In a presentation today (Jan. 7) at the annual meeting of the American Astronomical Society, astronomers from the University of California, Berkeley, report the latest results from the Pulsar Timing Array. The array is a set of distant millisecond pulsars being tracked by a group of radio telescopes looking for changes in the pulsars' very precise flashing as a sign that they are being perturbed by gravitational waves.

"The results will have significant impact on models of the Big Bang and the ensuing formation of galaxies in the early universe," said graduate student Andrea Lommen of UC Berkeley.

One particular millisecond pulsar, called B1855+09, has been observed by two different telescopes for 17 years, the longest baseline of any millisecond pulsar. Lommen and UC Berkeley professor of astronomy Donald Backer reconciled the last four years of observations at an upgraded 305-meter telescope at the National Astronomy & Ionosphere Center Arecibo Observatory in Puerto Rico, with previous observations at Arecibo and at the National Radio Astronomy Observatory's 42-meter telescope in Green Bank, W. Va. While smaller, the Green Bank telescope data collected by Backer were crucial in connecting data sets from before and after the period when the Arecibo telescope was being upgraded.

The measurements and analysis required keeping track of 91,481,630,234 rotations of the star with microsecond precision. The parameters of the analysis provide exquisite measurement of the star's location in the sky and its motion across the sky owing to the combined motions of the solar system and the pulsar.

Lommen and Backer found no unexplained perturbations in this long record of the B1855+09 pulsar that could be the result of the gravitational wave background radiation. The analysis extends the time base of previous work by Vicky Kaspi, now at McGill University in Montreal, from eight years to 17 years. The longer time base makes the UC Berkeley analysis sensitive to gravitational waves with longer periods, which, in turn, leads to a reduction in the upper limit of energy density in the gravitational wave background radiation by a factor of 10.

"The standard energy density for comparison in astrophysics is not the cosmic microwave background radiation, but rather the energy density whose mass equivalent would lead to the universe being just gravitationally bound," Backer said. "In these terms, our limit is about a billion times smaller. More importantly, we are now reaching a level where modern models of the universe of massive black holes suggest that we may expect to detect a signal."

Backer and Andrew Jaffe of University College, London, have worked on such models, which predict higher amplitude gravitational waves at increasingly longer wavelengths.

The gravitational waves the team is looking for come from massive black holes orbiting one another, the expected result of the merger of two galaxies with black holes at the center. Various theories predict that this occurs frequently in the universe, and the long-period gravitational waves produced by the circling black holes should permeate the universe.

"Pairs of massive black holes eventually will coalesce into a single hole as they radiate gravitational radiation, a prediction of Einstein's General Theory of Relativity," said Backer. "Gravitational waves propagate as a ripple in the fundamental structure of space-time."

The team is looking for long-period gravitational waves, with periods up to 10 years, in contrast to ground-based detectors such as the Laser Interferometer Gravitational Wave Observatory, or LIGO, which are searching for waves with periods of milliseconds.

The current view of the universe holds that galaxies like our Milky Way grow by mergers of smaller star systems. Over the past decade, astronomers have become increasingly sure that many galaxies harbor in their centers massive black holes with masses hundreds of millions times larger than the mass of the sun.

A galaxy merger event with two central massive black holes will lead to the two holes sinking to a common center and forming a binary pair in orbit about each other. This sinking of the holes happens for the same reason that satellites orbiting the Earth experience a drag force and plunge into the atmosphere.

Pairs of massive black holes eventually will coalesce into a single hole as they radiate gravitational radiation, a prediction of Einstein's General Theory of Relativity. Gravitational waves propagate as a ripple in the fundamental structure of space-time.

Stars in our galaxy called pulsars act as precision celestial clocks, and allow astronomer to detect, or place limits, on gravitational radiation passing through the solar system. Pulsars are rapidly rotating, highly magnetized neutron stars. The gravity wave signal from any single massive black hole system is too weak to detect, but astronomers are nearing the moment of detection of the cacophony of gravitational radiation from all the coalescing massive black hole events throughout the universe, Backer said.

"This radiation that we seek to detect is exceedingly weak - a million times less energetic than that in the cosmic microwave background radiation which, itself, is a billion times weaker than that in a household oven," he said.

Pulsars emit intense microwave beacons that we detect as pulses each rotation of the star. A class of pulsars with millisecond rotation periods provide the highest precision, pulse arrival time measurements using largest radio telescopes. The millisecond pulsars have also been found to be the most stable rotators, hence their usefulness as celestial clocks.

The measurements were performed at two wide (112 MHz) bands in radio frequency centered at 1410 and 2380 MHz. Dual-frequency measurements were necessary to remove effects of interstellar dispersion. Pulsars pulse more sharply at these relatively high radio frequencies and the signal is less disturbed by interstellar weather. The large bandwidth allowed Lommen and Backer to acquire more signal. Overall the combination of multiple high radio frequency with large bandwidth allowed for optimum accuracy in timing.

Lommen, Backer and colleagues at Princeton University, NRAO, Nancay, France, and Parkes, Australia, are monitoring an array of millisecond pulsars to improve on current results.