[imatfaal]

I voted for the watches because it's closest to what I think happened: The chronometer wasn't started and stopped accurately enough

It would be a bit embarrassing if it turned out that was the cause, but I suspect it's quite tricky to properly synchronize two clocks at different locations. Anyone know what the procedure is, and how much tolerance is acceptable for this experiment?

Nah - synchronizing the clocks is a comparative doddle (using doddle almost completely wrongly).

Metrology institutes around the world have long established the common view GPS method of synchronization/comparison.  This algorithm and protocol for synchronizing geographically remote clocks is not new and has been exhaustively tested (there remains the possibility that Gran Sasso didnt follow the instructions properly).

I seem to remember that with the upgraded beacons and after the swiss and german metrology institutes (now I bet they are a bunch of fun and hoopy guys) got involved that the max error was single nanaoseconds

[imatfaal]

If in fact photons are slowed by the small amount of inter stellar gas could this not solve the supernova event anomaly as the density of gas in the local region is about 1000 times higher than in inter galatic space

The refraction/scattering by gas/dust that affected light but not the neutrinos was indeed the cause of the slight delay in observations from SN1987a.  The neutrinos were a few hours earlier than the photons - but if Gran Sasso speed differentials are correct then the neutrinos should have been 3 years or so early - maybe they were and we never noticed them or maybe different energy neutrinos behave differently

[Geezer (OP)]

I voted for the watches because it's closest to what I think happened: The chronometer wasn't started and stopped accurately enough

It would be a bit embarrassing if it turned out that was the cause, but I suspect it's quite tricky to properly synchronize two clocks at different locations. Anyone know what the procedure is, and how much tolerance is acceptable for this experiment?

Nah - synchronizing the clocks is a comparative doddle (using doddle almost completely wrongly).

Metrology institutes around the world have long established the common view GPS method of synchronization/comparison.  This algorithm and protocol for synchronizing geographically remote clocks is not new and has been exhaustively tested (there remains the possibility that Gran Sasso didnt follow the instructions properly).

I seem to remember that with the upgraded beacons and after the swiss and german metrology institutes (now I bet they are a bunch of fun and hoopy guys) got involved that the max error was single nanaoseconds

Well yes, but how does it actually work? If science was based on "following the instructions", we'd still be hanging from a gumm tree.

[syhprum]

Imatfaal

has any research been on the slowing of photons by low density gas I know that it almost impossible to produce a vacuum as low as interstellar space but are there any figures published as to how c varies as near vacuum is approached so that any trend might apparent.
We should not be too surprised at photons being slowed down we see examples of it happening around us all the time the very reason we can see anything is due to the slowing down as they pass thru the lenses of our eyes.

[yor_on]

I don't know if Gran Sasso, 730 km away, uses  Carrier phase tracking? But "Tracking carrier phase signals provides no time of transmission information. The carrier signals, while modulated with time tagged binary codes, carry no time-tags that distinguish one cycle from another. The measurements used in carrier phase tracking are differences in carrier phase cycles and fractions of cycles over time. At least two receivers track carrier signals at the same time. Ionospheric delay differences at the two receivers must be small enough to insure that carrier phase cycles are properly accounted for. This usually requires that the two receivers be within about 30 km of each other. Carrier phase is tracked at both receivers and the changes in tracked phase are recorded over time in both receivers." and that's not 700 km

Still "All carrier-phase tracking is differential, requiring both a reference and remote receiver tracking carrier phases at the same time.
Unless the reference and remote receivers use L1-L2 differences to measure the ionospheric delay,  they must be close enough to insure that the ionospheric delay difference is less than a carrier wavelength.
Using L1-L2 ionospheric measurements and long measurement averaging periods, relative positions of fixed sites can be determined over baselines of hundreds of kilometers.
Phase difference changes in the two receivers are reduced using software to differences in three position dimensions between the reference station and the remote receiver. High accuracy range difference measurements with sub-centimeter accuracy are possible. Problems result from the difficulty of tracking carrier signals in noise or while the receiver moves. Two receivers and one SV over time result in single differences."

It's? Maybe, maybe not, working.

[syhprum]

I am sure they did it many many times and averaged out varying ionospheric conditions.
It is possible to compute c from measurements of magnetic and electrostatic phenomena how accurately can these be measured and how closely does the the measured value of c compare.

[imatfaal]

I voted for the watches because it's closest to what I think happened: The chronometer wasn't started and stopped accurately enough

It would be a bit embarrassing if it turned out that was the cause, but I suspect it's quite tricky to properly synchronize two clocks at different locations. Anyone know what the procedure is, and how much tolerance is acceptable for this experiment?

Nah - synchronizing the clocks is a comparative doddle (using doddle almost completely wrongly).

Metrology institutes around the world have long established the common view GPS method of synchronization/comparison.  This algorithm and protocol for synchronizing geographically remote clocks is not new and has been exhaustively tested (there remains the possibility that Gran Sasso didnt follow the instructions properly).

I seem to remember that with the upgraded beacons and after the swiss and german metrology institutes (now I bet they are a bunch of fun and hoopy guys) got involved that the max error was single nanaoseconds

Well yes, but how does it actually work? If science was based on "following the instructions", we'd still be hanging from a gumm tree.

Both ends have atomic clocks - whilst one satellite is overhead and both sites are receiving the signal, a record is made of the offset/delay between each clock and the gps signal (which is also atomic clock based)

If you then subtract the delay of Cern from the delay of Gran Sasso you end up with the difference between the two clocks (and the signal time & its error).  As we know the signal propagation time and error very accurately (it is the basis of GPS) we can calculate the difference between the clocks.  Only real error introduced is through differences in atmospheric conditions (which can be covered by repeating regularly and at night/day), or and by the ephemeris error.  The max error for intercontinental long range common view is about 10 nanoseconds - for such a short baseline the major error (ephemeris) is lowered significantly

And frankly Geezer - scientific experimentation is very much about following instructions.  You need to be consistent in your operations; and when one area of expertise has developed a technique and it has been found to be accurate then you use that knowledge.  ie when a group of nuclear physicists want to know how to measure time over a great distance they use a  technique developed by the NIST/USNO, and it is checked and implemented by a national Metrology institute - so in effect they just follow instructions.  So as long as Gran Sasso followed the technique carefully and did not vary the protocol then the validity of the result and more importantly the error has already been shown by experts/academics.

[CliffordK]

We already know that the speed of light varies based on the medium.  And there is apparently some interaction of light with local space...  causing the space where the measurement is taken to influence the actual measurement of the speed of light.

So...
if there is an asymptotic maximum speed for particles in local-space, it isn't too surprising to find particles that interact less with what is around them, and thus travel slightly faster than light.

The big question I would have is whether all neutrinos from natural sources distant from Earth travel to Earth at the same speed, or perhaps they are dependent on the relative velocity of the source.

[Geezer (OP)]

And frankly Geezer - scientific experimentation is very much about following instructions.  You need to be consistent in your operations; and when one area of expertise has developed a technique and it has been found to be accurate then you use that knowledge.  ie when a group of nuclear physicists want to know how to measure time over a great distance they use a  technique developed by the NIST/USNO, and it is checked and implemented by a national Metrology institute - so in effect they just follow instructions.  So as long as Gran Sasso followed the technique carefully and did not vary the protocol then the validity of the result and more importantly the error has already been shown by experts/academics.

Fully agree that a lot of it is simply following the instructions, and very rigorously at that. That usually works very well, but when there is an anomaly, everything is up for grabs. Anomalies (real or apparent) force scientists to ask very uncomfortable questions that move science forward.

My only point was that if nobody ever challenged the "instructions", the Earth would still be at the centre of the Universe.

[Geezer (OP)]

Metrology institutes around the world have long established the common view GPS method of synchronization/comparison.  This algorithm and protocol for synchronizing geographically remote clocks is not new and has been exhaustively tested

I'm sure these guys know what they are doing, but I found it slightly disconcerting that the description says things like "A and B, receive a one-way signal simultaneously"

If they know it's simultaneous, why bother? Seems a bit circular to me, but then, I'm well known for being nitpicky.

[imatfaal]

Clifford - not quite sure of the effects you are talking about - could you elaborate?  The problem with measuring neutrinos is that we are so bad at detecting them that we can only be certain of a source in very rare circumstances.  Supernova SN1987a is one such event - as Syhprum explained the neutrinos arrived a few hours before the light did, it is possible we missed the "othersort of neutrino" that were superluminal.  But to work out how fast Neutrinos travel in space we need to know precisely what caused them and when/where that was - and there are precious few opportunities for that.

[imatfaal]

Metrology institutes around the world have long established the common view GPS method of synchronization/comparison.  This algorithm and protocol for synchronizing geographically remote clocks is not new and has been exhaustively tested

I'm sure these guys know what they are doing, but I found it slightly disconcerting that the description says things like "A and B, receive a one-way signal simultaneously"

If they know it's simultaneous, why bother? Seems a bit circular to me, but then, I'm well known for being nitpicky.

Doesn't say they receive the same bit of the signal simultaneously - or that the signals are synchronized.

If Mrs Geezer puts on her favourite Balck Sabbath album at full volume at the other end of the house it is quite acceptable to say that you and she are listening simultaneously - although with the finite speed of sound, combined with the magnificence of geezer manor she might hear each of OO's dulcet tones a few seconds before you. 

In fact if you did a bit of measuring (distance, air pressure, humidity etc) you could tell which of your two clocks is slow.  ie you both look at the clocks in your respective rooms and note the time of the very last bit of Paranoid.

If you clocked it at 15 seconds exactly past the hour, and Mrs G at 10 seconds past exactly - Mrs G is 3.2 m from the stereo and you are 340 m in the other direction (it's a nice dry day at your beach house by the way and about 60F due to the aircon) - after you and Mrs G compare notes; you can work out that your clock is 4 seconds slower than Mrs G's.

Common view relies on a signal being received simultaneously - but possibly not in synch.  The fact that the message is the same - and you can the compare notes afterwards allows you to workout the difference in your clocks without worrying about synchronization.   The common view part allows you to remove the clock on the satellite from the equation - if I haven't screwed up, the above example would still work if your played Ozzie and friends at 45 or 78.  The only sources of error are where the satellites actually are (and we know that fairly well) - and the possibility that one leg of the signal gets bounced around a bit more than the other.  At 734 km baseline these errors reduce to fractions of a second  (2.3ns)

[Geezer (OP)]

Metrology institutes around the world have long established the common view GPS method of synchronization/comparison.  This algorithm and protocol for synchronizing geographically remote clocks is not new and has been exhaustively tested

I'm sure these guys know what they are doing, but I found it slightly disconcerting that the description says things like "A and B, receive a one-way signal simultaneously"

If they know it's simultaneous, why bother? Seems a bit circular to me, but then, I'm well known for being nitpicky.

Doesn't say they receive the same bit of the signal simultaneously - or that the signals are synchronized.

If Mrs Geezer puts on her favourite Balck Sabbath album at full volume at the other end of the house it is quite acceptable to say that you and she are listening simultaneously - although with the finite speed of sound, combined with the magnificence of geezer manor she might hear each of OO's dulcet tones a few seconds before you. 

In fact if you did a bit of measuring (distance, air pressure, humidity etc) you could tell which of your two clocks is slow.  ie you both look at the clocks in your respective rooms and note the time of the very last bit of Paranoid.

If you clocked it at 15 seconds exactly past the hour, and Mrs G at 10 seconds past exactly - Mrs G is 3.2 m from the stereo and you are 340 m in the other direction (it's a nice dry day at your beach house by the way and about 60F due to the aircon) - after you and Mrs G compare notes; you can work out that your clock is 4 seconds slower than Mrs G's.

Common view relies on a signal being received simultaneously - but possibly not in synch.  The fact that the message is the same - and you can the compare notes afterwards allows you to workout the difference in your clocks without worrying about synchronization.   The common view part allows you to remove the clock on the satellite from the equation - if I haven't screwed up, the above example would still work if your played Ozzie and friends at 45 or 78.  The only sources of error are where the satellites actually are (and we know that fairly well) - and the possibility that one leg of the signal gets bounced around a bit more than the other.  At 734 km baseline these errors reduce to fractions of a second  (2.3ns)

Ah! Right, I see what you mean. It might have been better if they had left out "simultaneous" and said "A and B receive the same transmission" or similar.

How accurately do the distances have to be known for this to work, or does it not require any knowledge of distance? Or does the distance measurement depend on a highly accurate common view of time?

[CliffordK]

As far as clocks.
It would be easy enough to put the two clocks together.
Synchronize them.
Separate them and run the experiment.
Bring them back together and verify that they are still in sync.

Or, perhaps have a third physical "master clock", that would be moved from one location to the other.
And, then just repeatedly verify that the "slave clocks" on each end match the master clock.

If jet aircraft affect atomic clocks, then carry them in shielded containers in trucks, trains, or cargo ships on the ground.

[syhprum]

It is not Jet aircraft that disturb atomic clocks but the reduced gravity at altitude and the speed of travel cause time dilation effects which of course you cannot shield against.

[imatfaal]

As far as clocks.
It would be easy enough to put the two clocks together.
Synchronize them.
Separate them and run the experiment.
Bring them back together and verify that they are still in sync.

Or, perhaps have a third physical "master clock", that would be moved from one location to the other.
And, then just repeatedly verify that the "slave clocks" on each end match the master clock.

If jet aircraft affect atomic clocks, then carry them in shielded containers in trucks, trains, or cargo ships on the ground.

Clifford

I think Syhprum has covered most of points in your post - however one other thing, they did move an accurate clock between the two.  I would recommend a quick scan of the experimental write up (Mike helpfully posted it above) you can read about the timing on pages 7-9.  Basically common view synchronization is using a third clock - ie the one on the gps satellite

As Geezer and I mentioned above - when you want great accuracy in a technique you go to those who have great expertise, and the way that the metrologists do this synchronization of clocks is incredibly accurate.  The Atomic clocks spread over the USA are accurate to within 1 ns of each other - bearing in mind that some are in Flagstaff Arizona and others are at sea-level you cannot just move one to another as a clock at 2000m and one at sea-level will run at different rates

[imatfaal]

Metrology institutes around the world have long established the common view GPS method of synchronization/comparison.  This algorithm and protocol for synchronizing geographically remote clocks is not new and has been exhaustively tested

I'm sure these guys know what they are doing, but I found it slightly disconcerting that the description says things like "A and B, receive a one-way signal simultaneously"

If they know it's simultaneous, why bother? Seems a bit circular to me, but then, I'm well known for being nitpicky.

Doesn't say they receive the same bit of the signal simultaneously - or that the signals are synchronized.

If Mrs Geezer puts on her favourite Balck Sabbath album at full volume at the other end of the house it is quite acceptable to say that you and she are listening simultaneously - although with the finite speed of sound, combined with the magnificence of geezer manor she might hear each of OO's dulcet tones a few seconds before you. 

In fact if you did a bit of measuring (distance, air pressure, humidity etc) you could tell which of your two clocks is slow.  ie you both look at the clocks in your respective rooms and note the time of the very last bit of Paranoid.

If you clocked it at 15 seconds exactly past the hour, and Mrs G at 10 seconds past exactly - Mrs G is 3.2 m from the stereo and you are 340 m in the other direction (it's a nice dry day at your beach house by the way and about 60F due to the aircon) - after you and Mrs G compare notes; you can work out that your clock is 4 seconds slower than Mrs G's.

Common view relies on a signal being received simultaneously - but possibly not in synch.  The fact that the message is the same - and you can the compare notes afterwards allows you to workout the difference in your clocks without worrying about synchronization.   The common view part allows you to remove the clock on the satellite from the equation - if I haven't screwed up, the above example would still work if your played Ozzie and friends at 45 or 78.  The only sources of error are where the satellites actually are (and we know that fairly well) - and the possibility that one leg of the signal gets bounced around a bit more than the other.  At 734 km baseline these errors reduce to fractions of a second  (2.3ns)

Ah! Right, I see what you mean. It might have been better if they had left out "simultaneous" and said "A and B receive the same transmission" or similar.

How accurately do the distances have to be known for this to work, or does it not require any knowledge of distance? Or does the distance measurement depend on a highly accurate common view of time?

Two distances involved - not sure which you are asking about. 

Distance to satellite. You do need to know where the satellite is and the distance (and more importantly the error) to each of the base-stations; but the speed of the signal is so damn high that even estimates with high ephemeris error will lead to a nano-second error at such a short baseline

Distance on baseline.  As you know the distance to a common known point (the gps sat) then you do know the baseline - but it is not part of the calculation

I am not aware of the error algorithms - but everything I have read suggested that at over 3000km you need to start being careful of errors close to 10ns.  But 730km is actually quite a short baseline distance for this technique and the max error is 2.3ns.  The subtraction of one data set from the other means that errors are cancelled not accrued - most of the error in a signal coming from a satellite 22000 km away to two posts 730km apart will be the same.

[syhprum]

In the days before telegraphs etc London watch makers had their timepieces synchronised by a courier carrying a watch synchronised at Greenwich to their establishments.

[Airthumbs]

Bulk jumpers  []

[Geezer (OP)]

Two distances involved - not sure which you are asking about. 

Distance to satellite. You do need to know where the satellite is and the distance (and more importantly the error) to each of the base-stations; but the speed of the signal is so damn high that even estimates with high ephemeris error will lead to a nano-second error at such a short baseline

Distance on baseline.  As you know the distance to a common known point (the gps sat) then you do know the baseline - but it is not part of the calculation

I am not aware of the error algorithms - but everything I have read suggested that at over 3000km you need to start being careful of errors close to 10ns.  But 730km is actually quite a short baseline distance for this technique and the max error is 2.3ns.  The subtraction of one data set from the other means that errors are cancelled not accrued - most of the error in a signal coming from a satellite 22000 km away to two posts 730km apart will be the same.

Thanks!

What I'm (sort of) getting at is that, for this experiment, both the distance and time have to be very accurately known. However, if I understand properly how this works, the accuracy of the time synchronization depends on a very accurate measurement of the baseline. If the baseline is a bit off I think it will introduce an error into the time synchronization, although it's probably very small.

It's probably highly unlikely that would account for the observed results, but it might be a contributing factor.

BTW, you only know the distance from the satellite to the receiver if they have a common view of time, and the speed of the signal really has nothing to do with it.