Two unexpected effects are plaguing a quixotic satellite designed to measure a subtle effect of gravity. But project scientists say they are optimistic that they can still extract a significant finding from their data.

The problems have delayed results from the satellite, known as Gravity Probe-B, until spring 2007, according to project head Francis Everitt, a physicist at Stanford University in California.

Gravity Probe-B is designed to detect a tiny effect of gravity known as frame dragging. According to Einstein's general theory of relativity, a spinning mass such as the Earth drags space-time around it like a mixer dragging batter. Gravity Probe-B hoped to detect the effect with four almost perfectly spherical gyroscopes attached to a telescope. The telescope is trained on a guide star, and the gyros are set spinning. With the satellite aligned to the star, the drag should tilt the gyros off their axes by a minuscule amount, around 0.000011 degrees in a year.

Gravity Probe-B's journey from conception into orbit was unusually long. It began with modest funding from NASA in 1964, but the high-precision instruments it required took around two decades to develop. The 1986 Challenger disaster further set back the launch of the probe, which was originally scheduled to fly aboard a space shuttle (see Nature 426, 380; 2003). After several near-cancellations, the satellite finally lifted off in April 2004. By then it had racked up a price tag of US$700 million, and another project, Lageos, had become the first to measure the frame-dragging effect (see Nature 431, 958; 2004).

Gravity effects: wobbles and torques are hampering the interpretation of space data. Credit: K. STEPHENSON/STANFORD UNIVER., LOCKHEED MARTIN

After an initial calibration phase, Gravity Probe-B took a year's worth of data, completing operations in September 2005. Results were expected by last summer, but the announcement never came.

Nature has learned that this is because two unanticipated effects are clouding the team's frame-dragging result. The first is a slight, circular wobble created by a combination of tiny imperfections in the gyroscopes and a gradual slowing of their spinning. Gravity Probe-B scientists anticipated the wobble, but their calculations had ruled out the possibility that its period would change. “But it happened,” says Everitt. “Nature decides what it wants to do.”

The second problem is a torque on the four gyros. The twisting force came about when the satellite's telescope occasionally slipped off the guide star, and it seems to be related to tiny electric fields between various components of the probe. The team didn't discover the effect until after the satellite finished taking data, but are now trying to disentangle it from the result. The problem appeared quite serious during the summer, but Everitt says the team is now working on a “very elegant” solution.

It is unclear how the problems will affect the mission's result. “The Gravity Probe-B team is playing its cards very close to its chest,” says Bernard Schutz of the Max-Planck Institute for Gravitational Physics in Potsdam, Germany. “Nobody outside the project knows whether they can understand and measure the effects well enough to reach their original objective.”

That objective was to measure the frame-dragging effect to within about 1%, which would be ten times more accurate than the Lageos result. Clifford Will, a physicist at Washington University in St Louis, who chairs NASA's scientific advisory board for the project and has been briefed by the team, says he is optimistic that there will be some sort of result. “But will it be what they promised NASA? I don't know.”

Everitt says that his team will announce a definitive result at the American Physical Society meeting in April 2007. He says he is still worried about whether they can produce a meaningful finding: “I don't think you should be relaxed on $700 million,” he says. “But I wouldn't have agreed to give a talk if I wasn't reasonably confident.”