Blog Entries: 9
Recognitions:
Gold Member

## No redshift in a freely falling frame

 Quote by GAsahi I derived it myself, at the beginning of this thread, here.
Ah, I see. But this derivation only holds for purely radial motion; in fact it only holds when one object is static at a constant height and the other is moving purely radially. The case stevendaryl talks about, as here...

 Quote by stevendaryl In the case of two observers at the same radius R, moving away from each other at speed v, your derivation gives the wrong answer.
...is more general. You gave a "hint" as to how a more general formula could be derived, but you haven't actually done the derivation. (A fully general formula would need to include the effects of motion, both radial and non-radial, for *both* objects, not just one.)

Also, as you noted in a later post, what you call "v" in your formula is a coordinate velocity, not an actual observed relative velocity.

Finally, since you mentioned Ashby's paper, it's worth noting that your derivation uses Schwarzschild coordinates, but his paper does not. He uses Earth Centered Inertial (ECI) coordinates, which differ from Schwarzschild coordinates in several important respects:

(1) They are isotropic;

(2) The coordinate time dt is scaled to the rate of time flow of observers on the "geoid" (the equipotential surface at "sea level") who are at rest relative to the actual Earth (i.e., rotating with it); however, the simultaneity convention is that of hypothetical inertial observers moving with the Earth's center of mass but *not* rotating with it (which would be the simultaneity convention of Schwarzschild coordinates centered on the Earth);

(3) The "potential" $\Phi_{0}$ in the metric, which is the potential on the "geoid", includes not only the effect of the Earth's rotation, but also includes corrections for the Earth's quadrupole moment, so it differs in two ways from the Schwarzchild potential.

These differences don't affect the general points under discussion, but since you referenced Ashby's paper, I think it's worth pointing out the ways in which his equations and notation differ from those being used in this thread.

 Quote by PeterDonis Ah, I see. But this derivation only holds for purely radial motion; in fact it only holds when one object is static at a constant height and the other is moving purely radially.
1. stevendaryl denied for the longest time that it is correct, resulting into a monumental waste of time

2. the derivation is easy to extend to more complex cases, like the one of a stationary receiver and a rotating transmitter, the methodology is the same.

3. the derivation extends easily to other metrics (as in the Ashby paper). The point is that the methodology is STANDARD and that it produces predictions that are confirmed by experiment (Pound Rebka for radial motion, GPS for more complex motion, etc)

 ...is more general. You gave a "hint" as to how a more general formula could be derived, but you haven't actually done the derivation. (A fully general formula would need to include the effects of motion, both radial and non-radial, for *both* objects, not just one.)
...because the derivation is trivial

 These differences don't affect the general points under discussion, but since you referenced Ashby's paper, I think it's worth pointing out the ways in which his equations and notation differ from those being used in this thread.
Agreed. The point was to show stevendaryl that the applied methodology is STANDARD.

 Quote by GAsahi You need to get your ratios right: $$\frac{d \tau_1}{d \tau_2}=\frac{f_2}{f_1}$$
My mistake, but either way, it's incorrect in the case that I was talking about, namely two observers, at the equator, moving at relative speed v in opposite directions (east for one, west for the other). In that case
$$\frac{d \tau_1}{d \tau_2}$$ = 1.
$$\frac{f_2}{f_1}$$ = √((1-v/c)/(1+v/c))

 Quote by GAsahi 1. stevendaryl denied for the longest time that it is correct, resulting into a monumental waste of time
No, I said that your DERIVATION was wrong. From the very beginning, I said that it was your method that was incorrect, not the result. It happens to give the right answer in one situation, but not in other situations.

Blog Entries: 9
Recognitions:
Gold Member
 Quote by GAsahi Agreed. The point was to show stevendaryl that the applied methodology is STANDARD.
I agree, but it's worth noting that part of what is standard is the adoption of particular coordinates--you used Schwarzschild coordinates, the Ashby paper used ECI coordinates (which seem to be the "standard" for these kinds of computations). That includes adopting a particular simultaneity convention, which is crucial for defining the time differentials and intervals that appear in the equations. I agree it's a "natural" choice of simultaneity convention for the purpose, but it's still a specific choice.

Blog Entries: 9
Recognitions:
Gold Member
 Quote by stevendaryl I said that it was your method that was incorrect
The general method he used is correct, provided you adopt the particular coordinates and simultaneity convention he adopted. (The particular derivation he gave was for a special case, yes.) If you are saying those coordinates and that simultaneity convention are not the only possible ones, that's true; and in different coordinates you would use a different method for defining "relative clock rates". An observer on a satellite in low Earth orbit, for example, would see Earthbound clocks (clocks at rest on the Earth's surface and rotating with it) to be "running slow" if he used his own local inertial coordinates; but in the ECI frame his clocks would be "running slow" relative to Earthbound clocks (I believe that's right for low enough orbits--the GPS satellite orbits are very high, 4.2 Earth radii IIRC). However, the observed frequency shift for light traveling between the two observers would be independent of which coordinates you adopted.

 Quote by PeterDonis The general method he used is correct, provided you adopt the particular coordinates and simultaneity convention he adopted.
I don't see that his derivation works in the case of two observers on the surface of the Earth, one moving east and one moving west, with relative speed v. His method would give a redshift of zero (or at best, of the order of (v/c)2), which is not the correct answer. The correct answer should, in the limit as rs→0 approach the Doppler shift formula.

 (The particular derivation he gave was for a special case, yes.) If you are saying those coordinates and that simultaneity convention are not the only possible ones, that's true; and in different coordinates you would use a different method for defining "relative clock rates". An observer on a satellite in low Earth orbit, for example, would see Earthbound clocks (clocks at rest on the Earth's surface and rotating with it) to be "running slow" if he used his own local inertial coordinates; but in the ECI frame his clocks would be "running slow" relative to Earthbound clocks (I believe that's right for low enough orbits--the GPS satellite orbits are very high, 4.2 Earth radii IIRC). However, the observed frequency shift for light traveling between the two observers would be independent of which coordinates you adopted.
That's exactly my point: redshift is coordinate-independent, while "relative clock rate" is coordinate-dependent. The two are only equal in certain circumstances.

 Quote by stevendaryl My mistake, but either way, it's incorrect in the case that I was talking about, namely two observers, at the equator, moving at relative speed v in opposite directions (east for one, west for the other). In that case $$\frac{d \tau_1}{d \tau_2}$$ = 1. $$\frac{f_2}{f_1}$$ = √((1-v/c)/(1+v/c))
First off, the formula $\frac{f_2}{f_1} = \sqrt{((1-v/c)/(1+v/c))}$
does not apply for circular motion. I know for a fact that the relativistic Doppler is more complicated for accelerated motion. Note the presence of the terms in $\frac{v^2}{c^2}$ here.
More importantly, you do not get redshift as reflected in your formula, you get blueshift for half the circle (decreasing closing distance) and redshift only for the other half (increasing closing distance) so, on average, $\frac{f_1}{f_2}=1$ due to problem symmetry.
Lastly, we were discussing the correctness of my formulas/derivations for radial motion. Failing to prove your point , you quietly moved the goalposts to circular motion. But even then, you failed to acknowledge my repeated references to the Ashby paper that shows how this problem gets solved for the case of circular motion. The solution uses a different metric but the methodology is the same, all the effects (with the notable excption of the Sagnac effect) fall out the metric.

Blog Entries: 9
Recognitions:
Gold Member
 Quote by stevendaryl I don't see that his derivation works in the case of two observers on the surface of the Earth, one moving east and one moving west, with relative speed v. His method would give a redshift of zero (or at best, of the order of (v/c)2), which is not the correct answer. The correct answer should, in the limit as rs→0 approach the Doppler shift formula.
The correct answer for what? For the observed frequency shift of light sent between the two observers? Or for the relative clock rates of the two observers? You yourself pointed out that the two are not necessarily the same; this is a case where they are different.

A more general formula for the rate of a moving clock relative to coordinate time, generalizing the one that GAsahi derived, is this:

$$\frac{d\tau}{dt} = \sqrt{1 - \frac{2 G M}{c^{2} r}}\sqrt{1 - \frac{v^{2}}{c^{2}}}$$

which is similar to the formula in Ashby's paper, except that I have used Schwarzschild coordinates instead of the ECI coordinates that he used. (Also, he expands out the square roots and throws away higher order terms.) Note also that v is relative to a non-rotating observer, so an observer at rest on the Earth's equator has a v of about 450 m/s eastward.

So the ratio of clock rates for two observers, both at the same radius r but moving at different velocities, is:

$$\frac{d\tau_{1}}{d\tau_{2}} = \frac{\sqrt{1 - \frac{v_{1}^{2}}{c^{2}}}}{\sqrt{1 - \frac{v_{2}^{2}}{c^{2}}}}$$

where the terms in M cancel since both observers are at the same radius r. This does look different from the Doppler formula, but so what? Why would we expect the two to be the same for this case?

Blog Entries: 9
Recognitions:
Gold Member
Missed a couple of items in an earlier post, just wanted to clarify:

 Quote by stevendaryl I think that you agree that the instantaneous ratio (d$\tau$1/dt)/(d$\tau$2/dt) is always defined for any coordinate system, but is coordinate-dependent.
Yes.

 Quote by stevendaryl Isn't "there is a timelike Killing vector field" and "there is a coordinate system in which the components of the metric are time-independent" the same thing?
Yes.

 Quote by stevendaryl Let me do an explicit calculation to prove my point. In Rindler coordinates (X,T), we have two clocks, one at X = X1, and one at X2. The Rindler interval is: d$\tau$2 = X2 dT2 - dX2 So for clocks at rest in the X,T coordinates, we have: d$\tau$ = X dT So the ratio of the rates is: d$\tau$1/d$\tau$2 = X1/X2 Conclusion: the "higher" clock (with greater X) runs faster. Now, do the same calculation in the coordinate system (x,t) related to (X,T) through: x = X cosh(gT) t = X/c sinh(gT) So d$\tau$2 = dt2 - 1/c2 dx2 = dt2 (1 - v2/c2) where v = dx/dt = the speed of the clock. So d$\tau$ = $\sqrt{1-(v/c)^{2}}$ dt 1)The ratios of the rates in this coordinate system is given by: d$\tau$1/d$\tau$2 = $\sqrt{1-(v_{1}/c)^{2}}$/$\sqrt{1-(v_{2}/c)^{2}}$ At t=0, $v_{1}$ = $v_{2}$ = 0. So the ratio starts off equal to 1, not X1/X2.
 Quote by GAsahi x = X cosh(gT) t = X/c sinh(gT) So d$\tau$2 = dt2 - 1/c2 dx2 = dt2 (1 - v2/c2) where v = dx/dt = the speed of the clock. So d$\tau$ = $\sqrt{1-(v/c)^{2}}$ dt The ratios of the rates in this coordinate system is given by: d$\tau$1/d$\tau$2 = $\sqrt{1-(v_{1}/c)^{2}}$/$\sqrt{1-(v_{2}/c)^{2}}$ 2)This part of the derivation is in error. If you did it correctly, you would have gotten that the correct result is $\frac{d \tau_1}{d \tau_2}=\sqrt{\frac{1-v/c}{1+v/c}}$ where $v$ is the instantaneous speed of the rocket containing the two clocks wrt the launcher frame.

Originally Posted by PeterDonis View Post

 It looks to me like these two quantities refer to two different things. The first refers to Schwarzschild spacetime; the second refers to Rindler coordinates on Minkowski spacetime. The answers for those two cases will not be the same, because Schwarzschild spacetime is curved and Minkowski spacetime is flat.

 Quote by stevendaryl No, both are about flat spacetime. The difference is that √(1-(v1/c)2)/√(1-(v2/c)2) is the ratio of the two clock rates, as measured in the "launch" frame, while √(1-v/c)/√(1+v/c) is the redshift formula for the case in which the "lower" clock sends a signal while at rest, and the signal is received by the "upper" clock when that clock is traveling at speed v. (Since the light signal takes time to propagate, the upper clock will have achieved a nonzero velocity while the light signal is in flight). 3)My point is that the redshift formula is NOT the same as the ratio of clock rates, 4)except in very specific circumstances. Those circumstances actually hold for Rindler coordinates and for Schwarzschild coordinates, but they don't hold for arbitrary coordinates. The conditions for being able to equate "relative clock rates" with "redshift" are: (1) The metric tensor is independent of time, and (2) the sender and receiver are at rest in the coordinate system.
Perhaps you could clarify a point of confusion.
1)You stated that wrt the launch frame the front and rear clocks would be related by the ratio of gammas.

2)GAsahi declared that incorrect and said they would be related by the Doppler factor.
Indicating that they were not equivalent.

3)You then agreed with him that they were not equivalent in general

4) But were equivalent in this instance

So my question is:
Is the ratio of gammas as measured in the launch frame exactly the same ratio as derived from the Rindler metric ?

Ratio of Velocity gammas =Doppler ratio=Rindler ratio??

Or not?

thanks

 Similar discussions for: No redshift in a freely falling frame Thread Forum Replies Introductory Physics Homework 5 Introductory Physics Homework 3 Introductory Physics Homework 4 Introductory Physics Homework 3 Introductory Physics Homework 5