By combining observations at gamma-ray and radio wavelengths, astronomers have rapidly increased the known number of millisecond pulsars in the Milky Way. The newly discovered pulsars, found using NASA's Fermi Gamma-ray Space Telescope and ground-based radio telescopes, could form part of a galactic-scale observatory for detecting hitherto elusive gravitational waves.

Gravitational waves are ripples in the fabric of space–time that Einstein's theory of general relativity predicts will occur whenever massive bodies are accelerated. Gravity waves at relatively high frequencies of about 1 Hz could be produced by the merger of neutron stars, for example. Physicists hope to detect such waves using a number of huge laser interferometers. These devices rely on the interference of two laser beams arranged at right angles to one another, with any passing wave producing a miniscule shift in the interference pattern.

An alternative, and cheaper, way to detect gravitational waves is to utilize millisecond pulsars – the extremely dense remnants of burnt-out stars that emit well-defined beams of electromagnetic radiation that sweep round like a beam from a lighthouse. Certain kinds of older pulsar can spin extremely rapidly and send out hundreds of pulses per second. These frequencies vary by less than a microsecond over very long periods of time and such pulsars therefore make ideal timekeepers.

All in the timing

The time it takes for these pulses to arrive at a radio telescope would be changed very slightly by any gravitational wave passing between pulsar and telescope. Therefore, by measuring the changes in the relative timing of large numbers of such pulsars, each with a different line of sight to the Earth, it should be possible to reveal the presence of a gravitational wave as well as record its direction of propagation and polarization.

Separated from one another by thousands of light years, these pulsars would enable the detection of waves with frequencies of nanohertz, which could be produced from sources such as black-hole binaries, which form when galaxies merge, or the "cosmic strings" believed to have existed in the early universe.

Millisecond pulsars were first discovered almost 30 years ago but the huge amount of telescope and computing time needed to find them using sky surveys at radio wavelengths has meant they have remained a relative rarity. Until recently astronomers had uncovered only about 150 of them, some 90 of which were grouped tightly together in star clusters and therefore unsuited for detecting gravitational waves. Now, however, Paul Ray of the Naval Research Laboratory in Washington, DC and colleagues have accelerated this discovery process by using the Fermi satellite's Large Area Telescope.

This instrument has been surveying the heavens at gamma-ray wavelengths since August 2008 and in that time has uncovered some 1000 new sources of gamma rays. Ray suspected that some of these sources were pulsars and so set up a team of astronomers to analyse them in finer detail using a number of radio telescopes around the world, including the Green Bank Telescope in West Virginia and the Parkes Observatory in Australia. After just three months of analysis on a fraction of the Fermi data, Ray and colleagues have found 17 new millisecond pulsars, and he reckons that they might double this number by studying the remaining unidentified sources.

With a little luck

Ray believes that "with continued good luck discovering millisecond pulsars and enough dedicated telescope time", gravitational waves might be detected using a pulsar array within the next decade. He points out that astronomers in the US, Europe and Australia are currently putting together proposals to increase the fraction of time devoted to pulsar observations on facilities such as the Green Bank and Parkes telescopes and to develop advanced software to process the huge amounts of data involved. He adds there is a small chance that a pulsar array might scoop the interferometers in making the first direct detection of gravity waves.

One of the physicists working on interferometers, Jim Hough of the University of Glasgow, agrees that pulsar timing is a good way to search for gravity waves at extremely low frequencies. He believes that if astronomers observe 20 pulsars with a timing precision of better than 100 nanoseconds for five years then they would "have a very good possibility of observing gravitational-wave signals."

The research was reported at a meeting of the American Astronomical Society in Washington, DC.