While light or sound waves propagate through space and time, gravitational waves are propagating ripples of space and time itself. Such distortions of spacetime are generated by catastrophic astrophysical phenomena involving extremely compact objects (like e.g. colliding and merging black holes or neutron stars, or the collapsing core of a star in a supernova event).

Despite its origin in such a violent environment, a gravitational wave arriving on Earth will have a relative signal strength of at most 10^-20, which is only one part in 100,000,000,000,000,000,000. This corresponds to a spatial displacement of only about 1/100th of the size of an atomic nucleus in a 100 km long measuring rod!

Although gravitational waves were predicted by Einstein already at the beginning of the 20th century, only now we have the technology to venture the first successful detection. Currently, five kilometer-class Laser interferometer detectors are actively scanning the sky for incoming gravitational waves from both Galactic and extragalactic events (see Fig. 1). They are supported by resonant detectors, which in contrast to the current interferometers have their optimal sensitivity only in a very narrow frequency band. In the forseeable future, the existing experiments will be upgraded for better sensitivity, new already planned detectors are scheduled to be built, and the space-based satellite LISA will complement the detectors on Earth.

A successful direct detection will not only unequivocally prove Einstein's bold prediction, but even more interestingly, will open a completely new "window" onto the universe. By routinely observing gravitational waves, astrophysicists will gain new and otherwise entirely unattainable insights into such fascinating objects like black holes, the enigmatic cosmic gamma ray bursts, or the driving engines behind stellar supernova explosions.

One example of a promising source of gravitational waves is the gravitational collapse of the rotating core of a massive star to a neutron star and the subsequent explosion of the star in a spectacular supernova event: During the collapse, within fractions of a second, a mass larger than that of our sun is compressed to densities exceeding 100 millions of tons per cubic centimeter, until the contraction is stopped by the core bounce. However, so far the search for gravitational waves from a supernova was impeded by the rather crude knowledge of the expected burst signal from the core bounce.

A first approach to produce templates of gravitational wave signals from stars dying as a supernova was already successfully undertaken several years ago at the Max Planck Institute for Astrophysics (see Research Highlight November 2001). However, these templates predicted a large variability of the way the collapse and bounce proceeds. The corresponding uncertainty of the wave signal hampered an application of the most efficient available filters in the detector signal analysis.

Greatly improving on these previous results, again scientists from the Max Planck Institute for Astrophysics in cooperation with Christian D. Ott (previously Max Planck Institute for Gravitational Physics, Golm, Germany; now Department of Astronomy and Steward Observatory, University of Arizona, U.S.A.) and supported by Ian Hawke (School of Mathematics, University of Southampton, U.K.) as well as by Erik Schnetter and Burkhard Zink (both Center for Computation & Technology, Louisiana State University, U.S.A.) have now accomplished new multi-dimensional supercomputer simulations of many rotating stellar core collapse models. In contrast to the old simpler computations, now much better initial models for the pre-collapse stellar cores were used, a more realistic microphysical description of matter was implemented, and the effects of neutrinos during the contraction phase were taken into account.

Most interestingly, with all these important improvements taken into account, the previously obtained diversity of the gravitational wave signals disappears, with only one signal type surviving (see Fig. 2). This type is known as Type I and is characterized by a positive pre-bounce signal rise and a large negative spike during bounce, followed by a ring-down phase during which the signal slowly dies off.

The finding that for improved models the gravitational wave templates become more generic has important implications for a possible detection of such an event. Obviously, with better knowledge of the expected signal and more robust predictions, fine-tuned data analysis methods can be applied to extract such signals from the data streams measured in the detectors. More importantly, as the frequencies of the signal templates cluster in a very narrow range, the interferometer detectors themselves can be fine-tuned to search predominantly in a narrow frequency band (see Fig. 3). In addition, resonant detectors, which are already narrow-banded by construction, can very effectively join the search.

With such techniques and a combined effort of several detectors, the detection range can probably be extended far beyond our Galaxy in the near future. This greatly increases the prospects for a signal detection, as otherwise hopes would solely rest on the rather rare event of a supernova in our own Galaxy.

H. Dimmelmeier, H.-T. Janka, A. Marek, E. Müller

Publication

C.D. Ott, H. Dimmelmeier, A. Marek, H.-T. Janka, I. Hawke, B. Zink, and E. Schnetter,
"3D Collapse of Rotating Stellar Iron Cores in General Relativity with Microphysics",
Physical Review Letters, submitted; astro-ph/0609819

H. Dimmelmeier, C.D. Ott, H.-T. Janka, A. Marek, and E. Müller,
"Generic Gravitational Wave Signals from the Collapse of Rotating Stellar Iron Cores",
Physical Review Letters, submitted; astro-ph/0702305

C.D. Ott, H. Dimmelmeier, A. Marek, H.-T. Janka, B. Zink, I. Hawke, and E. Schnetter,
"Rotating Collapse of Stellar Iron Cores in General Relativity",
Proceedings of the New Frontiers in Numerical Relativity Conference, AEI Golm, Germany, 2007; astro-ph/0612638