- #1
nutgeb
- 294
- 1
I want to discuss energy conservation and time dilation relative to the cosmological redshift. Because the elements of the redshift are coordinate specific, I’ll focus just on the FRW metric using proper distance coordinates (not comoving coordinates).
As explained in Tamara Davis’s cover article in the July issue of Scientific American, and by Bunn & Hogg in a paper she references, in FRW coordinates the cosmological redshift can be thought of as photons experiencing Doppler shifts along their trajectory. This interpretation assumes that comoving galaxies move through space, and are not dragged apart by the expansion of space itself.
In Minkowski coordinates, Doppler shift arises from a single end-to-end velocity differential between the emitter and observer. But in FRW coordinates, Doppler shift occurs in many tiny increments all along the trajectory. Relative to the emitter’s reference frame, each tiny segment of a photon’s worldline is characterized by a higher proper recession velocity than the previous segment the photon passed through. This recession velocity differential is a feature of Hubble’s Law: proper recession velocity increases in direct proportion to the proper distance from the emitter. The Hubble velocity in each segment is HD, where H is the current Hubble rate and D is the proper distance.
The photon is constrained to always maintain a local peculiar velocity of exactly c. So its proper velocity relative to the emitter starts simply as V = c, but thereafter is progressively boosted by the increasing Hubble velocities it passes through: V = c + HD (where V is proper velocity relative to the emitter). Because we are using FRW coordinates, there is no problem with the proper velocity exceeding c, just as the Hubble velocity itself can exceed c.
The Doppler shift occurs because, when light crosses into a segment of the worldline which has a faster Hubble recession velocity than the previous segment, the arrival time between each successive wave crest is slightly delayed. Relative to any given reference frame, the light waves progressively stretch out as they travel, because an earlier wave crest is always slightly farther away from the emitter than a subsequent wave crest, and therefore is constrained to have a higher proper velocity away from the emitter than the subsequent wave crest does at each instant in time. So the proper distance between any two wave crests progressively increases as a function of proper distance from the emitter (and to a lesser extent also as a function of the ever-changing Hubble rate). This is why each increment of redshift is fully realized immediately en route, rather than being realized entirely only upon arrival at the observer, even though the photon is not observed in the intermediate segments it passes through. The amount of redshift would be very different if calculated as a single end-to-end event instead of as an accumulation of progressive increments.
Energy conservation
I would like to discuss how the redshift conserves energy. The Davis article points out that the locally measured energy of a photon progressively decays as it redshifts incrementally while moving from frame to frame along its trajectory. But she concludes that, like all Doppler shifts, the apparent loss of energy merely reflects a change in perspective, that is, a change in reference frame, rather than a violation of energy conservation. The universe therefore is not 'leaking energy' as a consequence of redshift.
It is misleading to focus only on standalone local energy measurements. When the emitter and observer are in motion relative to each other, energy conservation is meaningful only if each emitter-observer pair is considered in isolation, and then the results are aggregated for the universe as a whole. Energy is strictly conserved if one takes the Doppler velocity differentials along the photon’s worldline fully into account.
Cosmological Doppler shifts can be thought of as momentum decay that is required by energy conservation. A photon’s momentum decays as it redshifts, in proportion to the decay in its frequency.
As I described, the photon gains proper velocity relative to the emitter along its trajectory. An increase in proper velocity does not come as a free lunch; for the photon to avoid gaining energy relative to the emitter's reference frame, its momentum must decay by the same ratio as its proper velocity increases. That is exactly what occurs. Proper velocity increases in proportion to the expansion of the cosmic scale factor a, while momentum decays by the inverse proportion, 1/a.
At each location along the photon's trajectory, multiplying the magnitude of the momentum p and proper velocity V together always yields a constant. I think that constant relates the photon's energy conservation back to the emitter's reference frame. Photon energy normally is calculated as pc. The photon’s proper velocity is c at the instant of emission, but thereafter as it gains proper velocity, the equation pc calculates only the locally measured energy. In order to relate the energy back to the emitter’s frame, momentum must be multiplied by proper velocity V rather than local velocity c: E = pV.
Doppler shift equation and time dilation
Mathematically, the incremental Doppler shifts are multiplied together to calculate the total accumulated redshift. The equation for this is the multiplicative series (1+Ht1dDt1)(1+Ht2dDt2)...(1+Ht0dDt0). Here dD is the change in proper distance from the emitter that occurs within each tiny segment of the worldline, and t0 is the time of observation. So HtdD simply calculates the Hubble recession velocity differential, and therefore the amount of Doppler shift, that the photon experiences when it crosses one segment.
Note that for each segment, dD = dt if units of c = 1 are used for time and distance. Htdt calculates how much the scale factor of the universe has expanded during the time duration of one segment of the photon’s travel. These also are aggregated together multiplicatively to calculate the total expansion. The fact that Htdt = HtdD demonstrates unequivocally that the kinematic Doppler solution is exactly equal to the traditional cosmological redshift formula whereby the redshift is equal to the proportional amount by which the cosmic scale factor has expanded during the photon’s travel. Thus, neither observations nor math provide a basis for preferring either the kinematic paradigm or the expanding space paradigm for the expansion of the universe.
It is very interesting that the cosmological Doppler shift formula in FRW coordinates is in the same form as the classical (non-relativistic) Doppler shift formula for a moving emitter and stationary observer: 1 + V/c. This shows that the cosmological Doppler shift, when calculated in FRW coordinates, does not include any element of SR time dilation. This is because of a characteristic unique to the FRW metric -- no time dilation occurs between comovers despite their unbounded recession velocities relative to each other. The clocks of all comovers run at the same rate and record the same elapsed time since the Big Bang.
To illustrate this point, consider the empty Milne model. In that case, SR Minkowski coordinates are converted to FRW coordinates by applying a Lorentz length expansion gamma factor 1/[tex]\gamma[/tex] (see Peacock p. 88), which offsets the SR time dilation gamma factor [tex]\gamma[/tex] for comovers (but not for peculiar velocities). Compared to the SR velocity in Minkowski coordinates, the GR velocity in FRW coordinates is increased by the factor arctanh:
[tex] V_{FRW} = \frac{1}{2} ln\left( \frac{1 + v_{sr} }{ 1 - v_{sr} } \right) [/tex]
Also consider that in FRW coordinates Hubble comoving recession velocities are always added together directly without using the SR formula for addition of velocities. The SR formula obviously wouldn’t work because Hubble velocities can exceed c.
I would like to discuss this point. You may think that a Doppler equation that does not include the effects of SR is flawed, but consider whether that logic applies specifically to comovers in FRW coordinates. You may prefer to think of this equation as being equivalent to the SR Doppler shift, but with the time dilation element eliminated due to the use of FRW coordinates. The element of SR time dilation is the only difference between the classical and SR Doppler shifts.
I should emphasize that as the segment lengths are reduced toward zero, the SR time dilation element approaches zero anyway. So as a practical matter the difference between including and omitting SR time dilation from the equation can be made vanishingly small. However it is an interesting theoretical question.
And by way of comparison, if Schwarzschild coordinates are used instead of FRW, then the redshift equation includes elements of both SR and gravitational time dilation, neither of which appears directly in the FRW version of the redshift equation.
As explained in Tamara Davis’s cover article in the July issue of Scientific American, and by Bunn & Hogg in a paper she references, in FRW coordinates the cosmological redshift can be thought of as photons experiencing Doppler shifts along their trajectory. This interpretation assumes that comoving galaxies move through space, and are not dragged apart by the expansion of space itself.
In Minkowski coordinates, Doppler shift arises from a single end-to-end velocity differential between the emitter and observer. But in FRW coordinates, Doppler shift occurs in many tiny increments all along the trajectory. Relative to the emitter’s reference frame, each tiny segment of a photon’s worldline is characterized by a higher proper recession velocity than the previous segment the photon passed through. This recession velocity differential is a feature of Hubble’s Law: proper recession velocity increases in direct proportion to the proper distance from the emitter. The Hubble velocity in each segment is HD, where H is the current Hubble rate and D is the proper distance.
The photon is constrained to always maintain a local peculiar velocity of exactly c. So its proper velocity relative to the emitter starts simply as V = c, but thereafter is progressively boosted by the increasing Hubble velocities it passes through: V = c + HD (where V is proper velocity relative to the emitter). Because we are using FRW coordinates, there is no problem with the proper velocity exceeding c, just as the Hubble velocity itself can exceed c.
The Doppler shift occurs because, when light crosses into a segment of the worldline which has a faster Hubble recession velocity than the previous segment, the arrival time between each successive wave crest is slightly delayed. Relative to any given reference frame, the light waves progressively stretch out as they travel, because an earlier wave crest is always slightly farther away from the emitter than a subsequent wave crest, and therefore is constrained to have a higher proper velocity away from the emitter than the subsequent wave crest does at each instant in time. So the proper distance between any two wave crests progressively increases as a function of proper distance from the emitter (and to a lesser extent also as a function of the ever-changing Hubble rate). This is why each increment of redshift is fully realized immediately en route, rather than being realized entirely only upon arrival at the observer, even though the photon is not observed in the intermediate segments it passes through. The amount of redshift would be very different if calculated as a single end-to-end event instead of as an accumulation of progressive increments.
Energy conservation
I would like to discuss how the redshift conserves energy. The Davis article points out that the locally measured energy of a photon progressively decays as it redshifts incrementally while moving from frame to frame along its trajectory. But she concludes that, like all Doppler shifts, the apparent loss of energy merely reflects a change in perspective, that is, a change in reference frame, rather than a violation of energy conservation. The universe therefore is not 'leaking energy' as a consequence of redshift.
It is misleading to focus only on standalone local energy measurements. When the emitter and observer are in motion relative to each other, energy conservation is meaningful only if each emitter-observer pair is considered in isolation, and then the results are aggregated for the universe as a whole. Energy is strictly conserved if one takes the Doppler velocity differentials along the photon’s worldline fully into account.
Cosmological Doppler shifts can be thought of as momentum decay that is required by energy conservation. A photon’s momentum decays as it redshifts, in proportion to the decay in its frequency.
As I described, the photon gains proper velocity relative to the emitter along its trajectory. An increase in proper velocity does not come as a free lunch; for the photon to avoid gaining energy relative to the emitter's reference frame, its momentum must decay by the same ratio as its proper velocity increases. That is exactly what occurs. Proper velocity increases in proportion to the expansion of the cosmic scale factor a, while momentum decays by the inverse proportion, 1/a.
At each location along the photon's trajectory, multiplying the magnitude of the momentum p and proper velocity V together always yields a constant. I think that constant relates the photon's energy conservation back to the emitter's reference frame. Photon energy normally is calculated as pc. The photon’s proper velocity is c at the instant of emission, but thereafter as it gains proper velocity, the equation pc calculates only the locally measured energy. In order to relate the energy back to the emitter’s frame, momentum must be multiplied by proper velocity V rather than local velocity c: E = pV.
Doppler shift equation and time dilation
Mathematically, the incremental Doppler shifts are multiplied together to calculate the total accumulated redshift. The equation for this is the multiplicative series (1+Ht1dDt1)(1+Ht2dDt2)...(1+Ht0dDt0). Here dD is the change in proper distance from the emitter that occurs within each tiny segment of the worldline, and t0 is the time of observation. So HtdD simply calculates the Hubble recession velocity differential, and therefore the amount of Doppler shift, that the photon experiences when it crosses one segment.
Note that for each segment, dD = dt if units of c = 1 are used for time and distance. Htdt calculates how much the scale factor of the universe has expanded during the time duration of one segment of the photon’s travel. These also are aggregated together multiplicatively to calculate the total expansion. The fact that Htdt = HtdD demonstrates unequivocally that the kinematic Doppler solution is exactly equal to the traditional cosmological redshift formula whereby the redshift is equal to the proportional amount by which the cosmic scale factor has expanded during the photon’s travel. Thus, neither observations nor math provide a basis for preferring either the kinematic paradigm or the expanding space paradigm for the expansion of the universe.
It is very interesting that the cosmological Doppler shift formula in FRW coordinates is in the same form as the classical (non-relativistic) Doppler shift formula for a moving emitter and stationary observer: 1 + V/c. This shows that the cosmological Doppler shift, when calculated in FRW coordinates, does not include any element of SR time dilation. This is because of a characteristic unique to the FRW metric -- no time dilation occurs between comovers despite their unbounded recession velocities relative to each other. The clocks of all comovers run at the same rate and record the same elapsed time since the Big Bang.
To illustrate this point, consider the empty Milne model. In that case, SR Minkowski coordinates are converted to FRW coordinates by applying a Lorentz length expansion gamma factor 1/[tex]\gamma[/tex] (see Peacock p. 88), which offsets the SR time dilation gamma factor [tex]\gamma[/tex] for comovers (but not for peculiar velocities). Compared to the SR velocity in Minkowski coordinates, the GR velocity in FRW coordinates is increased by the factor arctanh:
[tex] V_{FRW} = \frac{1}{2} ln\left( \frac{1 + v_{sr} }{ 1 - v_{sr} } \right) [/tex]
Also consider that in FRW coordinates Hubble comoving recession velocities are always added together directly without using the SR formula for addition of velocities. The SR formula obviously wouldn’t work because Hubble velocities can exceed c.
I would like to discuss this point. You may think that a Doppler equation that does not include the effects of SR is flawed, but consider whether that logic applies specifically to comovers in FRW coordinates. You may prefer to think of this equation as being equivalent to the SR Doppler shift, but with the time dilation element eliminated due to the use of FRW coordinates. The element of SR time dilation is the only difference between the classical and SR Doppler shifts.
I should emphasize that as the segment lengths are reduced toward zero, the SR time dilation element approaches zero anyway. So as a practical matter the difference between including and omitting SR time dilation from the equation can be made vanishingly small. However it is an interesting theoretical question.
And by way of comparison, if Schwarzschild coordinates are used instead of FRW, then the redshift equation includes elements of both SR and gravitational time dilation, neither of which appears directly in the FRW version of the redshift equation.