special relativity's second postulate is invalid
On 1/12/13 3:07 PM, Koobee Wublee wrote:
Years of educating Tom, he has finally given in that this
synchronization of clocks between the ground and the satellites.
Sorry to but in, but Tom has it exactly right!
One of the requirements of the GPS is that GPS time remain equal to UTC within one microsecond, modulo leap seconds (UTC uses them, GPS does not). Without the correction to the satellite clock dividers, as calculated using GR, this requirement could not be met; indeed, the ground segment and the space segment could not be used together. We KNOW this because the first GPS satellites were operated for a few weeks without the modified divider, and their clocks drifted as predicted by GR; this drift is MUCH larger than measurement resolutions.
Yes, daily corrections to the individual satellites are generated by the ground segment and uploaded to the satellites for rebroadcast to the GPS receivers. There are a large number of small corrections; the largest are usually orbit corrections, but they also include clock drift, effects of sun and moon, current ionosphere parameters, etc. The GR correction corresponds to about 38 microseconds per day; the daily corrections (other than orbit) are typically a few nanoseconds -- a thousand times smaller.