Observations of neutron stars and pulsars extend over more than 19 decades of the electromagnetic spectrum, from low radio frequencies (around 30 MHz) to high gamma-ray energies (above 200 GeV). The techniques used in telescopes between these extremes range from the coherent detection of radio waves to photon detection techniques more usually associated with nuclear physics. There are nevertheless elements in common over the whole range, which we will refer to in this brief survey.

1. (1) The signal is weak, requiring large collecting areas and long integration times.

2. (2) Identification of objects requires accurate positions and discrimination from adjacent sources.

3. (3) Pulsed sources require high timing accuracies, often around 1 microsecond.

4. (4) Measurements must discriminate against unwanted backgrounds, either of astronomical origin, such as radio emission or cosmic rays from the Milky Way Galaxy, or from terrestrial sources, especially man-made radio signals.

The terrestrial atmosphere is transparent to radio waves (except at short millimetric wavelengths where molecular absorption occurs, and at long metric wavelengths where ionospheric refraction and reflection occur). Radio telescopes can therefore be built at ground level, and can extend in size almost indefinitely, giving both high sensitivity and high angular resolution. X-rays and gamma-rays are absorbed in the atmosphere, and direct detection of such high-energy photons can only be achieved using space-based telescopes, where telescope apertures are limited by the capabilities of launch vehicles to a few metres in diameter.

Over the whole electromagnetic spectrum from radio to the very high energy gamma-rays detected by air shower telescopes, pulsars radiate beams giving pulse profiles which are unique signatures differing from pulsar to pulsar. For some individual pulsars, notably the Crab Pulsar described in Chapter 9, very similar profiles are observed over the whole spectrum. The majority of pulsars are observable only in radio; here the individual pulses are often very variable and it is by averaging the pulse shapes of many, sometimes hundreds, of pulses that the characteristic shape is seen.

The advent of the Fermi LAT satellite telescope and the air shower Cerenkov detectors HESS and VERITAS has opened a new window for observing the high-energy beamed radiation from pulsars, and the wind nebulae. Many young pulsars, and especially those with large spin-down energy, are observable with the LAT, along with many of the millisecond pulsars. In the high-energy regimes the radiation is detected as individual photons, arriving so infrequently that integration over many millions of pulse periods is needed before the profile emerges.

These integrated radio and high-energy profiles are the key to understanding the geometry and the physical processes within the magnetosphere. In this chapter we start with the radio profiles, which have provided an astonishingly wide range of phenomena, both in their detailed shapes and in their variations over various time scales.

The surveys

A glance through the catalogue of known pulsars shows at once that they are mostly found in the Milky Way. The normal pulsars show the clearest concentration towards the plane of the Galaxy, while the millisecond pulsars, most of which can only be detected at smaller distances, are more isotropic. The normal pulsars must therefore be young Galactic objects, and it might be assumed that their distribution through the Galaxy is similar to that of young massive stars and supernovae. Although this is nearly correct, it can only be established by reading the catalogue in conjunction with a description of the surveys in which the pulsars were found; many of these surveys in fact concentrated on the plane of the Galaxy, giving an obvious bias to the catalogue, while others show considerable variations of sensitivity over the sky.

The first surveys to cover large areas of the sky were comparatively insensitive, and necessarily gave rather meagre evidence. For example, Large & Vaughan (1971) found only 29 pulsars in 7 steradians of the southern sky. Nevertheless this catalogue, combined with a northern hemisphere catalogue covering low Galactic latitudes (Davies, Lyne & Seiradakis 1973) showed that there must be at least 105 active pulsars in the Galaxy.

There are now nearly 2000 known normal pulsars, over half of which have been discovered in surveys carried out at frequencies near 1.4 GHz (see Chapter 3). The entire sky has now been surveyed to a reasonably well calibrated flux density limit, both for normal pulsars for millisecond pulsars, while surveys with greater sensitivity cover low Galactic latitudes.

Astrophysics provides many examples of rotating and orbiting bodies whose periods of rotation and revolution can be determined with great accuracy. Within the Solar System the orbital motion of the planets can be timed to a small fraction of a second, while the rotation of the Earth is used as a clock that is reliable to about one part in 108 per day. Outside the Earth there is, however, no other clock with a precision approaching that of pulsar rotation.

The arrival times of the radio pulses from pulsars are easy to study, and a surprising amount can be learned from them. Not only do they provide information on the nature of the pulsed radio source, they can also give an accurate position for the source; and they can explore the propagation of the pulses through the interstellar medium. All three kinds of information were noted by Hewish and his collaborators in the discovery paper of 1968. They showed that the shortness of the pulses, and their short and precise periodicity, implied that the source was small, and that it might be a rotating neutron star. They showed also that the pulse period was varying because of the Doppler effect of the Earth's motion round the Sun; this annual variation implied that the source lay outside the Solar System. Finally, they showed that the arrival time of a single pulse depended on radio frequency; this dispersion effect was found to be in accord with the effect of a long journey through the ionised gas of interstellar space.

The spectrum of thermal radiation from a neutron star with surface temperature of order 106 K peaks in the X-ray spectrum at a photon energy around 1 keV. The first observation of an X-ray source outside the Solar System, made in 1962 using a rocket-borne instrument (Giacconi et al. 1962), revealed an unexpected and powerful source, designated Sco X-1. The explanation of this source was given by Shklovsky in 1967; it is indeed a thermal source, but it is accreting matter in a hot circumstellar disc surrounding a neutron star in a binary system. Sco X-1 is now the prototype of a class of binary X-ray sources known as Low Mass X-ray Binaries (LMXBs).

Confirmation of the nature of Sco X-1 and other X-ray sources in the Galaxy revealed by the first X-ray astronomy satellite UHURU (launched in 1970) came when the source Cen X-3 was shown to be pulsating with a period of 4.8 seconds. Following the same arguments as in the interpretation of the binary radio pulsars, it soon became clear that the source must be a rapidly rotating neutron star in a binary system. The orbital periods are typically several days, indicating that the binary systems are close enough for mass transfer to occur.

Single jets are seen to protrude from quasars as well as the cores of powerful radio galaxies. There is no physical principle which precludes intrinsically asymmetric systems, although a collapsed spinning object might be expected to possess a high degree of mirror symmetry. On the other hand, a system of merging binary black holes would certainly not. Nevertheless, the appearance of jets as one-sided led to the hypothesis that intrinsically symmetric twin jets contain radiating plasma that moves at relativistic speeds. The flux of one jet is Doppler-boosted for an observer lying within an emission cone centred on the jet direction, while the flux of the receding jet is reduced.

The discovery of knots of synchrotron emission seen to move in the sky away from quasar cores at speeds exceeding the speed of light turned into convincing evidence for the relativistic interpretation. The so-called superluminal motion was recorded in many quasar jets in the 1970s. The radio galaxy 3C 120 (z = 0.033) was observed at two epochs which suggested a speed of two to three times the speed of light (Shaffer et al., 1972). The quasar 3C 345 was observed at four epochs in 1974 and 1975. The apparent transverse speed was reported as eight times the speed of light (Cohen et al., 1976). As a result, an interpretation described in this chapter based on highly relativistic jet flows at a small angle to the line of sight became accepted.

By late 2010 five large-scale laser interferometer gravitational wave detectors had been operating for several years at unprecedented sensitivity. They were searching for gravitational wave signals created by matter in its most extreme and exotic form – neutron stars, black holes and the Big Bang itself. The detectors were the most sensitive instruments ever created, able to detect fractional changes in spacetime geometry at the level of parts in 1023, corresponding to the measurement of energy changes of less than 10-31 joules per hertz of bandwidth. Despite this extraordinary achievement, the sensitivity was about 10 times below the level where we could be confident of detecting predicted signals. For example, the mean time between detectable chirp signals from the coalescence of pairs of neutron stars was likely to be once every 50 years, so that in a year of operation the chance of detection was only about 2%.

Despite this pessimistic prognosis, many of the 1000 physicists in the worldwide collaborations involved with the above detectors remained optimistic that nature might to kind enough to provide a first signal. Optimism was high enough that a system had been put in place to alert optical telescopes to slew to the part of the sky corresponding to the arrival times of any significant event.

On 16 September 2010 a coincident signal appeared in LIGO detectors spaced 2000 kilometres apart in the USA. It was immediately recognised as a significant event, especially after it was also identified in the data of the Virgo detector in Italy.

The suppression of seismic ground motion is a significant challenge for ground based gravitational wave interferometric detectors. It is generally achieved through different combinations of active isolation and passive isolation. This chapter will look at two different vibration isolation systems. Part 1 will describe the mostly active vibration isolation system used by LIGO. Part 2 will discuss the mostly passive isolation systems developed at the University of Western Australia, which have features in common with isolation systems developed for many other detectors.

Seismic isolation for Advanced LIGO

Lantz for the LIGO Scientific Collaboration

One of the most significant improvements for Advanced LIGO will be to move the lower edge of the detection band from 40 Hz down to 10 Hz. This will allow us to start tracking the final inspiral of compact binary systems earlier in their evolution, and will also allow us to observe the inspirals not only of pairs of neutron stars, but also of binary systems containing more massive objects, such as black holes with 10 to 30 times the mass of the Sun, as described in Chapter 4. In order to measure gravitational waves at 10 Hz, the requirements for the motion of the Advanced LIGO mirrors are very strict. The Advanced LIGO interferometer cannot distinguish between differential length changes in the arms caused by the spacetime distortion of a passing gravitational wave and differential length changes caused by mirror motion. For an interesting discussion about interferometer configurations where this might not be true, see recent work by Kawamura and Chen (2004) and others.

This chapter first introduces gravitational wave detection from a very general point of view, before looking at the particular methods of detection across the spectrum from nanohertz to kilohertz. It finishes by focusing specifically on terrestrial laser interferometers.

Introduction

The discovery of radio waves by Heinrich Hertz in 1886 unleashed the communications revolution which has transformed our lives. Optimisation of radio receivers required understanding and integration of two concepts. The first was the concept of the antenna, which taps energy from a wave freely propagating in space and converts it into a signal which can be amplified and detected. The second was the receiver, which processes this energy by detection (converting it to a slowly time-varying voltage), amplification (increasing its amplitude without changing its frequency) or modulation (changing its frequency).

Designing gravitational wave receivers is analogous to designing radio receivers, except that electric charges moving freely in conductors are replaced by test masses floating freely in space. This concept was illustrated in Figure 1.2 in Chapter 1, showing how a ring of test particles is deformed by a passing gravitational wave. The first gravitational wave receivers were constructed by Joseph Weber in the 1960s. They took the form of large test masses in which gravitational waves could induce quadrupole vibrations. Weber went on to develop the Weber bar, in which one searched for excitations in the fundamental longitudinal vibrational mode of a cylinder. In this case, the receiver can be idealised as a pair of point masses joined by a mechanical spring.

This chapter features the USA-based LIGO, the Laser Interferometer Gravitational-Wave Observatory – the first of three case studies covering different worldwide interferometric gravitational wave detectors. In addition to describing the basic interferometer operation and its various components, we discuss the technological challenges that have been overcome for its successful operation.

Introduction

The prediction of gravitational waves (GWs), oscillations in the spacetime metric that propagate at the speed of light, is one of the most profound differences between Einstein's general theory of relativity and the Newtonian theory of gravity that it replaced. As discussed in Chapter 1, GWs remained a theoretical prediction for more than 50 years until the first observational evidence for their existence came with the discovery and subsequent observations of the binary pulsar PSR 1913+16, by Russell Hulse and Joseph Taylor (Weisberg and Taylor, 2005). In about 300 million years, the PSR 1913+16 orbit will decrease to the point where the pair coalesces into a single compact object, a process that will produce directly detectable gravitational waves. In the meantime, the direct detection of GWs will require similarly strong sources – extremely large masses moving with large accelerations in strong gravitational fields. The goal of LIGO, the Laser Interferometer Gravitational-Wave Observatory (Abramovici et al., 1992), is just that: to detect and study GWs of astrophysical origin. Achieving this goal will mark the opening of a new window on the Universe, with the promise of new physics and astrophysics. In physics, GW detection could provide information about strong-field gravitation, the untested domain of strongly curved space where Newtonian gravitation is no longer even a poor approximation.

The data from widely spaced networks of gravitational wave detectors can be combined to act as a single detector with optimum sensitivity and directional resolution. We first provide a basic mathematical framework and characterise the detection capacity and relative performance of various networks. A systematic approach is then provided for the construction of network detection statistics, stable waveform extraction, null-stream construction, and source localisation. At the end, wediscuss the angular resolution of an arbitrary detector network and issues relevant to the field of multi-messenger gravitational wave astronomy.

Introduction

As discussed in the previous chapters, several large-scale interferometric gravitational wave (GW) detectors have reached or exceeded their design sensitivity, and have been coordinating to operate as a global array. These include the LIGO detectors at Louisiana and Washington states of the USA, the Virgo detector in Italy and the GEO600 detector in Germany. The three US-based interferometric GW detectors LIGO have completed their ground-breaking fifth science run in November 2007. An integrated full year's worth of data has been accumulated with all three interferometers in coincidence. Advanced LIGO will enable a 10-fold improvement in sensitivity, allowing detectors to monitor a volume of the Universe 1000 times larger than can be achieved by current detectors. These advanced detectors are predicted to detect tens to hundreds of events per year (see Abadie et al., 2010). The detection of the first GW signal is virtually assured with Advanced LIGO.

Many experiments in modern physics set highly demanding requirements on their laser light sources. Precision metrology, laser cooling experiments and the quantum engineering of atoms and molecules are some example areas in which very stable lasers are indispensable. The first generation of interferometric gravitational wave detectors have been among the laser applications with the most challenging requirements, simultaneously requiring low fluctuations in power, frequency and beam pointing as well as high power levels of 10W. Laser sources for second generation gravitational wave detectors need to fulfill even more demanding requirements, which we will discuss in the first section of this chapter. The second section is devoted to the design of lasers for advanced detectors followed by a section in which we discuss their stabilisation. The last section covers some laser concepts for third generation gravitational wave detectors.

Requirements on the light source of a gravitational wave detector

One of the fundamental noise sources of laser interferometric gravitational wave detectors directly related to the laser light is the shot noise in the interferometer readout. The ratio of a potential gravitational wave signal to the readout shot noise is proportional to the square root of the light power in the interferometer. Hence, gravitational wave detectors need high-power lasers in combination with resonant optical cavities to achieve high circulating power levels in the interferometer. First generation detectors use lasers with approximately 10Wlight power and second generation instruments will require power levels of order 200W. In general, increasing the light power in the interferometer improves the sensitivity until the noise introduced by fluctuating radiation pressure forces on their mirrors reaches the same level as the readout shot noise.

This chapter describes the Virgo interferometric gravitational wave detector. We first discuss the overall detector design before describing the individual subsystems in detail. We finish by outlining the commissioning and upgrades required to achieve second generation sensitivity.

Introduction

The gravitational wave detector Virgo1, funded by CNRS (France) and INFN (Italy), is a recycled Michelson interferometer with arms replaced by 3 km long Fabry–Perot cavities. Virgo is located at the European Gravitational Observatory (EGO), close to Cascina (Pisa, Italy), and is the only ground-based antenna which, from its conception, aimed at reducing the detection band lower threshold down to 10 Hz. This goal was achieved by isolating the interferometer mirrors from ground seismic noise which, in the tens of hertz range, is many orders of magnitude larger than the small variation of the arm length induced by the gravitational wave passage.

The construction of Virgo started in the second half of 1990s and, after a commissioning phase of several years, the interferometer performed two scientific runs (VSR1, lasting four months in 2007, and VSR2, from July 2009 to January 2010), in coincidence with the antennae of the LIGO project.

Once the antenna is isolated from ground seismic noise, thermal displacements induced by pendulum dissipations in the mirror suspensions (pendulum thermal noise) limit the sensitivity up to a few tens of hertz. At higher frequency, the thermal noise induced by dissipation inside the mirror (mirror thermal noise) is dominant, up to a few hundreds of hertz. Above this frequency, the antenna sensitivity is mainly suppressed by the shot noise in the photo detection.

The detection of gravitational waves is sometimes described as the Holy Grail of Modern Physics. This is somewhat of a misnomer. Like the search for the holy grail, the search has appeared endless and fruitless, especially to non-scientific observers who cannot believe that it could take so long to make a detector, test it and come up with a firm answer. But unlike the search for the holy grail, physicists know that gravitational waves exist, not only from the beauty and elegance of Einstein's General Theory which predicts their existence, but also from the observations of binary pulsar systems which lose energy exactly in accordance with the theoretical predictions. This work by Joseph Taylor was rewarded with the 1993 Nobel Prize in physics.

The saga of gravitational wave detection goes back a long way: Einstein believed they existed but thought they were not physically detectable. Eddington queried their existence: he suggested that ‘they travel at the speed of thought’. But in the 1950's Pirani, Feynman, Bondi and later Isaacson proved their physical reality, and in about 1960 Joseph Weber began to develop his famous resonant mass detectors. One now resides in the Smithsonian museum and another at one of LIGO's gravitational wave observatories. About 1970 his claims of detection (which turned out to be false) fired up a whole community.

This chapter introduces the different classes of gravitational wave sources targeted by terrestrial and space-based detectors. The possibility and implications of gravitational wave emissions from supernovae and coalescing binary systems of neutron stars and/or black holes are discussed, as well as the possible connection between gravitational wave sources and gamma-ray bursts. The chapter also discusses continuous gravitational wave sources and describes how a stochastic gravitational wave background could be produced from astrophysical sources or from events in the early Universe.

Introduction

Astrophysics provides us with a variety of candidate systems which should be observable in the spectrum of gravitational waves. However, it is important to remember that our powers of prediction of new phenomena are limited, so any list of sources is almost certain to be incomplete.

Amongst stellar mass systems we expect detectable gravitational radiation from the formation of black holes and neutron stars (Fryer et al., 2002), and from the coalescence of binary neutron stars and final collapse of such binaries to form a black hole (Phinney, 1991). We would expect not only discrete sources, but also continuous stochastic backgrounds created from large numbers of discrete sources. In our Galaxy the very large populations of binary stars create a stochastic background in the 10-3 to 10-5 Hz range (Hils et al., 1990; Cutler, 1998; Nelemans et al., 2001).