Re: satellite clock gravitometer

[Leigh Palmer <palmer@SFU.CA>]

At 9:53 AM -0500 1/28/01, Chuck Britton wrote:
> At 11:33 PM -0800 1/27/01, Leigh Palmer,  you wrote about Re:
> satellite clock gravitometer:
> > In fact, the time difference is still unaccountably discrepant after
> > relativistic modeling is done. The correction used is an empirical
> > one, but it has the correct sign. I've now told you how to calculate
> > the result; it may surprise you.
>
> I was indeed surprised when the GR course had us calculate the two
> 'corrections'.

At least you've had a GR course - I haven't. I need to learn some of
this fast because I'm going to be teaching a course with some GR
content. I can do things like gravitational redshifts, of course, but
it gets pretty viscous when I try to wade through those strings of
Greek indices. Anyone know a Royal Road to Enlightenment in GR?

> How big is the discrepancy (we weren't TOLD about THAT detail!)?

I don't know how large the discrepancy is. I was told about it by a
colleague at the University of Maryland. He is working on it. It is
not large, but it would significantly degrade the accuracy of the
GPS were it to be ignored. Since it is a systematic error one simply
measures it and compensates accordingly. The fact that there is no
fundamental theory behind the correction is no bar to using the system.

> Are there proposed sources for the discrepancy?

There are always wild ideas out there. If you look up the paper to
which I refer in an accompanying posting you will see one of them.
(My colleague is cited in that paper, by the way.)

Leigh