Notion of Absolute Clock Synchronization in SR
in [Relativity]
|
From: Tom Roberts on 13 Mar 2010 12:59 GSS wrote: > On Mar 11, 10:07 pm, Tom Roberts <tjrob...(a)sbcglobal.net> wrote: >> But it conflicts with actual observations of the world we inhabit. Atomic clocks >> on earth and in satellites do not remain synchronized. Effective clocks >> intrinsic to elementary particles vary widely (factors of 10,000 or more in >> rate) with their speed relative to the lab. >> > Elementary particles do not 'carry' any intrinsic clocks. Sure they do, effectively. The decay of unstable particles proceeds with a constant probability per unit time, so the number remaining decreases exponentially with time. This can be measured very accurately, as a function of the particle's speed relative to the lab. At any given speed the exponential decrease is an excellent model, and the decay constant of that exponential behaves as a function of speed quite accurately in agreement with the prediction of SR. >> There is no such definition, as "absolute clock synchronization" does not exist, >> either in SR or in the world we inhabit. >> > But the definition of clock synchronization does exist. The definition of clock synchronization includes the requirement that synchronized clocks REMAIN in synch. That manifestly does not happen for clocks that are not at rest in the same locally-inertial frame [#]. Just because a definition of clock synchronization exists does not mean it is "absolute". [#] Or certain other special cases, such as clocks at rest on earth's geoid. These special cases all rely on various effects of their non-inertial motion canceling out. For instance, the geoid is such that the latitude-dependent effect of speed relative to the ECI is canceled by gravitation. > Is there no > alternative method of clock synchronization other than e- > synchronization? There are an infinite number of ways one could adjust the offsets of a pair of clocks. Most of them would not merit the name "synchronization". Except for the above special cases, all the ones that merit that name apply only when the clocks are at rest in the same locally-inertial frame, and in that frame they are all equivalent to Einstein's synchronization method. That is, if you adjust two clocks' offsets, and they then fail any of Einstein's methods, then nobody would call those clocks "synchronized". Exercise for the reader: How is this applied to clocks at rest on the geoid? Hint: there is a reason the GPS uses the ECI frame. >> It is NOT true that "there is no relative motion between the two clocks" -- that >> is valid ONLY relative to the rotating earth, which is not an inertial frame. It >> is invariably wrong to attempt to make conclusions based on non-inertial frames >> without careful considerations of the implications of their non-inertialness. >> Relative to any inertial frame, these two clocks do have different velocities. > > Do you mean to imply that the clock rates or their frequencies will no > longer match (even when both clocks are at rest on the surface of > earth or say geoid) just because both A and B appear to have different > 'velocities' in ECI or that the line segment AB is seen to be rotating > in ECI? You need to read what I wrote. For some methods of comparison of their frequencies, the answer is that they are equal, and for other methods of comparison the answer is they are different. This simple and obvious method is in the first set: send a light signal between them along a path fixed on the rotating earth and compare one clock to the signal from the other clock. > [...] all clocks on > the surface of earth, say geoid, must remain synchronized irrespective > of their rotation about the earth axis. No. You discussed COMPARING FREQUENCIES, not being synchronized. The frequencies will compare equal as long as they are all at rest on the geoid. That's why the geoid is important for planet-wide clock synchronization. But clocks at other altitudes, or clocks that are moving on the geoid, will not have frequencies equal to those at rest on the geoid (for sensible methods of comparison). THAT is my point -- there is nothing "absolute" or "universal" here. You only discussed comparing the clocks' frequencies, going in short steps around the equator. If, instead, you had discussed pairwise synchronization, you would find that when you got back to your starting point, the first and last clocks are NOT synchronized, even though each adjacent pair of clocks is. And the difference would depend on the direction you went around. Metrologists call this the "Sagnac effect" and account for it when comparing clocks at different labs. Exercise for the reader: GSS's argument could easily be modified to synchronize clock pairs rather than compare frequencies. Find the fallacy in that. Then explain why his argument works for frequency comparisons but not for synchronization. Hint: this second point is nontrivial -- remember what I said above about special cases needing effects to cancel. Tom Roberts
From: Sue... on 13 Mar 2010 17:15 On Mar 13, 12:59 pm, Tom Roberts <tjroberts...(a)sbcglobal.net> wrote: > GSS wrote: > > On Mar 11, 10:07 pm, Tom Roberts <tjrob...(a)sbcglobal.net> wrote: > >> But it conflicts with actual observations of the world we inhabit. Atomic clocks > >> on earth and in satellites do not remain synchronized. Effective clocks > >> intrinsic to elementary particles vary widely (factors of 10,000 or more in > >> rate) with their speed relative to the lab. > > > Elementary particles do not 'carry' any intrinsic clocks. > > Sure they do, effectively. The decay of unstable particles proceeds with a > constant probability per unit time, so the number remaining decreases > exponentially with time. This can be measured very accurately, as a function of > the particle's speed relative to the lab. At any given speed the exponential > decrease is an excellent model, and the decay constant of that exponential > behaves as a function of speed quite accurately in agreement with the prediction > of SR. Can you offer the name of a metrology laboratory that has exploited this "very accurate" method of marking time? Sue... http://en.wikipedia.org/wiki/Lorentz_ether_theory#Later_activity_and_Current_Status > > >> There is no such definition, as "absolute clock synchronization" does not exist, > >> either in SR or in the world we inhabit. > > > But the definition of clock synchronization does exist. > > The definition of clock synchronization includes the requirement that > synchronized clocks REMAIN in synch. That manifestly does not happen for clocks > that are not at rest in the same locally-inertial frame [#]. > > Just because a definition of clock synchronization exists does not mean it is > "absolute". > > [#] Or certain other special cases, such as clocks at rest on > earth's geoid. These special cases all rely on various effects > of their non-inertial motion canceling out. For instance, the > geoid is such that the latitude-dependent effect of speed > relative to the ECI is canceled by gravitation. > > > Is there no > > alternative method of clock synchronization other than e- > > synchronization? > > There are an infinite number of ways one could adjust the offsets of a pair of > clocks. Most of them would not merit the name "synchronization". Except for the > above special cases, all the ones that merit that name apply only when the > clocks are at rest in the same locally-inertial frame, and in that frame they > are all equivalent to Einstein's synchronization method. > > That is, if you adjust two clocks' offsets, and they then fail any of Einstein's > methods, then nobody would call those clocks "synchronized". > > Exercise for the reader: How is this applied to clocks at rest > on the geoid? Hint: there is a reason the GPS uses the ECI frame. > > >> It is NOT true that "there is no relative motion between the two clocks" -- that > >> is valid ONLY relative to the rotating earth, which is not an inertial frame. It > >> is invariably wrong to attempt to make conclusions based on non-inertial frames > >> without careful considerations of the implications of their non-inertialness. > >> Relative to any inertial frame, these two clocks do have different velocities. > > > Do you mean to imply that the clock rates or their frequencies will no > > longer match (even when both clocks are at rest on the surface of > > earth or say geoid) just because both A and B appear to have different > > 'velocities' in ECI or that the line segment AB is seen to be rotating > > in ECI? > > You need to read what I wrote. For some methods of comparison of their > frequencies, the answer is that they are equal, and for other methods of > comparison the answer is they are different. This simple and obvious method is > in the first set: send a light signal between them along a path fixed on the > rotating earth and compare one clock to the signal from the other clock. > > > [...] all clocks on > > the surface of earth, say geoid, must remain synchronized irrespective > > of their rotation about the earth axis. > > No. You discussed COMPARING FREQUENCIES, not being synchronized. The frequencies > will compare equal as long as they are all at rest on the geoid. That's why the > geoid is important for planet-wide clock synchronization. > > But clocks at other altitudes, or clocks that are moving on the geoid, will not > have frequencies equal to those at rest on the geoid (for sensible methods of > comparison). THAT is my point -- there is nothing "absolute" or "universal" here. > > You only discussed comparing the clocks' frequencies, going in > short steps around the equator. If, instead, you had discussed > pairwise synchronization, you would find that when you got back > to your starting point, the first and last clocks are NOT > synchronized, even though each adjacent pair of clocks is.. And > the difference would depend on the direction you went around. > Metrologists call this the "Sagnac effect" and account for it > when comparing clocks at different labs. > > Exercise for the reader: GSS's argument could easily be modified > to synchronize clock pairs rather than compare frequencies. Find > the fallacy in that. Then explain why his argument works for > frequency comparisons but not for synchronization. Hint: this > second point is nontrivial -- remember what I said above about > special cases needing effects to cancel. > > Tom Roberts
From: BURT on 13 Mar 2010 18:53 On Mar 13, 2:15 pm, "Sue..." <suzysewns...(a)yahoo.com.au> wrote: > On Mar 13, 12:59 pm, Tom Roberts <tjroberts...(a)sbcglobal.net> wrote: > > > > > > > GSS wrote: > > > On Mar 11, 10:07 pm, Tom Roberts <tjrob...(a)sbcglobal.net> wrote: > > >> But it conflicts with actual observations of the world we inhabit. Atomic clocks > > >> on earth and in satellites do not remain synchronized. Effective clocks > > >> intrinsic to elementary particles vary widely (factors of 10,000 or more in > > >> rate) with their speed relative to the lab. > > > > Elementary particles do not 'carry' any intrinsic clocks. > > > Sure they do, effectively. The decay of unstable particles proceeds with a > > constant probability per unit time, so the number remaining decreases > > exponentially with time. This can be measured very accurately, as a function of > > the particle's speed relative to the lab. At any given speed the exponential > > decrease is an excellent model, and the decay constant of that exponential > > behaves as a function of speed quite accurately in agreement with the prediction > > of SR. > > Can you offer the name of a metrology laboratory that has > exploited this "very accurate" method of marking time? > > Sue... > > http://en.wikipedia.org/wiki/Lorentz_ether_theory#Later_activity_and_... > > > > > > > >> There is no such definition, as "absolute clock synchronization" does not exist, > > >> either in SR or in the world we inhabit. > > > > But the definition of clock synchronization does exist. > > > The definition of clock synchronization includes the requirement that > > synchronized clocks REMAIN in synch. That manifestly does not happen for clocks > > that are not at rest in the same locally-inertial frame [#]. > > > Just because a definition of clock synchronization exists does not mean it is > > "absolute". > > > [#] Or certain other special cases, such as clocks at rest on > > earth's geoid. These special cases all rely on various effects > > of their non-inertial motion canceling out. For instance, the > > geoid is such that the latitude-dependent effect of speed > > relative to the ECI is canceled by gravitation. > > > > Is there no > > > alternative method of clock synchronization other than e- > > > synchronization? > > > There are an infinite number of ways one could adjust the offsets of a pair of > > clocks. Most of them would not merit the name "synchronization". Except for the > > above special cases, all the ones that merit that name apply only when the > > clocks are at rest in the same locally-inertial frame, and in that frame they > > are all equivalent to Einstein's synchronization method. > > > That is, if you adjust two clocks' offsets, and they then fail any of Einstein's > > methods, then nobody would call those clocks "synchronized". > > > Exercise for the reader: How is this applied to clocks at rest > > on the geoid? Hint: there is a reason the GPS uses the ECI frame. > > > >> It is NOT true that "there is no relative motion between the two clocks" -- that > > >> is valid ONLY relative to the rotating earth, which is not an inertial frame. It > > >> is invariably wrong to attempt to make conclusions based on non-inertial frames > > >> without careful considerations of the implications of their non-inertialness. > > >> Relative to any inertial frame, these two clocks do have different velocities. > > > > Do you mean to imply that the clock rates or their frequencies will no > > > longer match (even when both clocks are at rest on the surface of > > > earth or say geoid) just because both A and B appear to have different > > > 'velocities' in ECI or that the line segment AB is seen to be rotating > > > in ECI? > > > You need to read what I wrote. For some methods of comparison of their > > frequencies, the answer is that they are equal, and for other methods of > > comparison the answer is they are different. This simple and obvious method is > > in the first set: send a light signal between them along a path fixed on the > > rotating earth and compare one clock to the signal from the other clock.. > > > > [...] all clocks on > > > the surface of earth, say geoid, must remain synchronized irrespective > > > of their rotation about the earth axis. > > > No. You discussed COMPARING FREQUENCIES, not being synchronized. The frequencies > > will compare equal as long as they are all at rest on the geoid. That's why the > > geoid is important for planet-wide clock synchronization. > > > But clocks at other altitudes, or clocks that are moving on the geoid, will not > > have frequencies equal to those at rest on the geoid (for sensible methods of > > comparison). THAT is my point -- there is nothing "absolute" or "universal" here. > > > You only discussed comparing the clocks' frequencies, going in > > short steps around the equator. If, instead, you had discussed > > pairwise synchronization, you would find that when you got back > > to your starting point, the first and last clocks are NOT > > synchronized, even though each adjacent pair of clocks is. And > > the difference would depend on the direction you went around. > > Metrologists call this the "Sagnac effect" and account for it > > when comparing clocks at different labs. > > > Exercise for the reader: GSS's argument could easily be modified > > to synchronize clock pairs rather than compare frequencies. Find > > the fallacy in that. Then explain why his argument works for > > frequency comparisons but not for synchronization. Hint: this > > second point is nontrivial -- remember what I said above about > > special cases needing effects to cancel. > > > Tom Roberts- Hide quoted text - > > - Show quoted text -- Hide quoted text - > > - Show quoted text - Aether flows over energy in the direction of energy's propagation through space. Speeding up energy slows the aether flow by one rate out of two. The flow is always in the same direction as the energy is moving into. Mitch Raemsch; one flow direction over energy and field is the simplest concept
From: BURT on 13 Mar 2010 19:03 On Mar 11, 7:35 am, GSS <gurcharn_san...(a)yahoo.com> wrote: > As per Newtonian notion of absolute space and time, clocks can be > synchronized in absolute terms such that identical precision atomic > clocks located anywhere within the solar system and in any state of > motion, will read the same time t1 when a standard master clock reads > t1. This notion of absolute clock synchronization implies the notion > of absolute simultaneity. > > However, as per SR, spatial distance and time measurements have been > rendered 'relative' and cannot be the same value for different > observers in different states of motion. As per SR the notion of > global 'absolute simultaneity' is fundamentally invalid for different > observers in different states of motion. Therefore, the notion of > global 'absolute clock synchronization' (in contrast to e- > synchronization) is no longer valid in SR. > > Since the term 'absolute clock synchronization' is often used in > discussions, I would like to request some Relativity experts to kindly > clarify the precise definition of absolute clock synchronization in > SR. Kindly illustrate the procedure, through some 'thought experiment' > or 'gedanken', to achieve absolute clock synchronization for all > observers in different states of motion within our solar system. > > Further, I also need some expert opinion on the following situation, > involving clock synchronization. > > Two identical precision atomic clocks are positioned side by side at > point A on the surface of earth and mutually synchronized to ensure > that > (a) their clock rates or frequencies are exactly matched or > synchronized > (b) their instantaneous timing offsets are eliminated to ensure that a > common trigger pulse yields the same timing reading t1 from both > clocks. > > Assuming the inherent drift of the two atomic clocks is identical and > well within 100 ps per day, it can be demonstrated that while the two > clocks remain side by side, their synchronization, after a period of > one day, is retained at well within one ns accuracy. > > Let us shift one of the synchronized atomic clocks to a position B > such that distance AB is about 30 km. As per Newtonian notion of > absolute space and time, the mutual synchronization of the two clocks, > positioned at points A and B, will be retained in tact and this > synchronization can be referred as 'absolute synchronization'. But > according to SR, the mutual synchronization of the two clocks will > 'breakdown' during the shifting of one of the clocks from point A to > point B. Since 'after' shifting of one clock to point B on the surface > of earth, there is no relative motion between the two clocks, their > time rates or frequencies will again 'become' synchronized. Therefore, > the only persisting effect of the 'synchronization breakdown' during > shifting or repositioning of the two clocks, will be a motion induced > constant time offset, say dT, in the instantaneous readings of the two > clocks. > > My question to the learned Relativity experts is: > What is the order of magnitude of this 'relative motion induced' > timing offset dT between the two clocks? > Can it be precisely calculated in SR? Is it likely to be within a few > nanoseconds or less? > > Suppose we now shift the clock at point B to bring it back to point A, > (with an identical speed and acceleration profile), will this timing > offset dT now increase to 2.dT or reduce to zero? > > I shall be thankful to the Relativity experts for their valuable > opinions and clarifications. > > GSS Relative motion in the distance has no gamma. Only accelerated energy exepriences changes in math gamma. What has never accelerated does not change in its clock. You cannot change anothers clock by changing your own motion. Only your clock can change by your motion through space. Mitch Raemsch; energy flowing faster has a slower clock by motion rate
From: harald on 15 Mar 2010 06:20
On Mar 12, 5:13 pm, Tom Roberts <tjrob...(a)sbcglobal.net> wrote: > harald wrote: > > [clocks] can also be > > synchronized at a certain point in time (only at that time) without > > having the same frequency. > > We do not call this "synchronized". The whole point of synchronizing clocks is > so they can be used together to make related measurements of something. Two > clocks that indicate the same time only once cannot be used together. > > After all, a stopped "clock" is correct twice a day! > (In physics, we don't call such a device a "clock"; it is > merely a clockface with unmoving hands.) > > Tom Roberts That misses the point of my correction of the OP (both of which you snipped) and my last remark was in general*. The OP wrongly thought that two clocks that have the same frequency are necessarily synchronized. The OP thus missed an essential point about synchronizing clocks. * It's unclear who "we" are. For physicists it's rather standard to synchronize running clocks at the start of an experiment without assuming that they run perfectly in sync during the experiment, and this is also common language of textbooks (see for an online example http://mamacass.ucsd.edu/people/pblanco/physics2d/handout1/index.html: "we can arrange for the clocks to synchronize when O' passes O, i.e. t'=t=0"). Harald |