James Hough outlines the last 30 years of gravitational-wave astronomy, from building prototype detectors to making a revolutionary discovery
“Gravitational waves, yet to be convincingly detected, promise to open a new astronomical window.” Those are the words I wrote for Physics World in early 1989. Today, with six detections of gravitational waves confirmed over the past three years, I am delighted to see how many of the predictions I made in that article have come to fruition.
Almost exactly a century after they were predicted by Albert Einstein in his general theory of relativity, the first detection of gravitational waves – produced via the collision and subsequent merger of two black holes – was made by the Laser Interferometer Gravitational-wave Observatory (LIGO) detectors in the US, on 14 September 2015. Since then, four more black hole coalescences have been reported. Although the initial observation took all of us completely by surprise, it was a much-awaited discovery. These observations provide the first direct proof that black holes exist; that they can be in binary orbits; and that there is a family of black holes of tens of solar masses, which were not thought to exist.
To add to the excitement, we have now also detected gravitational waves and gamma rays produced by two coalescing neutron stars. The observation was made possible once the newly upgraded Virgo detector in Italy joined forces with LIGO in August 2017. Not only did these detectors witness the spectacular merger of the two ultracompact dead stars, but more than 70 other telescopes and arrays world over picked up the electromagnetic radiation the event produced. Such emissions are known to be associated with a “kilonova” – a type of event that may be responsible for producing a large fraction of the heavy elements, such as gold and platinum, in our universe. Observing such a source so soon after Virgo joined the search was amazing for those of us who had been waiting so long to see anything.
Looking back
So how did we get from the situation I described 30 years ago to where we are now? Well, it is an intriguing and sometimes tortuous story.
During the 1980s, prototype interferometric gravitational-wave detectors were being developed by teams in Garching, Germany; in Glasgow, Scotland; and at Caltech and the Massachusetts Institute of Technology (MIT) in the US. During the following decade, detector sensitivities were achieved at a level such that, if scaled up to much longer baselines, they would potentially allow the detection of signals from coalescing compact binary systems. Improving detector sensitivity was a real struggle, but I recall how it was spurred on by a strong spirit of competition between the research groups involved.
Long-baseline detectors compare the lengths of two arms at right angles to one another, in a Michelson-type interferometer arrangement, with each arm holding a large mirror (hung as a pendulum) at its far end. A measurement is made by splitting a laser beam between the two arms, and comparing the phase of the returning light, as it bounces back from the mirrors. As a gravitational wave of a particular polarization propagates normal to the plane of the detector, one arm will initially increase in length and the other decrease and vice versa. Longer arms create a more sensitive instrument because gravitational waves cause a strain in space–time, and so the distance change between the mirrors is greater the further apart they are. In addition the arms are artificially lengthened by forming Fabry–Perot cavities – resonant optical cavities – in each arm.
Early in 1983 a study for possible long-baseline instruments in the US was published on the instigation of Rainer Weiss, Kip Thorne and Ronald Drever (last year, Weiss and Thorne, together with Barry Barish, won the Nobel Prize for Physics for their efforts in building LIGO). Further proposals for such instruments were worked up in Germany and Italy starting in 1985; while a year later, the UK planned a detector based in Scotland. By 1989 progress on all fronts was encouraging and prototype sensitivities continued to improve. Research groups in Germany and the UK were also encouraged by their respective funding agencies to stop competing and join together, to submit a single proposal for a detector system to be located in either Germany or Scotland.
Based on the earlier US study, and bolstered by results from the prototype experiments, a construction proposal for the LIGO interferometers was submitted by Caltech and MIT to the US National Science Foundation. This proposal was for three long-baseline interferometers: two with arms 4 km long, and one with 2 km arms. One 4 km detector and the 2 km instrument were to be in the same vacuum system – subsequently constructed at Hanford in Washington State – while the other 4 km instrument, in its own vacuum system, was constructed in Louisiana.
Change afoot
Policy-makers in Germany and the UK were also seriously discussing funding for such efforts, with the Science and Engineering Research Council (SERC) in the UK considering stumping up cash for the first stages of construction. But in 1990 two significant events put an end to the possibilities of a detector in Scotland. First, a new chairperson was appointed to SERC who was not a supporter of gravitational-wave research – he found a “black hole”, not in the universe, but in the finances of the council. This meant that there would not be any UK funding for a long-baseline detector. Second, reunification occurred in Germany. The former East Germany was far behind what had been West Germany in terms of investment and scientific infrastructure. In order to start reinvigorating research in the former East, the priority for science funding moved to setting up new institutes and attracting academics to this part of Germany.
This was a significant blow to me and my colleagues, particularly in Scotland. We had fought hard to get planning permission for a detector site in Tentsmuir Forrest, near the former Royal Air Force base at Leuchars on the east coast of Scotland. The opposition from locals was very vocal as many thought we were planning a new secret weapons facility, while others thought the lasers would harm the wildlife. Persuading the relevant authorities that neither was the case had been challenging and fun; but it seemed it had all been for nothing.
It wasn’t all plain sailing for our US colleagues either, as internal difficulties with the project management were causing major problems. However, by 1994 these issues with LIGO had been resolved. Luckily there was also a change in the research funding structure in the UK, with the formation of the Particle Physics and Astronomy Research Council, and gravitational waves were once more considered for funding. The situation had also improved in Germany, and under the leadership of Karsten Danzmann, a new joint German–UK detector called “GEO 600” was conceived, and construction began at Ruthe near Hannover.
Novel features
The LIGO detectors in the US were designed to have 4 km long arms, while the Virgo detector in Italy – the brain child of Alain Brillet and Adalberto Giazotto – was to have 3 km long arms. However, both financial and site restrictions limited the arm length of the German–UK detector to 600 m, meaning that GEO 600 would have further challenges to overcome for it to compete with the longer detectors. To do this, we incorporated some new features into the device: we used a technique known as “signal recycling”, which allowed tunable signal enhancement over its operating band. The device also used fused-silica fibres rather than metal wires to suspend the silica mirrors, to reduce thermal noise effects, particularly at low frequency; as well as lasers of advanced design.
Both LIGO and GEO 600 were meant to begin operating in early January 2001. It was an exciting but stressful time for those of us in GEO, as we were having some “technical difficulties” with the detector, and felt our honour was at stake. Just when it seemed as though LIGO would begin before us, an earthquake in the US damaged one of its detectors, allowing us to catch up. Soon after, LIGO and GEO began operating (joined by Virgo in 2003) and ran until 2010. As is now well known, no gravitational-wave signals were detected during that entire observational period. This was not unexpected though, and indeed the original LIGO proposal had acknowledged that sensitivity enhancements would have to be made at a later time for there to be a real chance of success.
It was decided that both LIGO detectors would be upgraded, and would adopt the advanced technologies already trialled by GEO – giving birth to the experiment’s current incarnation, the “Advanced LIGO” (aLIGO) system. More sophisticated active seismic isolation systems were developed in the US; the suspension systems for the silica mirrors, using silica fibres, were supplied by the UK; the lasers were made in Germany; while parts of the optics were fashioned in Australia.
The aLIGO interferometers at Hanford and Livingston – both with two 4 km arms, and no 2 km interferometer – became operational in September 2015. The revolutionary first detection was made almost immediately – in fact, while the detectors were still being calibrated. Thanks to the new seismic isolation, and lower thermal noise from the suspensions, aLIGO had achieved the sensitivity needed to make this first discovery in 2015. The upgraded Virgo detector, meanwhile, came online in August 2017 – just in time to enable better sky location of the colliding neutron stars, allowing us to confirm that the collision was indeed the source of the short gamma-ray burst.
Looking ahead
Both LIGO and Virgo will be further enhanced to continue improving their sensitivity in the coming years. A new detector, KAGRA, in the Kamioka mine in Japan is also expected to join the network in 2019, while a version of the aLIGO detector will soon be under construction in India. What we are experiencing is the opening up of a new area in observational astronomy – that of gravitational multimessenger astronomy, which has enormous potential to help us better understand how our universe formed and evolved, and what it is today.
Much more is to be discovered, and to do so, our detectors will also extend into space, with the Laser Interferometer Space Antenna (LISA) – a joint mission between the European Space Agency and NASA that is expected to launch in around 2035. This space-based gravitational-wave observatory will consist of three identical drag-free spacecraft – each holding proof mass mirrors – placed at the vertices of a virtual equilateral triangle in space, with sides 2.5 million kilometres long, in a heliocentric orbit.
A new cosmic messenger
The relative lengths of the “arms” – the distance between each craft – will be monitored by laser interferometry. LISA will extend the frequency range of observations down to the sub-millihertz regime (a frequency region not accessible to ground-based detectors). It should therefore allow us to study, among other areas, the interactions of supermassive black holes in merging galaxies, as well as make it possible to track some coalescing systems from very low frequency, through to the higher frequencies detected here on Earth.
Third-generation ground-based detectors with longer baselines and higher sensitivity – such as the Einstein Telescope in Europe and Cosmic Explorer in the US – are also being proposed. Together, LISA and the Earth-bound observatories should allow us to pick up gravitational signals from the majority of the known universe.
The last three decades have led us to what has been a remarkable and groundbreaking time in astronomy, and I believe that the next three will hold a wealth of new science, with many surprises still to come.
