Distance Of Far Away Stars And Galaxies

[miklops]

Hi all

This is my first post - saw the opportunity to ask questions to professional astronomers and I've had this one on the back of my mind for a while.

I understand that the distances of stars are calculated from their redshift - that is, since the Universe is expanding at faster speeds the farther away from Earth you look, the more pronounced redshift would indicate a more distant star/galaxy.

My question is: how can we be so sure? Isn't that a big generalization? What if:

- the redshift isn't caused by a doppler-like effect? or
- the expansion of the Universe isn't constant in all directions? or
- the expansion of the Universe isn't constant at all distances?

I also read something about a class of variable stars that is used to 'calibrate' these redshift calculations - the reasoning being that since they basically always have the same brightness you can ascertain their distance from how bright they look... What if... their brightness wasn't constant across the Universe?

Seems to me like a kind of circular reasoning... What is it that breaks the circle and makes everyone so sure about redshifts and distances? (there must be something, it's just me who don't know what it is - therefore my question.)

Regards
Miguel
Lisbon - Portugal

[SpaceErnie]

This is a really good question. Hopefully, I can give you a good answer!

The short answer is that we can not be sure of any of these things, but a lot of experiments have beeb node to try to reduce the uncertainty. A few general things first:

The way Red-shift is measured is by looking for characteristic patterns in a star's light. Certain elements in the star (say, Hydrogen) emit light at a particular set of wavelengths. If you see light from a Star which has those same patterns of emission, but shifted in wavelength, then its a good bet that you are looking at red-(or blue-)shifted light from that element. The amount of shifting in wavelength tells you how fast the star is moving, relative to you.
For example, if the light spectrum of the element normally looks like this:
_____/\____/\/\_____

but in the distant star, it looks like this:
_/\____/\/\___________
Then, the amount of shift tells you how fast it is moving. This only works if you can find easily identifiable patterns though!

Regarding constancy of expansion, that is something that is confirmed by lookgin an all directions and seeing the same kind of red-shifts. Edwin Hubble did that when he discovered the expansion of the Universe. More recently, the COBE spacecraft did a much higher resolution job.

On the subject of variable stars, you are right to question the logic. The only way to confirm this theory is to find variable stars and measure their distance by yet another method. And that is exactly what is done. In the end, all astronomical measurements are based on paralax observations for near-by stars. (If you don't understand what I mean by paralax, I can explain that later&#33 Essentially, you measure the distance to the nearest stars using reliable methods (just geometry), and use those measurements to validate new methods which work at greater distances (such as the variable star method).

SpaceMan Ernie

[miklops]

Thank you!

However there are, to me, still a couple of 'loose ends' to that reasoning (again in my humble opinion):

- the same redshifts being observed in all directions could prove that redshifts are a reliable method of determining distance. However, couldn't two objects at the *precisely* the same distance have different redshifts (maybe they could be running away from us a different speeds?)? What if their distance can't be measured accurately through parallax (I understand for very distant objects this becomes impossible)? Then you'd have two different objects, at the same distance, but because of different redshifts, astronomers on Earth would conclude they are at very different distances from Earth than they actually are. How can that possibility be ruled out for very far away objects?

- the other is that you measure the distance of the variable stars by parallax and then you can deduce the distance of the more distant variables. But what if their brightness (the most distant ones) had no relation to the ones nearer to us? What if there's a class of variable stars that is even brighter, or dimmer, than the ones we 'know' close to us and therefore wreak all the assumptions? Is there a way of ruling out that possibility?

Thank you again
Miguel
Lisbon - Portugal

[SpaceErnie]

I'll answer the second point first... You are completely right. This is a real danger for this method of measurement, and one that has to be very carefully considered by any astronomer who uses it. It is fundamentally based on the assumption that the rules of Physics are constant at all places and times. If we see familiar phenomena in the spectrum of distant stars, we ASSUME that they are undergoing the same processes as those close to us. As you say though, it IS an assumption. A lot of effort has been spent to remove doubt from the process, but there is always some!
I'm not sure if I can give you a definite answer to the first question. One thing to note is that if two objects are at the same distance, but moving at different speeds, then that means either (a) they have not travelled in a straight line or (B) they started at different places, but travelled in a straight line.
(B) would make the Big Bang theory invalid.
(a) would mean that one had undergone significant acceleration that the other did not.
Either could be true. As you mention, Paralax measurement does not work at large distances because modern instruments are not sensitive enough. The best proof of the red-shift verses distance theory is to compare it with other measurements. As I said, locally, we can use paralax. Then you use that to calibrate measurement using star types with predictable brightness (assuming that they obey the same rules as local ones&#33. Then you use those measurements to verify other measurements with ,say, brightness of supernovas or size of galaxies. At each step, you rely on the accuracy of the previous step. And at each step, you can check to see if red-shift is still a good measure of distance. So far, nobody has found a good reason to doubt it. So far...!

[miklops]

Thank you again for your patient explanations!

A last question: has the optical redshift effect ever been replicated by experimental means (not including the comparison with the Doppler effect)? Can an experiment be devised to demonstrate the effect? I'm assuming it's not currently possible due to the limitations of current technology - namely we can't build a fast enough space probe, but maybe it could be tested in some other way, even if the detected effect is minuscule.

Very best regards
Miguel
Lisbon - Portugal

[SpaceErnie]

Hi Miguel,

Good to know that my answers are useful! Here's one for your last question.

Red-shift can be reproduced much closer to home. In fact, most solar-system exploration mission have had to cope with it in their communication links. The effect is much smaller, since the relative velocities are small (only a few 10s of km per second&#33. For instance, when the Cassini mission to Saturn releases the Huygens atmospheric probe on approach to Titan, the two spacecraft will have to deal with a noticeable shift in the received/transmitted frequency of the communication link.
Particle accelerator experiments (like CERN) produce far higher velocities and the effects of red-shift are much stronger in the light given off by the particles streams, moving at almost the speed of light.
Actually, now I come to think of it, some atomic clocks and optical cooling systems (for ultra-low temperature physics) rely on the doppler shift to produce the extremely low energy states that they need. That's another story though!

SpaceMan Ernie, UK

[GOURDHEAD]

Then you'd have two different objects, at the same distance, but because of different redshifts, astronomers on Earth would conclude they are at very different distances from Earth than they actually are. How can that possibility be ruled out for very far away objects?

Some uncertainity remains. To sort out the causal types of velocities of objects of this type, especially galaxies (stars are not discernable) at the larger distances, the relative magnitudes of the velocities due to the different causes can safely be assumed to be at large variance. Conditions that would cause a galaxy to have a "local" non-cosmological expansion caused velocity (on the order of < 10^+05 meters/second) comparable in magnitude to that of the expansion velocity (on the order of > 10^+07 meters/second) commensurate with its distance are not believed to exist, and if they did (i.e., gravitational slingshoting amongst members of a local group), would likely destroy or strangely modify the structure of the galaxy such that it would no longer be recognizable as such.

Similar arguments apply to nearer galaxies appearing to have redshifted velocities commensurate with much larger cosmological redshifts of more distant galaxies and with blueshifts otherwise masking their true cosmological expansion velocities.

Velocity magnitude is a handy theoretical limiter.