General relativity passes cosmic test
Einstein's theory holds in extreme gravitational fields.
Einstein's general theory of relativity has passed its toughest test yet.
By looking at a pair of neutron stars the densest objects in the Universe after black holes scientists have made the most accurate measurement yet of how strong gravitational fields bend radio waves. This reduces the room for theories that predict Einstein's equations will fail near very dense masses.
Albert Einstein's theory of general relativity says that gravity results from the curvature of space-time. He developed a series of equations, more complex and less famous than E=mc2, to describe how space-time is curved.
These equations predict how objects such as neutron stars, which are about twice as massive as our Sun but only tens of kilometres across, will distort the space around them. The theory also predicts that when stars orbit each other they create gravitational waves that distort space-time as they ripple outwards, and that the stars move towards each other as a result.
Now a team of astrophysicists, led by Michael Kramer at the University of Manchester, UK, have tested these notions by measuring how neutron stars' gravitational fields increase the time it takes radio waves to reach Earth, and how the stars are moving in relation to each other.
Their measurements show that the equations of general relativity are accurate to at least one part in 20,000, or 0.05%, beating the old record of one part in 500. The results are published in Science this week1.
The system they studied, called PSR J0737-3039A/B, is the first known binary star system made of two pulsating neutron stars, or pulsars.
Pulsars are ideal for testing general relativity. Their gravitational fields are 100,000 times stronger the Sun's, and they emit radio waves at very regular intervals, making any distortions easy to measure. "They are like cosmic stopwatches," says Kramer.
And at only 20 km across, they can be thought of as dots instead of three-dimensional spheres, which makes the equations much easier to solve.
General relativity has been tested in the strong gravitational fields of binary systems before2, but this is the first system to be found where both stars are pulsars, making measurements even more reliable.
"We knew soon after the system was discovered that it was going to be an impressive laboratory for testing general relativity, and it's great to see the first high-precision results to come out of it," says physicist Clifford Will of Washington University in St Louis, Missouri.
General relativity has been tested in our Solar System to a precision of 0.002%3. But this was in a mild gravitational field. Some theorists think that it is only in extreme conditions that general relativity breaks down.
"We know that general relativity breaks down at very small scales when quantum mechanics takes over. We are trying to see if it breaks down at very large scales," says Kramer.
The most popular alternatives to general relativity involve what are called 'tensor-scalar' fields. If such fields exist, it could have weird consequences for the laws of physics: momentum may no longer be conserved, and Newton's gravitational constant may change over space and time.
The demonstration that general relativity works for pulsar pairs will not necessarily deter those who support competing theories. It could be that relativity breaks down only under even more extreme conditions, or it only deviates by an infinitesimal amount.
The Universe holds more extreme places to test relativity, says Kramer. "We would like to observe systems with even stronger gravitational fields such as a pulsar orbiting a black hole".
Other researchers are also trying to shore up Einstein's ideas. It is hoped that a ground-based detector will soon spot gravitational waves; a space-based detector will be online in the next 6 to 7 years. Their measurements will complement this study, says Kramer.
Visit our newsblog to read and post comments about this story.
- Kramer M., et al. Science, (2006).
- Van Straten W., et al. Nature, 412 . 158 - 160 (2001).
- Bertotti B., et al. Nature, 425 . 374 - 376 (2003).