Venus during the day, a test of relativity?

[Vast]

At the moment you can easily see Venus at any time during the day with a small set of binoculars, or even with the naked eye if you look closely enough and know where to look-around about 31 degrees away from the sun. I was thinking about the experiment to test relativity, where starlight was bent by the suns gravity during an eclipse, and wondered whether Venus could be used in the same manner. Mind you, the glare from the sun, increases the closer it is, and for this experiment to work, Venus needs to be on the other side of the sun to us, and consequently reduces in brightness. Nonetheless, with a modest telescope, wouldn’t one still be able to conduct this experiment?

 

[Nugatory]

In principle, yes, but in practice not with a modest telescope because of the sun's glare. (There are of course serious safety issues around telescopic observations of the sun as well).

Eddington did this in 1919, but it required sending an expedition to the southern hemisphere to catch a solar eclipse to eliminate the glare - and then he used distant stars instead of Venus because that's what was behind the sun at the time of the eclipse.

It's more practical to look for distant galaxies distorting the light from even more distant objects (google for "gravitational lensing") but that is most certainly not a job for a modest telescope.

 

[PeterDonis]

he used distant stars instead of Venus because that's what was behind the sun at the time of the eclipse

He used the particular distant stars that were behind the Sun at the time of eclipse, but that's not the only reason he used distant stars. He used distant stars because their position on the sky relative to each other does not change over a period of six months or so, so he could measure their position on the sky six months before or after the eclipse, and then measure it again during the eclipse, and compare the two. That won't work with a planet like Venus because its position on the sky does change on a shorter timescale than six months.

 

[Nugatory]

That won't work with a planet like Venus because its position on the sky does change on a shorter timescale than six months.

The position does change, but we know enough about planetary motion to be calculate where Venus "ought" to be at the time of the eclipse... the analysis would be somewhat analogous to the way that Romer was able to estimate the speed of light from the eclipses of Jupiter's moon Io.

But of course this is all beside the point because Venus wasn't in the right place, and even if it were it's not obvious that the calculation would be any easier than the fixed star one.

 

[PeterDonis]

we know enough about planetary motion to be calculate where Venus "ought" to be at the time of the eclipse

I could see doing this as a demonstration if light bending by the Sun is already accepted. But I'm not sure it would be a valid test of that theoretical prediction if it were not already accepted.

 

[mfb]

The deflection is at most 1.75 arcseconds if Venus is just in the right place, or 900 km at the orbit of Venus. With radar astronomy it is easy to measure and predict its place that good, but with the methods of 1919? In addition the deflection is less than the size of Venus: You have to resolve its disk and measure its geometrical center or something similar.

 

[Ibix]

He used distant stars because their position on the sky relative to each other does not change over a period of six months or so, so he could measure their position on the sky six months before or after the eclipse, and then measure it again during the eclipse, and compare the two.

To answer this objection, one could measure the apparent speed of Venus compared to distant stars as it approached the Sun's limb. Because light deflection means you can see it when naive ray tracing would suggest it's behind the Sun its apparent speed must dip before it passes behind the Sun and pick up again after it emerges. You would not expect this behaviour in a transit.

As noted by others, the precision needed is beyond casual observation.

 

[PeterDonis]

its apparent speed must dip before it passes behind the Sun and pick up again after it emerges

I'm not sure this will work because the light from the distant stars will be bent as well. But I have not done the math.

If this worked, it wouldn't be a "dip" so much as a gradual decrease, then a gradual increase. The light bending changes smoothly as a function of angle--with radio waves it has been measured, IIRC, for angles as far as 90 degrees away from the Sun.

 

[mfb]

The light bending changes smoothly as a function of angle--with radio waves it has been measured, IIRC, for angles as far as 90 degrees away from the Sun.

Gaia has to take it into account with visible light. It looks away from the Sun but the deflection is still a few milliarcseconds while it measures star positions with an uncertainty of several microarceconds.