Gravity Probe B

The December 21-28 2006 issue of Nature (v. 444, p. 978-979) contains a short news article stating that Nature has learned that “two unanticipated effects are clouding the [GP-B] team's frame-dragging results” and also that “results were expected by last summer but the announcement never came.”

The two issues referred to in Nature have been regularly reported to NASA and our GP-B Science Advisory Committee (SAC) and publicly via these status updates. They are: 1) The effect of polhode motion of the gyros on readout calibration (see the polhode story in last month's update) and 2) misalignment torques observed and calibrated during the post-science instrument calibration phase in August-September 2005 (see the four weekly updates of September 2005).

In August 2005, a three-phase data analysis plan was devised in order to properly handle these and other issues. As first reported in May 2006, our intent--reached in agreement with NASA--has been to make the first science announcement in April 2007. This is still our plan.

Note: the recent claims by Iorio that he has measured the frame-dragging precession from the Mars Polar Orbital Surveyor High-precision measurement of frame-dragging with the Mars Global Surveyor spacecraft in the gravitational field of Mars and found it to be (0.9937 ± 0.0053) times the GR prediction has been refuted by Kris Krogh in today's eprint: Iorio’s “high-precision measurement” of frame-dragging with the Mars Global Surveyor.

Iorio had claimed too high a precision on an orbit that was not that well determined - 'angular momentum wheel desaturation' thrusters had been regularly used for one thing and also there would have been unmodelled atmospheric braking for another.

We show here that the recent claim of a test of the Lense-Thirring effect with an error of 0.5% using the Mars Global Surveyor is misleading and the quoted error is incorrect by a factor of at least ten thousand. Indeed, the simple error analysis of [1] neglects the role of some important systematic errors affecting the out-of-plane acceleration. The preliminary error analysis presented here shows that even an optimistic uncertainty for this measurement is at the level of, at least, ~ 3026% to ~ 4811%, i.e., even an optimistic uncertainty is about 30 to 48 times the Lense-Thirring effect. In other words by including only some systematic errors we obtained an uncertainty almost ten thousand times larger than the claimed 0.5% error.

Since lunar ranging currently provides a part in a thousand fit to the gravitomagnetic contributions to the lunar orbit, this feature of post-Newtonian gravity is not adjustable to fit any anomalous result beyond the 0.1% level from Gravity Probe-B without disturbing the existing fit of theory to the 36 years of LLR data.