Gravity Probe B (GP-B) seems like the experiment that never ends. The spacecraft set records for the longest-running development project at NASA — 40 years — and data analysis is stretching into its second year. The GP-B team may one day announce the most precise measurements yet of a long-sought effect of general relativity. But that didn't happen on 14 April, when project scientists presented an interim report to the American Physical Society meeting in Jacksonville, Florida.

Team members promise a final report by December, when money for the $760-million experiment runs out. But it is clear that unexpected systematic errors will make it a real challenge to reach the original mission goals. For the most subtle effect measured, the GP-B team needs to make the experiment's uncertainty 100 times lower.

NASA's Gravity Probe B aims to confirm two predictions of general relativity. Credit: K. STEPHENSON, STANFORD UNIV., LOCKHEED MARTIN CORP.

GP-B is a simple concept that in reality proved overwhelming. The experiment, proposed in 1964, required four perfect spinning gyroscopes in Earth orbit to measure how the spinning planet drags the fabric of space-time around with it — a phenomenon predicted by Einstein's general theory of relativity and called 'frame-dragging'. The mission survived every NASA attempt to cancel it (see Nature 426, 380–381; 2003) and was finally launched in April 2004.

The main goal was to measure frame-dragging to within 1% accuracy. That would be ten times better than the best measurements so far, which were taken by bouncing laser signals off the Earth-orbiting LAGEOS satellites.

GP-B has had one success. At the meeting, the team announced the first experimental measurement of geodetic precession, another small distortion of space-time. The geodetic precession of the Moon around the Earth has previously been verified to an accuracy of about 0.7%. GP-B has now measured the effect on their probe to within 1.5%. The team hopes to reduce this further — but it is the much smaller frame-dragging effect that everyone cares about.

The problems plaguing the analysis are systematic errors: electrostatic effects on the spheres at the core of the gyroscopes cause misalignments and wobbles that vary unexpectedly over time. The data were also recorded in chunks because solar flares required the system to be rebooted.

Physicist Clifford Will, who chairs NASA's scientific advisory committee for the project, says such errors are a real headache: “There's art involved — it's a slightly nebulous business; that's why they are being careful.” Still, the team remains bullish about the remaining data analysis. “I'm not interested in being disappointed,” says Francis Everitt of Stanford University, the project's principal investigator.

Meanwhile, the LAGEOS findings could be further improved before the end of the year by incorporating a new model of Earth's gravitational field, as gathered by the GRACE spacecraft. Erricos Pavlis of the University of Maryland in Baltimore County, who works with the LAGEOS data, says that it would be nice to beat GP-B, but even so he doesn't want to see it fail.

“After all these millions spent and decades of people's work,” he says, “it's only fair that they get something out of this project.”