Doppler redshift in a vacuum requires physical source

[iantresman]

As far as I understand it, the Doppler redshift requires either a moving physical source or physical observer. Does this mean that a photon moving in a vacuum can not be redshifted UNTIL it comes into contact with some form of moving matter (the observer)? And presumably a redshifted photon in a vacuum must have been produced by a moving physical sources?

Is there any energy exchange with the source or observer (radiative transfer?).

Regards,
Ian Tresman

[Cougar]

A crucial word is missing in your statements and questions: the word "relative". Corrections, then, would read as follows:

As far as I understand it, the Doppler redshift requires relative motion between source and observer.

Does this mean that a photon moving in a vacuum can not be redshifted UNTIL it comes into contact with some form of moving matter (the observer)?

Oh, I see what you're getting at. But again, it's a relativity thing. The photon can be redshifted as soon as it reaches a different frame of reference from its source.

And presumably a redshifted photon in a vacuum must have been produced by a moving physical sources?

Moving relative to some other frame of reference.

[Fortis]

It's akin to talking about the kinetic energy of a rocket in space. It depends on the inertial reference frame that you are interested in.

To discuss the wavelength/frequency of a photon, you need to define the inertial reference frame as well, otherwise it is a meaningless concept. Once you have done that, then you can work out what the wavelength/frequency is in another frame. You don't, however, have to explicitly detect it in the other frame to know what it would be.

[Tim Thompson]

As far as I understand it, the Doppler redshift requires either a moving physical source or physical observer. Does this mean that a photon moving in a vacuum can not be redshifted UNTIL it comes into contact with some form of moving matter (the observer)?

I think that is correct, but only in the case where the emitter & receiver share the same inertial reference frame. Since the Doppler shift depends on the relative motion of emitter & receiver, with respect to each other, one cannot determine redshift, blueshift, or noshift, until both have been sufficiently specified. But in the case of accelerated reference frames, one can argue that the photon is redshifted independent of an observer, by virtue of the reference frame. But, of course, you do need an observer, to make an observation.

And presumably a redshifted photon in a vacuum must have been produced by a moving physical sources?

Or a stationary source & a moving receiver (observer).

Is there any energy exchange with the source or observer (radiative transfer?).

I don't really understand the point of the question. The photon is usually (maybe "always") destroyed by the act of detecting it (certainly "always" is the case in astronomy). That means the energy of the photon is transferred to the detecting device, in these days that would likely be a CCD detector. In that case, the energy of the photon is transferred to the electrons in the detector, and the electrons are then read out of the detector to determine the number & energy of "detected" photons.

[Nereid]

[snip]

Is there any energy exchange with the source or observer (radiative transfer?).

I don't really understand the point of the question. The photon is usually (maybe "always") destroyed by the act of detecting it (certainly "always" is the case in astronomy). That means the energy of the photon is transferred to the detecting device, in these days that would likely be a CCD detector. In that case, the energy of the photon is transferred to the electrons in the detector, and the electrons are then read out of the detector to determine the number & energy of "detected" photons.

I too would like a clarification Ian.

However, my read of the question is different from others' in this thread (so far).

I understand the OP question to be about the existence of photons, and their nature (energy, wavelength, ...), in the absence of an observer (a bit like Schrödinger's cat). Or, expressing it in a different way, do photons have a 'doppler' property, analogous to the 'spin' property used (for example) in experiments on the EPR paradox?

[Sam5]

As far as I understand it, the Doppler redshift requires either a moving physical source or physical observer. Does this mean that a photon moving in a vacuum can not be redshifted UNTIL it comes into contact with some form of moving matter (the observer)? And presumably a redshifted photon in a vacuum must have been produced by a moving physical sources?
...
Regards,
Ian Tresman

Davis and Lineweaver said in their Scientific American article last year:

“The primary observation that the universe is expanding emerged between 1910 and 1930. Atoms emit and absorb light of specific wavelengths, as measured in laboratory experiments. The same patterns show up in the light from distant galaxies, except that the patterns have been shifted to longer wavelengths. Astronomers say that the galactic light has been redshifted. The explanation is straightforward: As space expands, light waves get stretched. If the universe doubles in size during the waves' journey, their wavelengths double and their energy is halved.”

Full article here:

http://www.sciam.com/print_version.c...2383414B7F0147

[iantresman]

I too would like a clarification Ian.

The Wikipedia redshift description says that "A single photon propagated through a vacuum can redshift in three distinct ways.". I wasn't sure whether this was accurate.

This seems accurate for the cosmological redshift and relativistic redshifts, but not the Doppler redshift. Sure, you can have a conceptual moving source or observer, but in practice, the Doppler redshift must require a REAL physical moving source of observer. In which case it doesn't take place in a vacuum.

It had been suggested that "curved space" would cause a Doppler redshift... doesn't curved space imply a cosmological or relativistic redshift? And if we observed such a redshift, how could we distinguish between Doppler due to relative motion of the source, and that due to curved space; how do we know that there is no motion, and the Doppler effect is not due entirely to curved space?

As for energy exchange, I was trying to compare the Doppler redshift with scattering effects, where energy is transfered from the photon to matter. If the Doppler redshift requires a source or observer, then is the Doppler effect due to an energy transfer, but I think I now understand it is due to a change of frame of reference, and there is no energy transfer.

Regards,
Ian Tresman

[antoniseb]

This is all because you are nitpicking the language in Wikipedia? Come on Ian, you know that Wiki tries to provide information on everything, but it is not always written for precision. I agree with you that during the photon's travels through the vacuum (if you don't count the instants of the beginning and end of the path), it is not doppler shifted.

[iantresman]

This is all because you are nitpicking the language in Wikipedia? Come on Ian, you know that Wiki tries to provide information on everything, but it is not always written for precision. I agree with you that during the photon's travels through the vacuum (if you don't count the instants of the beginning and end of the path), it is not doppler shifted.

And does curved space cause a Doppler redshift, or is this indeed a cosmological or relativistic redshift?

Yes, I am indeed nitpicking the language, Wikipedia is an encyclopedia, not a scentific text, and I think that it can be both accurate, and descriptive.

Regards,
Ian Tresman

[antoniseb]

1. does curved space cause a Doppler redshift, or is this indeed a cosmological or relativistic redshift?
2. Yes, I am indeed nitpicking the language, Wikipedia is an encyclopedia, not a scentific text, and I think that it can be both accurate, and descriptive.

1. I wouldn't call any redshift that happens to a photon because of traversing 'curved space' doppler. Doppler to me is just about relative velocities of the source and destination.

2. I'm honored (on behalf of the the entire community) that you are checking with us before sending a corrective note to the Wiki people.

[Eroica]

I think it's misleading to talk about a photon being redshifted. Photons don't change. Redshift is a change in the way the observer perceives the photon. It's possible for an identical stream of photons to be observed redshifted by one observer, blueshifted by a second observer, and "unshifted" by a third observer.

The problem with first establishing the relative motion of source and observer is that it seems to presuppose that both coexist. Imagine a quasar emitting a photon 10 billion years ago. The Earth didn't exist then, and perhaps today when the photon is observed the quasar no longer exists. So what happens here?

Do we simply observe by how much the light has been redshifted and then work out what the relative motion would have to be? If so, that's doing things the other way round.

[Ken G]

I think the key point that has not yet emerged from this thread is that a redshift is an observed thing, end of story. It requires no relativity, relativity is a theoretical concept we use to try to explain the redshift (just as gravity is a theoretical idea we use to understand why things fall, but they do whether we understand it or not). But the important point is, in general relativity you get redshifts in coordinate independent ways. But people trying to explain why the redshift is there are not usually content with saying "because Einstein's equations say so", so they go a step further and choose a coordinate system . It is the coordinatization that generates the reference frames, and also the words we use to describe what happened. This is as true in everyday special relativity as it is in general relativity-- you can't say when a Doppler shift occured in any kind of absolute way, it depends on the chosen coordinatization. That's why I think we should lump Doppler shifts together with shifts that appear due to coordinates that change in time and call that whole business a generalized Doppler shift, but I've never seen this done.

Cosmological redshifts, on the other hand, are a net effect of contributing factors. I would argue that cosmological redshifts comprise of two uniquely determinable components, and here I deviate completely from the standard nomenclature. I say there is a gravitational part and a generalized Doppler part, where the gravity part is from the unique curvature of spacetime, and the generalized Doppler part is the rest. The former comes from the curvature of the metric, and the latter comes from the other aspects of the metric that are not related to curvature, and can decompose arbitrarily into motion of source, motion of observer, and motion of coordinates themselves. Why on Earth gravitational redshifts are reserved for local effects is beyond me-- the local effects, both near source and observer, are easy enough to consider independently from the average behavior of the matter in the universe as a whole and should be thought of in a completely different category than the cosmological redshifts, although they are also net effects that can be broken down in similar ways.

[iantresman]

2. I'm honored (on behalf of the the entire community) that you are checking with us before sending a corrective note to the Wiki people.

I do try and check things first, or have at least some form of citation.

I don't supposed some of you guys would like to help in editing the Wikipedia redshift article?

Regards,
Ian Tresman

[iantresman]

I think the key point that has not yet emerged from this thread is that a redshift is an observed thing, end of story.

I think this is where the Wikipedia Redshift article has gone astray, and it is not clear what is observed, what is theory, and why.

Regards,
Ian Tresman

[Ken G]

Yes, I agree with that, from the flavor of the quote you included.

[Tim Thompson]

As I see it, redshift is a measured quantity. After all, it is defined in terms of the difference between a rest-frame wavelength, and an observed wavelength. So how can you know what the redshift is, until you have an "observed" wavelength? That much should be obvious.

The open questions are: (1) What causes the redshift, and (2) Is the redshift an intrinsic property of the photon, or is it strictly an effect of the relative motion of observer & source?

In the case of a Doppler redshift, the answer is obvious: It is strictly an effect of the relative motion of observer & source. Any individual photon could be observed to be redshifted, blueshifted, or noshifted, depending on the observer.

But in the case of cosmological redshifts, as interpreted in the framework of general relativity (GR), the answer is not so obvious (at least not to me). We do know for certain that in this case the cosmological redshifts are not Doppler shifts. And we know that the observed redshift is a combination of Cosmological redshift & Doppler shift. I am reluctant to grant "relativistic redshift" the independence that the Wiki article does, because it is a Doppler shift, but carried into special relativity. It is really only a special case of the more general Doppler shift, caused by relative motion between observer & source.

What we (or at least I) do not know for certain is this: Is the cosmological redshift an intrinsic property of the photon or not? I am inclined to answer in the affirmative, that a cosmological redshift is an intrinsic property of the photon. My reason is that this cosmological redshift in GR is an effect not of the relative motion between observer & source, but rather an effect of the spacetime that extends between observer & source. So it is no longer enough to know how the observer & source move relative to each other, it is necessary to have a model for the spacetime trajectory of the photon through the universe (in this case, a GR model). Outside of the esoterica of quantum field theory, reality & spacetime are inseperable. The intrinsic properties of the photon (and of everything else) are described in terms of spacetime. So any changes induced by spacetime must be considered intrinsic to the photon.

That's the way I see it.

[Sam5]

Starlight Redshifts were predicted by Doppler, 27 years before they were first observed and measured with a spectroscope by Huggins.

[iantresman]

As I see it, redshift is a measured quantity. After all, it is defined in terms of the difference between a rest-frame wavelength, and an observed wavelength. So how can you know what the redshift is, until you have an "observed" wavelength? That much should be obvious.

I agree with that. With regard to the relativistic redshift, you suggest that it is similar to the Doppler effect. Is it the same as the relativistic Doppler effect, and applies to objects moving toward the speed of light?

But isn't the relativistic redshift also known as the gravitational redshift, which might be exhibited by a stationary black hole?

Regards,
Ian Tresman

[Cougar]

I think the key point that has not yet emerged from this thread is that a redshift is an observed thing, end of story. It requires no relativity, relativity is a theoretical concept we use to try to explain the redshift (just as gravity is a theoretical idea we use to understand why things fall, but they do whether we understand it or not).

I'm not sure what your point is here, Kenny G. You didn't end the story where you said the story ended. You said relativity is not required, but then you went on to say...

But the important point is, in general relativity you get redshifts in coordinate independent ways.

So relativity is not necessary... unless you want to have some idea about what is going on? Is relativity just "a theoretical concept", or is it an accurate explanation about how Nature works? A natural law, maybe?

But people trying to explain why the redshift is there are not usually content with saying "because Einstein's equations say so", so they go a step further and choose a coordinate system.

I swear you've got "coordinate systems" on the brain. Yes, normally we choose the frame that we happen to be existing in. We don't have to, but it seems like a natural choice.

Cosmological redshifts, on the other hand, are a net effect of contributing factors.

Well, OK, except that the cosmological component is usually so much larger than all other factors combined, they are typically not even considered in the calculation, with minimal loss of accuracy.

I would argue that cosmological redshifts comprise of two uniquely determinable components, and here I deviate completely from the standard nomenclature.

I'll say.

there is a gravitational part and a generalized Doppler part, where the gravity part is from the unique curvature of spacetime...

Wouldn't it make more sense to call this the "spacetime curvature" part, I mean, since the term "gravitational redshift" is already taken?

...and the generalized Doppler part is the rest.... from the other aspects of the metric that are not related to curvature, and can decompose arbitrarily into motion of source, motion of observer, and motion of coordinates themselves.

Arbitrarily?

I used to think that the mathematics of doppler receding galaxies was equivalent to the mathematics of cosmologically expanding spacetime. Obviously the causes are different, but I thought the math worked out equivalently. It was pointed out to me on this board that this was not the case. Doppler does not decompose to cosmological, and I do wish people would stop referring to cosmological redshift as a doppler shift, which it is not.

Why on Earth gravitational redshifts are reserved for local effects is beyond me...

Perhaps because overall spacetime curvature has traditionally been considered and is so close to being flat, and hence not effecting a redshift?

[Ken G]

s relativity just "a theoretical concept", or is it an accurate explanation about how Nature works? A natural law, maybe?

I wouldn't try to draw the distinction, it seems pretty philosophical, except to say that it pays to keep a clear idea of the difference between laws of nature and nature itself (ask yourself, which law of nature is exact, and if none are, then how does nature figure out exactly what to do?).

I swear you've got "coordinate systems" on the brain.

Indeed I do, but it's because I continue to see people confusing "nature" with coordinatizations of nature. Here's an example just above:

I am inclined to answer in the affirmative, that a cosmological redshift is an intrinsic property of the photon. My reason is that this cosmological redshift in GR is an effect not of the relative motion between observer & source, but rather an effect of the spacetime that extends between observer & source.

I think what Tim really means by the word "intrinsic" is something that happened to the coordinates that is an invariant property that would have to happen to any coordinate chart that connected the observer to the source in such a way as to be locally inertial frames, independent of the motion of source and observer. But it's still not intrinsic to the photon, it's intrinsic to this approach to making coordinate charts. And you can't equate intrinsic = invariant without also including Doppler shifts, because even a Doppler shift is invariant to the coordinatization.

Well, OK, except that the cosmological component is usually so much larger than all other factors combined, they are typically not even considered in the calculation, with minimal loss of accuracy.

No I don't think that's true, because even a universe with little or no gravity could still have cosmological redshifts. I would say that the generalized Doppler component traces back to the initial expanding state, i.e. the tradeoff between spatial curvature and temporal curvature in the comoving frame, whereas the gravitational component is due to spacetime curvature. So not all the cosmological redshift is gravitational, by this meaning of the term, and both net contributions are coordinate invariant.

Wouldn't it make more sense to call this the "spacetime curvature" part, I mean, since the term "gravitational redshift" is already taken?

Not if you think it has been taken by a poor pedagogical stance. Gravitational redshift should be all the redshift that comes from the presence of gravity, I would hate to cede that phrase to the camp that maintains it can only be caused by local gravitational perturbations. But then, I'm not an expert, perhaps if I had a deeper understanding I'd see it. Still, there are many things about which I do have a deeper understanding where I see poor choice of terms, and I try to help replace them.

Doppler does not decompose to cosmological, and I do wish people would stop referring to cosmological redshift as a doppler shift, which it is not.

You misunderstand me here. I did not say Doppler decomposed to cosmological, I said cosmological decomposed in part to gravitational, and in part to Doppler, which in turn decomposed into motion of source, of observer, and of coordinates themselves. It is this final decomposition that I said was arbitrary, because it is-- it is not coordinate invariant. After all, there are only two possibilities: arbitrary, and invariant. Net effects are invariant, but contributing effects are often arbitrary. I am trying to reduce the amount of arbitrariness by choosing better meanings for the words, this is my entire goal. The underlying truth to all this is that comoving spatial coordinates just make a heck of a lot of sense, so they kinda don't seem arbitrary, but they still are, and I refuse to toe the line and pretend otherwise.

Perhaps because overall spacetime curvature has traditionally been considered and is so close to being flat, and hence not effecting a redshift?

No, it is the overall spatial curvature that is flat, if you use comoving coordinates. Spacetime curvature is highly nonflat, and that is a large part of the cosmological redshift. You see why I have coordinates on the brain-- it is hard to keep them straight!

[iantresman]

Well, OK, except that the cosmological component is usually so much larger than all other factors combined, they are typically not even considered in the calculation, with minimal loss of accuracy.

So when one measures such a redshift, how does one ascribe such a proportion to Doppler redshift or Cosmological redshift? Are there independent measurements that corroborate one or the other?

Regards,
Ian Tresman

[Tim Thompson]

So when one measures such a redshift, how does one ascribe such a proportion to Doppler redshift or Cosmological redshift? Are there independent measurements that corroborate one or the other?

No, the proportion of Doppler or Cosmological redshift is entirely model dependent.

But isn't the relativistic redshift also known as the gravitational redshift, which might be exhibited by a stationary black hole?

I see the Wiki folks put it all in the "relativistic" redshift category. I don't like that way of doing it. The "special relativistic" redshift is a Doppler shift, and nothing more. But it is crafted in terms of special relativity rather than Newtonian physics. The gravitational redshift is an entirely different beast, entirely general (not "special") relativistic. It also has nothing to do with relative motion between observer & source, and everything to do with the way gravity distorts spacetime.

[iantresman]

I see the Wiki folks put it all in the "relativistic" redshift category. I don't like that way of doing it. The "special relativistic" redshift is a Doppler shift, and nothing more.

The Wiki Redshift has a section on Redshift mechanism. The last one it calls "Relativistic effects"; is this the same as that under the Doppler effect which it mentions as "relativistic Doppler effect"?

So to clarify, if they are both Doppler effects, the only difference is that one is due to approaching the speed of light, in which case relativity is taken into account.

And I presume that in the Doppler effect, relativistic and non-relativistic, once a photon has left the source, it will not be subject to any other redshift, possibly until it reaches a moving observer.

And in which case, for a photon moving in a vacuum, between the source and observer, only the gravitational redshift might affect the photon?

Regards,
Ian Tresman

[Ken G]

And I presume that in the Doppler effect, relativistic and non-relativistic, once a photon has left the source, it will not be subject to any other redshift, possibly until it reaches a moving observer.

Once you choose a coordinate system, what I would call the Doppler effect involves the motion of the source relative to the coordinates, the motion of the spatial coordinates, and the motion of the observer relative to the coordinates. The answer comes out coordinate independent, so it doesn't matter that you chose a coordinate system, but you still have to if you want to say when and why the shifts occurred (the words depend on the coordinates). The gravitational piece, I would say, has to do not with the motion of the coordinates but the spacetime curvature itself, which is independent of the spatial coordinates or their motion in time. So I think any redshift is comprised of two pieces, Doppler and gravitational, and that decomposition is coordinate invariant so is "real". But I've never seen it described this way, and I can't say for sure that I am right. If I am, the current nomenclature is seriously jumbled.

[Tim Thompson]

The Wiki Redshift has a section on Redshift mechanism. The last one it calls "Relativistic effects"; is this the same as that under the Doppler effect which it mentions as "relativistic Doppler effect"?

Yes & No. That section, "Relativistic effects" includes two distinctly different and unrelated redshift mechanisms. One refers to time dilation in special relativity, and that is the same as the relativistic Doppler effect. The other is the gravitational redshift.

So to clarify, if they are both Doppler effects, the only difference is that one is due to approaching the speed of light, in which case relativity is taken into account.

What you have described is the difference between the relativistic & non-relativistic (i.e, Newtonian) Doppler shift.

And I presume that in the Doppler effect, relativistic and non-relativistic, once a photon has left the source, it will not be subject to any other redshift, possibly until it reaches a moving observer.

The Doppler effect deals with that part of the redshift which is due only to the relative motion between observer & source. Whether or not there is any other effect depends on what happens to the photon as it flies along. If you use only the Doppler formula, then you are assuming that there is no other, non-Doppler effect. If you have rfeason to believe that there is some other, non-Doppler effect, then you don't use the Doppler formula by itself.

And in which case, for a photon moving in a vacuum, between the source and observer, only the gravitational redshift might affect the photon?

The case where there is no relative motion between the observer & source.

[Ken G]

My problem is I'm not sure what one means by "no relative motion between observer and source". This cannot mean that the distance between them is not changing, because in cosmology the distance is increasing, but it is traditionally said that there is no relative motion so it's not a Doppler shift. I don't see how this is not just an arbitrary distinction. I think the terminology is intimately intertwined with the choice of coordinates, and little effort is made to sensibly extract it. To talk about a Doppler shift, you probably have to place the observer and source into locally inertial frames, and then connect those frames little by little as you move along, always analytically continuing to the next frame that would make the most sense from the inertial perspective, like painting a N-S-E-W grid on the surface of the Earth. If you started in an inertial frame moving with the source, and ended with an inertial frame that moved with the observer, then perhaps you would say there is no Doppler shift and all the redshift is gravitational. But in cosmology, there is no effort to use this coordinate system. Instead, comoving frame coordinates are used, which automatically have no relative movement of source and observer with respect to the coordinates (ignoring small perturbative effects). So you've loaded the dice, as it were, to give no Doppler shift, but you didn't connect the frames in the inertial way, rather you made them always move with the local matter. What this does is brings in the history of that matter in an implicit way, so it is something more than just what gravity is doing. We may have to satisfy ourselves that the names we give to various shifts will always depend on coordinates, but I'm wondering if there is not a way to choose our names such that the gravitational shift is always invariant, and the generalized Doppler shift is always invariant. I suggested a way that might work, but I just don't know for sure. Without such an effort, what is meant by "no relative motion"?

[iantresman]

So how does this look a a summary:

Doppler redshift: Redshift due to the relative motion of the source and observor. A photon in free space, is not subjected to further Doppler effects.

Cosmological redshift: Redshift due to the expansion of space, which can occur only in free space. It appear similar to the Doppler redshift since the source appears to be moving relative to the observer. Does it take into account whether the expansion of space is linear?

Gravitational redshift: Redshift due to time dliation effects, which may occur to a Doppler redshift (where the source of observer is approaching light speed), OR, due to any photo in free space, travelling though a gravitational well such that near a black hole or neutron star(?)

Regards,
Ian Tresman

[Jeff Root]

I'm going to comment on the first, simplest definition, and
ignore the other two.

Doppler redshift: Redshift due to the relative motion of the
source and observor.

The word 'motion' here is ambiguous. For Doppler effect, it
definitely means 'change in position' as opposed to 'speed'.
That is an important distinction. Unfortunately the word
'motion' is very handy but very easily misinterpreted.

I know this is only a draft, but I'll point out that the word
'observer' is misspelled.

A photon in free space, is not subjected to further Doppler effects.

That is vague and misleading. It also seems pointless, since
it doesn't say anything that the first sentence doesn't say.
You need to clarify exactly what the Doppler effect is/how it
works, but the second sentence doesn't do that.

Here is a GIF animation I made a few days ago which compares
Doppler effects in four different situations. It is intended
to show that Doppler effect is the result of changing distance
between the source and the observer, not their relative speed:

http://www.freemars.org/jeff2/Doppler3.htm

The animation shows Doppler effect on bullets. Of course, the
effect on light is different.

-- Jeff, in Minneapolis

[iantresman]

http://www.freemars.org/jeff2/Doppler3.htm

The animation shows Doppler effect on bullets. Of course, the
effect on light is different.

Show off! I wish I could put an animation together as quickly. That's great. What did you use to create it?

So is the distance between bullets equivalent to wavelength, and frequency as is, pixels = metres, frames = seconds?

Why not label the illustration with wavelength, frequency, metres and seconds, and add it to the Wikipedia article on the Doppler effect.

And if the gun was a "laser gun", it could fire photons?

Regards,
Ian Tresman

[iantresman]

Anyone reading this thread may be interested in another question I have posed in another thread, on Redshift and blurring.

Regards,
Ian Tresman