* GPS Does NOT require Einstein Relativity. Fred and Ryan discuss the claim that GPS systems rely on Einstein's special and general relativity. They debunk this claim by explaining how GPS works and how clock synchronization is the key, not relativity. As noted GPS scientist Dr. Tom Van Flandern quipped, "They have basically blown off Einstein."
GPS takes both general and special relativity into account every day.
This link is a paper written by Van Flandern that discusses GPS, relativity, and some related subjects.
Here's the relevant paragraph:
To offset these two effects [general and special relativity effects], the GPS engineers reset the clock rates, slowing them down before launch by 39,000 nanoseconds a day. They then proceed to tick in orbit at the same rate as ground clocks, and the system "works." Ground observers can indeed pin-point their position to a high degree of precision. In (Einstein) theory, however, it was expected that because the orbiting clocks all move rapidly and with varying speeds relative to any ground observer (who may be anywhere on the Earth's surface), and since in Einstein's theory the relevant speed is always speed relative to the observer, it was expected that continuously varying relativistic corrections would have to be made to clock rates. This in turn would have introduced an unworkable complexity into the GPS. But these corrections were not made. Yet "the system manages to work, even though they use no relativistic corrections after launch," Van Flandern said. "They have basically blown off Einstein."
The problem is that they they DO make relativistic corrections after launch...every day (and every couple of hours, too). The satellites are loaded with information that tells the user where the satellites are/were at a particular time and how fast they are/were going (Keplerian orbital parameters). The satellites transmit a signal that conveys that information to the users and also gives a time-hack of when the transmission left the satellite antenna. The user receives the signal some amount of time after, based on the speed of light, which gives a clock offset time that can be compared with the receiver's local time. The offset, using the speed of light, gives the user the distance from the receiver unit to the satellite. Do that same thing with 3 other satellites, and the user can pinpoint his own location and time error.
Clock rates aren't actually changed to account for relativistic corrections prior to launch as stated, but they can be changed after launch, based on direct observations of that signal I mentioned earlier, as received by special monitor stations that also have highly accurate atomic clocks. So the on-orbit clocks can be compared with ground-based clocks after the clocks are powered up (the clocks are not powered during launch). Corrections are made to accommodate both relativity and errors in clock time and rate early on in the power-up sequence, and usually those don't need to be changed much after that, although errors in clock performance might necessitate it sometimes.
The other thing that needs to be corrected sometimes (maybe once a year) is the actual orbit the satellite is flying in. Orbits can degrade for several reasons, so a change in velocity ("Delta-V") is performed regularly on every satellite. When the velocity is changed, the satellite either speeds up or slows down, which also changes the orbit altitude. The altitude then has a commensurate effect on the gravity acting on the satellite (higher altitude results in less gravitational pull), which then affects the time-keeping of the satellite's clock. So you can see that when a Delta-V is performed, the satellite has a change in both its speed (special relativity effect) and its gravity (general relativity effect), which offset each other, but not equally. This offset has been used, based on expected new orbit parameters during the planning of a Delta-V, to estimate the the change that will need to be applied to the transmitted clock parameters (also in that signal I mentioned earlier). While no orbital maneuver is ever completely predictable, the estimations based on Einstein's SR and GR theories have been on target whenever the Delta-V is accomplished, leading to a more rapid return to service after the Delta_V operation is completed (satellites are disabled from use during the Delta-V).
So, despite Van Flandern's insistence, he is not correct to say that anyone has "blown off" relativity in the use of GPS. However, he might be correct that a particular part of the corrections were/are not necessary, if those corrections are small or counteract each other.
