'Epic' GP-B NASA probe proves Einstein right

Nearly half a century in the planning, armed with 13 new technologies and four of the most perfect spheres ever created, NASA's $750 million Gravity Probe B (GP-B)'s troubled mission is finally over.

After 16 months of in-orbit data gathering and almost five years of data analysis, GP-B has mapped the space-time vortex that Einstein predicted would be created by the Earth's movement in space. That prediction, which required instruments of exquisite sensitivity, has been confirmed to "... a geodetic precession of 6.600 plus or minus 0.017 arcseconds and a frame dragging effect of 0.039 plus or minus 0.007 arcseconds," according to Stanford University physicist Francis Everitt, principal investigator of the Gravity Probe B mission.

The figures are not as precise as those gathered by other experiments, including laser rangefinding of mirrors left on the Moon by Apollo mission astronauts, although they do provide independent corroboration through entirely novel means and close loopholes of uncertainty that left some of the other results still open to question.

"This is an epic result," said Clifford Will of Washington University in St. Louis, chair of the National Research Council panel set up by NASA to monitor GP-B. "One day," he said, "this will be written up in textbooks as one of the classic experiments in the history of physics." However, a NASA review panel had recommended in 2008 that the data analysis be abandoned due to operational problems, leaving the researchers dependent on private funding.

The most subtle effect being detected, frame dragging, is the way that the gravitational distortion of space and time caused by an object changes when that object moves. GP-B detected this through a set of four gyroscopes, which remained spinning in one direction no matter how they are moved. However, if space-time was distorted around them, they remained true to that direction - but the direction relative to the rest of the universe changes.

That effect is very small and easy to mask by external noise. Thus, the GP-B gyroscopes are by a very large factor the most precise ever made. Each is based on a ball around an inch and a half across, made from fused quartz and within forty atoms of precisely spherical. The balls are coated in a very thin layer of superconducting niobium, which generates a small magnetic field when it rotates.

That field is precisely aligned with the direction of spin of the balls; it was measured on the satellite by SQUIDs — Superconducting Quantum Interference Devices — that introduced no drag on the rotation. The balls span at 10,000 RPM in extreme vacuum, kept much freer of gas than the space surrounding the satellite in its 400 mile-high orbit, and would have lost less than one percent of their rotational speed in a thousand years.

All this was kept within a superconducting shield that blocked the Earth's magnetic field. The satellite itself monitored the position of one of the balls to within a nanometre, and constantly adjusted its position with microthrusters as it orbited around the planet to prevent any of the balls touching the rest of the apparatus. It also monitored the position of a microwave-emitting star 329 light years away, IM Pegasi, whose position was precisely measured by radiotelescope from Earth and used as a reference against which the gyroscopes' alignment can be measured.

Such lavish use of superconductivity needed huge amounts of cooling, to around two degrees above absolute zero. The spacecraft started with 400 kg of liquid helium, which both provided the cooling and, as it slowly boiled off, was used to power the microthrusters that maintained alignment.

GP-B launched in 2004 during the one-second launch window necessary to hit the precise orbit required, and at first checked out well.

However, interference from solar flares made the data much noisier than expected, and other unexpected effects reduced the amount and quality of information returned. For example, the gyroscopes exhibited unexpectedly complex rotational behaviour called Polhode motion that had to be modelled after the data had been collected, and retrospectively allowed for.

Perhaps most dangerously, electrostatic forces and magnetic interactions within the spacecraft put twisting forces on the gyroscopes that couldn't be distinguished from those caused by the relativistic effects being sought, threatening the entire mission. The controllers deliberately misaligned the satellite for a few days, which changed the direction of the internal errors but left the external forces unchanged; by using this to measure the errors, they could apply a correction factor to all data collected and recover all of the external measurements.

Nonetheless, the increased chance of error caused by such a concatenation of factors helped lead to NASA's decision to halt work on the data in 2008, due to doubts that the results would be sufficiently rigorous. The Saudi Royal Family contributed most of the $3.2 million funding that let the investigators finish the job.