englishman) handled reporting tests of work by Einstein (a German speaker with a Germanic name) in 1919.

 

[stoomart]

I'd like to call attention to this paper: https://arxiv.org/ftp/physics/papers/0409/0409124.pdf
I'm not trying to question the validity of GR in bending of light near a star (such as the sun), as this paper apparently does. But surely Newtonian refraction of light passing through the sun's upper atmosphere must play a role. How does one separate the contributions due to Newtonian refraction and GR in computing the total observed effect in the bending of a light path?

Quick note about the paper and a later one: the author is not an endorser, they were not published in a peer-reviewed journal, and don't have any citations. Stars aside, my problem replacing gravitational lensing with refraction is there's no atmosphere around black holes. The author redefines BHs as a "shrunk neutron-star surrounded by a thick glassy skirt of atmosphere".

 

[vanhees71]

I even wonder, how this paper stayed on the arXiv! One very superficial but usually correct prejudice is that you shouldn't trust papers written in Word (at least not in theoretical physics).

 

[DrClaude]

Thread closed pending moderation.

 

[Nugatory]

I even wonder, how this paper stayed on the arXiv! One very superficial but usually correct prejudice is that you shouldn't trust papers written in Word (at least not in theoretical physics).

The paper was posted to arXiv back in 2004, before they had discovered the need for quality control. Today it would end up on vixra or worse.

 

[Nugatory]

Thread closed pending moderation.

And we can leave it closed - the source upon which it is based is clearly bogus.

 

[mfb]

One more comment: The Gaia telescope sees the deflection of starlight by the Sun with every single star observation - at 1 AU distance and with an angle of 45° to 135° relative to the Sun, far away from any sort of stellar atmosphere. Without a correction for this effect the Gaia data wouldn't make any sense; it is clearly there.
Gaia is expected to measure this deflection with a relative precision of a few parts in a million.