General Relativity (GR) describes the effect of gravity on matter in terms of space-time curvature, which in turn is generated by matter, but also has its own dynamics.
Several of GR's classic predictions have been verified, including the bending of starlight during eclipses, delay of propagation of radar in the solar system, the gravitational redshift, and the perihelion precession of Mercury. GR can be further tested by the dynamics of relativistic binary pulsars, as well as space-based experiments. In modern technology, general relativistic corrections are essential to maintaining the accuracy of the Global Positioning System (GPS).
GR provides us with a physical description of compact objects with strong self gravity, in particular black holes, objects so compact that even light cannot escape from their surface, the horizon. Physics around the black-hole horizon, though in some sense peculiar, can be treated consistently by GR. According to current astronomical understanding, rather solid evidence has been found that stellar-mass black holes exist in some X-ray binary systems (assuming GR is correct). These black holes are thought to be end products of stellar evolutions. On the other hand, super-massive black holes (with masses 10^4 and above) are thought to exist at the center of galaxies (for example in our galaxy, the Milky Way). A third class of intermediate-mass black holes have also been conceived to exist, and there has been some observational evidence.
On the large, cosmological scale, GR predicts the dynamics of the expansion of the universe, provided an initial condition, and an equation of state for matters in the universe.
GR also portrays a dynamical picture of space-time itself --- it predicts a wave of space-time curvature that propagates on its own, without the help of matter. These gravitational waves are of great importance to our fundamental understanding of space-time. Similar to any other types of waves, although they propagate on their own, they must be generated by some means. Gravitational waves are generated by accelerated matter motion, or oscillations in space-time curvature.
In laboratory and the microscopic world, gravity is the weakest among all four fundamental interactions (eletromagnetic, strong, weak, and gravity). It only manifests itself at large scales (strong and weak interactions are short-ranged), and for macroscopic systems (which usually have equal number of positive and negative charges). The strongest gravitational waves are produced by the high-energy processes in the universe, for example coalescing binary black holes/neutron stars/white dwarfs, spinning neutron stars with non-vanishing ellipticity --- either isolated or within a binary, oscillations in nascent neutron stars or black holes, cosmological explosions such as supernovae, as well as density fluctuations from the early universe.
Gravitational waves are usually generated by the bulk motion of masses or space-time curvature, and reflect a different aspect of the source as those revealed by electromagnetic radiations --- some potential gravitational-wave sources are not even visible from electromagnetic radiations; in particular GWs originating from the oscillations of black holes. Some sources of gravitational waves are also expected to reside in the highly non-linear regime of general relativity, which is currently not well understood. The gravitational-wave stochastic background can also probe an era much earlier than the Cosmic Microwave Background (CMB), since the interaction of gravitational waves with matter is much weaker than electromagnetic waves (and therefore they decouple from the rest of the universe at a much earlier time).
In summary, gravitational waves, aside from their unique significance for fundamental physics, can also provide an entirely “new window” to the universe --- the energetic, exotic and the early. However, the detection of gravitational waves requires extraordinary measurement sensitivity.
Interaction of Gravitational-waves with Matter
The effect of gravitational waves on closely separated (distance shorter than the wavelength) free test masses can be characterized by a time-dependent tidal force, which can be described by h, the strain it induces on the distance between two masses. (Formally h is the amplitude of the space-time metric perturbation.) As they arrive at the Earth from their distant sources, astronomical gravitational waves are extremely weak. For example, a pair of merging neutron stars near the center of our galaxy will generate an amplitude h~3×10-17 at around 100 Hz. For two test masses separated by 4 kilometers, this will generate a relative displacement of 10-13meter, which is 1/1,000 the size of a hydrogen atom! However, even such a “ loud” event is estimated to happen in our galaxy only once per 104--105 years, which means we must be able to reach further into the universe, to include at least ~104--105 “Milky-Way-equivalent” galaxies to be able to detect such an event in one year’s observation time. As it turns out, we need to be able to see h~10-23, over a million times better, in order to be able to make routine detections. This means we will need a displacement sensitivity of 1/10,000 the size of an atomic nucleus.
This example suggests that we need superb precision-measurement technology to detect gravitational waves. On the other hand, the fact that gravitational waves are so hard to generate also means that they do not interact easily with the matters (mainly dusts) they travel through in the universe --- and information of the source is preserved faithfully.
History of gravitationa-wave detection
The quest to detect gravitational waves began in the 1960s, when Weber build the first resonant mass detectors (also called the "bar detectors") to sense the tiny oscillations gravitational waves induce at their resonant frequencies (at a few hundred Hz to several kHz). Although these laboratory-scale (typically a few meters in size) resonant bars were severely limited by thermal fluctuations, as well as their narrowband implementation, they spurred the development of both gravitational wave experiment and theoretical studies of sources, while no positive results have been obtained. In the mean time, indirect evidence of gravitational waves has been discovered in the famous Hulse-Taylor binary pulsar, whose orbital period shrinks according to the prediction of general relativistic gravitational radiation-reaction. Three more binary neutron star systems have been discovered since then that will merge in less than the current age of the universe, including a recent one in which both objects are pulsars. Through the past four decades, sensitivities of bar detectors have been improving gradually. Currently operative bar detectors include: Allegro (Louisiana State University, US), Auriga (Padova, Italy), Explorer (CERN), Nautilus (University of Rome, Italy) and Niobe (University of Western Australia).
Schematic Drawing of a Laser Interferometer Gravitational-Wave Detector
Starting in the 1970s, practical development of a different type of detectors, laser interferometric gravitational-wave detectors with kilometer-scale arm lengths (interferometers for short), was started by Forward (at Hughes Aircraft Research Labs in Malibu, California) and Weiss (at MIT), and by Drever (at University of Glasgow, Scotland). These detectors use laser interferometry to measure the strains induced by gravitational waves on mirror-endowed test masses hanging from seismic isolation stacks (see Figure). Despite the technological challenge and the cost of construction, these devices have the fundamental advantage provided by their long baseline that the displacement of the optics is proportional to the product of the gravitational-wave strain and the arm length. Optical interferometry also provides the possibilities of broadband and tunable-band operations, which dramatically increases their scientific potential with respect to resonant bar detectors. Kilometer-scale baselines have been further enhanced by inserting Fabry-Perot cavities into the interferometer arms, and by adopting more advanced optical techniques such as power and signal recycling.
Current status and future outlook of ground-based gravitational-wave detection
After three decades of development, an international network of long-baseline laser interferometric gravitational-wave detectors, consisting of LIGO in the USA, VIRGO (an Italian-French collaboration), GEO600 (a German-British collaboration, operated by AEI-Hannover and University of Glasgow) and TAMA300 (Japanese), are beginning operation. The LIGO interferometers have reached sensitivity of h ~ 10-21 at 150 Hz, a factor of ~2 away from its design goal at this frequency, and are expected to reach its design goal in one year. First coincident science runs between the LIGO and GEO600 interferometers have been completed, yielding initial upper-limit results. Coincident science runs between the LIGO and the TAMA interferometers have been performed as well. Virgo, now in commissioning phase, is also expected to begin GW observation runs within the year and make coincidence science runs soon after that. Astrophysical estimates suggest that it is plausible for first-generation interferometers to make a convincing detection over an operation of three years. In particular, each interferometer in initial LIGO is already able to detect neutron-star-binary inspirals up to 14 Mpc away (this is averaged over relative orientations of the source and the detector; optimally oriented, it can reach 22 Mpc; 1 Mpc = 3 million light years), with an expected rate of once per several years to once per several tens of years. It is generally believed that upgrades must be made before a rich program of observational gravitational-wave astronomy can be carried out.
An upgrade of LIGO (named Advanced LIGO) is planned for around 2008, as a joint project by institutions in the US, UK and Germany. Advanced LIGO is planned to improve the sensitivity by a factor of 15, by lowering various thermal noises and the seismic noise, and by increasing optical power; the signal-recycling optical technique, currently being used by GEO 600, is also part of the upgrade to be performed.
A factor of 15 increase in sensitivity implies an increase of detection volume by 153~3000, which makes Advanced LIGO plausible to detect compact-binary inspirals in a regular basis, to detect continuous waves from several Low-Mass X-ray Binaries or from known and unknown pulsars, and to detect possible gravitational waves from cosmological supernova explosions. Similarly Virgo is expected to undergo an equivalent, but more incremental, upgrade in a similar time scale. In addition, a second-generation detector, the Large-scale Cryogenic Gravitational-wave Telescope (LCGT), is planned in Japan and is waiting for funding.
A generation of detectors beyond is also currently being conceived in Euope and the US. The increasing sensitivity of gravitational-wave detectors promise a fruitful observational program of gravitational-wave astronomy.
Space-based projects and other detection methods
Ground-based interferometers are generally believed to be limited to frequencies higher than 1Hz, fundamentally by the Newtonian gravity-gradient noise, namely time-dependent Newtonian gravitational forces acting differently on the test masses, due to motions of matters in the surrounding area, including seismic waves, atmospheric fluctuations, and human activities.
To extend sensitivity to lower frequencies, a space-based interferometer, the Laser Interferometer Space Antenna (LISA), has been planned by ESA and NASA for the coming decade, with launch scheduled for 2012. LISA targets gravitational waves with frequencies from 0.1 mHz to 0.1 Hz, which can be generated by processes involving super-massive black holes in centers of galaxies. LISA uses three drag-free satellites forming an equilateral triangle with separation of five million kilometers, together this formation circles the sun with a period of 1 year. Laser beams link each pair of the spacecraft, providing flexible interferometric configurations. NASA has already financed a design study for space-based GW detectors beyond LISA. One proposal is the Big-Bang Observer (BBO), which covers the frequency range of 0.01 Hz to 1 Hz, and aims ultimately at detecting stochastic GW background originated from the big bang.
Other methods of detecting gravitational waves include doppler tracking of spacecraft, pulsar timing, and the measurement of anisotropy in the polarization of the Cosmic Microwave Background (CMB).
Further Readings
• Kip S. Thorne, Black Holes and Time Warps: Einstein's Outrageous Legacy, W.W. Norton & Company, New York, 1994. [German edition: Gekrümmter Raumund verbogene Zeit (Droemer Knaur, Munich, 1994.]
• Marcia Bartusiak, Einstein's Unfinished Symphony: Listening to the Sounds of Space-Time, Joseph Henry Press, Washington, DC, 2000.