Room-temperature superconductivity?

[rcain]

Sorry Tom, my typo. I did mean usec.

I am curious about decay rates. I did some searching, but still nothing definitive on SR effects on isotope decay rates.

i'm supposing there's got to be some direct stochastic relationship to the existing atomic clock errors in there somewhere.

[ladajo]

Are you certain that the SR correction hasn't, by this time, random-walked away from its "factory offset" value?

Look at the orders of magnitude.
What ever drives the need for the 7usec correction, it is consistent. It is calculated as an SR effect, and for different orbits (speed) it holds. Again, we may be calling it SR, and be wrong, but it is "something" that is related to velocity, or at least appears to be.

[johanfprins]

I will quickly jump in at this point but this is not a return to post every day since I still have too much to do.

Quick question for Johan...

If there is no difference in time rate, barring delay in signal from orbit to earth, why do GPS satellites need to resync periodically?

Firstly there is a difference owing to gravity. The clocks in orbit actually keep time faster within their own reference frames than a clock on earth. This has to be corrected.

The SR effect is NOT an actual slowing down of the clock in orbit within its own inertial reference frame, but a slowing down relative to the reference frame on earth: And since these satellites are used to determine positions on eath, the clocks on the sattelite must be accordingly corrected as if they are actually slowing down within their own reference frames.

Are you certain that the SR correction hasn't, by this time, random-walked away from its "factory offset" value?

Look at the orders of magnitude.
What ever drives the need for the 7usec correction, it is consistent. It is calculated as an SR effect, and for different orbits (speed) it holds. Again, we may be calling it SR, and be wrong, but it is "something" that is related to velocity, or at least appears to be.

If DeltaV's random walk argument is correct then our whole discussion becomes irrelevant. But if a correction is required for gravity, then a correction is also required owing to SR. But this does not mean that it is required because the clock within the sattelite is actually slowing down within its own reference frame. It is required since the clock is going slower within the reference frame on earth where the distances are calculated. If you do not correct for both gravity and SR the distances will be incorrect.

But this does NOT mean that a clock that had been accelerated to move with a speed v and then decelerated to rejoin another clock that stayed behind, will show a different timer owing to SR. If this does happen, it violates the first postulate of Einstein. This would then prove that SR is wrong: I do not believe that the latter is the case and therefore, I believe that the experiments of flying clocks around the world in aeroplanes cannot be correct.

[DeltaV]

There's an SR effect, alright. An effect apparent to ground observers (ignoring random walk/noise/bias drift/jumps/updates/etc). This does not prove that SR is actually changing the aging rate onboard the satellite. In contrast with the GR effect, which is actually speeding up the satellite clocks by about 45 ns/day.

So the +45ns GR part of the offset is compensating for an actual change in aging and the -7ns SR part of the offset is compensating for an apparent change in aging. The fact that a combined offset is used does not mean that both parts are compensating for actual aging rates onboard the satellites.

I'm guessing the offset is implemented via a constant value in computer memory that determines how many atomic oscillations correspond to one clock "tick", using a digital counter.

Starting with bird 2, the early Navstar/GPS birds with Cesium clocks had the offset loaded at the factory. The newer birds, with jumpier Rubidium clocks, have a variable offset that is uploaded with each asynchronous update, referenced to the NIST/USNO ground clocks.

[johanfprins]

There's an SR effect, alright. An effect apparent to ground observers (ignoring random walk/noise/bias drift/jumps/updates/etc). This does not prove that SR is actually changing the aging rate onboard the satellite.

AMEN!

[ladajo]

So the +45ns GR part of the offset is compensating for an actual change in aging and the -7ns SR part of the offset is compensating for an apparent change in aging. The fact that a combined offset is used does not mean that both parts are compensating for actual aging rates onboard the satellites.

The offsets are to account for the difference in the rate of oscillations of the atoms. So if the GR component is +45us, and the SR component is -7us, arrving at an overall +38us, this amounts to a clear difference in the oscillation counts for the clock in orbit as compared to the clock on the ground counting out its 1 second worth of oscillations (Cs-133 international standard of 9,192,631,770). How can you say that one effect is real and the other apparent. The counts change or they don't.
The reason that the GR and SR components are split, is because in different orbits (GR - gravity/altitude, SR- velocity), the values both change independantly, and thus arrive at a different combined value. It is possible for either term to dominate the adjustment based on the height and speed of the vehicle carrying the clock.

As I stated above, I also wonder about isotope decay rates. If atomic osciallations demonstrabley change rate, so should isotope decay rates if SR is a real effect. We should be able to look at data or run an experiment on SR based on isotope decay rates if it is real. Available data and studies seem to indicate that isotope decay rates do not change, and that seems to counter SR. However, my digging has shown me so far that we really have not taken a hard look at this aspect yet.

I again think that Johan's idea of a velocity only based experiment, tied with a decay rate test would be informative.

[tomclarke]

So the +45ns GR part of the offset is compensating for an actual change in aging and the -7ns SR part of the offset is compensating for an apparent change in aging. The fact that a combined offset is used does not mean that both parts are compensating for actual aging rates onboard the satellites.

The offsets are to account for the difference in the rate of oscillations of the atoms. So if the GR component is +45us, and the SR component is -7us, arrving at an overall +38us, this amounts to a clear difference in the oscillation counts for the clock in orbit as compared to the clock on the ground counting out its 1 second worth of oscillations (Cs-133 international standard of 9,192,631,770). How can you say that one effect is real and the other apparent. The counts change or they don't.
The reason that the GR and SR components are split, is because in different orbits (GR - gravity/altitude, SR- velocity), the values both change independantly, and thus arrive at a different combined value. It is possible for either term to dominate the adjustment based on the height and speed of the vehicle carrying the clock.

As I stated above, I also wonder about isotope decay rates. If atomic osciallations demonstrabley change rate, so should isotope decay rates if SR is a real effect. We should be able to look at data or run an experiment on SR based on isotope decay rates if it is real. Available data and studies seem to indicate that isotope decay rates do not change, and that seems to counter SR. However, my digging has shown me so far that we really have not taken a hard look at this aspect yet.

I again think that Johan's idea of a velocity only based experiment, tied with a decay rate test would be informative.

You need v high velocity to do isotope decay, since it is difficult to measure rates very accurately.

[tomclarke]

There's an SR effect, alright. An effect apparent to ground observers (ignoring random walk/noise/bias drift/jumps/updates/etc). This does not prove that SR is actually changing the aging rate onboard the satellite. In contrast with the GR effect, which is actually speeding up the satellite clocks by about 45 ns/day.

So the +45ns GR part of the offset is compensating for an actual change in aging and the -7ns SR part of the offset is compensating for an apparent change in aging. The fact that a combined offset is used does not mean that both parts are compensating for actual aging rates onboard the satellites.

I'm guessing the offset is implemented via a constant value in computer memory that determines how many atomic oscillations correspond to one clock "tick", using a digital counter.

Starting with bird 2, the early Navstar/GPS birds with Cesium clocks had the offset loaded at the factory. The newer birds, with jumpier Rubidium clocks, have a variable offset that is uploaded with each asynchronous update, referenced to the NIST/USNO ground clocks.

A difference in rate, for a clock which never changes its position (at given point in orbit) is not observational. It is actual.

You can synchronise the satellite & ground clocks diectly and find over long time that one has more ticks than the other.

BTW your statement that SR effect is observational whereas GR is actual has no sense. Eitehr both are observational, or both are real, since they are measured in identical fashion.

[ladajo]

You need v high velocity to do isotope decay, since it is difficult to measure rates very accurately.

Or a long time.

If we had thought about it, it would have been easy, cheap, and probably no measurable impact to mission to add a small package on any of the deep space RTG vehicles to measure this. But to date, we have not done so as far as I can tell.

Oh well.

[DeltaV]

A difference in rate, for a clock which never changes its position (at given point in orbit) is not observational. It is actual.

If by "never changes its position" you mean a clock on a very tall tower (orbital altitude), experiencing an actual rate change due to gravitational blueshift, I have always agreed with this.
It would be the same actual rate change (ignoring higher order GR effects) as for a clock in a circular orbit at that altitude.

You can synchronise the satellite & ground clocks diectly and find over long time that one has more ticks than the other.

No one is disputing this. If the clocks were both brought to relative rest at the same location, the actual GR-caused difference would be observed and the apparent SR-caused difference would have evaporated.

BTW your statement that SR effect is observational whereas GR is actual has no sense. Eitehr both are observational, or both are real, since they are measured in identical fashion.

Your statement makes no sense, since there are no separate measures, to an observer on the ground, of the GR and SR effects on satellite clocks. There is only one type of measure performed, not "identical fashion" measures.
GR (actually) speeds up the satellite clock by 45 ns/day, hence the +45ns part of the "factory offset" (a digital counter scale change, mapping atomic oscillations to output "ticks").
SR (apparently) slows down the satellite clock, to a ground observer, hence the -7ns part of the "factory offset".
How does any of this mandate that either both must be observational or both must be real? It doesn't.

[DeltaV]

How can you say that one effect is real and the other apparent. The counts change or they don't.

Of course the counts change.
A ground observer sees a combination of 3 things: actual change due to GR, apparent change due to SR, and the synthetic correction of the "factory offset".
See my post above. Why is this so hard for you to understand?

The reason that the GR and SR components are split, is because in different orbits (GR - gravity/altitude, SR- velocity), the values both change independantly, and thus arrive at a different combined value. It is possible for either term to dominate the adjustment based on the height and speed of the vehicle carrying the clock.

I have never had a quarrel with this.

[johanfprins]

How does any of this mandate that either both must be observational or both must be real? It doesn't.

Bravo! Bravo!

Both are real within the reference frame on earth. GR is also real within the sattelite's reference frame. SR is only real within the earth's reference frame (as the formula for time dilation CLEARLY states) NOT within the satellite's own reference frame.

Since the satellites are used to calculate positions within the earth's reference frame, corections are required for both GR and SR; even though SR does NOT cause the clock on the sattelite to actually slow down relative to the reference frame of the sattelite itself.

[tomclarke]

A difference in rate, for a clock which never changes its position (at given point in orbit) is not observational. It is actual.

If by "never changes its position" you mean a clock on a very tall tower (orbital altitude), experiencing an actual rate change due to gravitational blueshift, I have always agreed with this.
It would be the same actual rate change (ignoring higher order GR effects) as for a clock in a circular orbit at that altitude.

You can synchronise the satellite & ground clocks diectly and find over long time that one has more ticks than the other.

No one is disputing this. If the clocks were both brought to relative rest at the same location, the actual GR-caused difference would be observed and the apparent SR-caused difference would have evaporated.

BTW your statement that SR effect is observational whereas GR is actual has no sense. Eitehr both are observational, or both are real, since they are measured in identical fashion.

Your statement makes no sense, since there are no separate measures, to an observer on the ground, of the GR and SR effects on satellite clocks. There is only one type of measure performed, not "identical fashion" measures.
GR (actually) speeds up the satellite clock by 45 ns/day, hence the +45ns part of the "factory offset" (a digital counter scale change, mapping atomic oscillations to output "ticks").
SR (apparently) slows down the satellite clock, to a ground observer, hence the -7ns part of the "factory offset".
How does any of this mandate that either both must be observational or both must be real? It doesn't.

"Observational" means that the signals coming from the moving satellite are so transformed (because of the movement) that there is apparent time dilation.

But there is no transformation between signals used to determine number of ticks:

day1, when satellite is exactly overhead (say) receve count 1.
After one orbit, when satellite is in identical position, receive count 2.

However long signals take, the velocity & position of satellite are same so the time difference is the same and we can exactly really compare satellite and ground counts.

Of course, as I've said above, but you have not understood, where the satellite stays in roughly same position any observational correction is bounded by light time to object, whereas the real change is unbounded and the time difference gets larger as time continues.

What is so difficult about that?

[tomclarke]

How does any of this mandate that either both must be observational or both must be real? It doesn't.

Bravo! Btavo!

Both are real within the reference frame on earth. GR is also real within the sattelite's reference frame. SR is only real within the earth's reference frame (as the formula for time dilation CLEARLY states) NOT within the satellite's own reference frame.

Since the satellites are used to calculate positions within the earth's reference frame, corections are required for both GR and SR; even though SR does NOT cause the clock on the sattelite to actually slow down relative to the reference frame of the sattelite itself.

Alas Johan that 7us/day correction prove sthis wrong.

There is no observational correction, because the relative velocity & distance of the satellite do not change from one orbit to the next. But there is a corretion which gets larger each day. It must be real - use Einstein clock synchronisation at two distinct times to see it. Errors are bounded by light speed to satellite.

[tomclarke]

How does any of this mandate that either both must be observational or both must be real? It doesn't.

Bravo! Btavo!

Both are real within the reference frame on earth. GR is also real within the sattelite's reference frame. SR is only real within the earth's reference frame (as the formula for time dilation CLEARLY states) NOT within the satellite's own reference frame.

Since the satellites are used to calculate positions within the earth's reference frame, corections are required for both GR and SR; even though SR does NOT cause the clock on the sattelite to actually slow down relative to the reference frame of the sattelite itself.

Alas Johan that 7us/day correction prove sthis wrong.

There is no observational correction, because the relative velocity & distance of the satellite do not change from one orbit to the next. But there is a corretion which gets larger each day. It must be real - use Einstein clock synchronisation at two distinct times to see it. Errors are bounded by light speed to satellite.