Astronomy › Section R · Black holes and curved spacetime › Topic R.5
Gravitational-wave astronomy
When black holes or neutron stars spiral together, they shake spacetime itself. Since 2015, detectors able to sense a change far smaller than a proton have been hearing these ripples, opening a completely new way to observe the universe.
Ripples in spacetime
General relativity says mass curves spacetime. When masses accelerate, for example two stars whirling around each other, the curvature changes, and the changes spread outwards as gravitational waves, travelling at the speed of light, \(c = 3.00 \times 10^{8}\) m/s. A passing wave alternately stretches space in one direction while squeezing it at right angles, then the reverse. The effect is described by the strain, \(h = \frac{\Delta L}{L}\), the fractional change in a length. Even for the most violent events, the strain reaching Earth is only about \(10^{-21}\), which is why Einstein, who predicted the waves in 1916, doubted they could ever be detected.
The first clue: a shrinking orbit
Indirect evidence came first. As "Tests of relativity and time in gravity" described, the orbit of the Hulse-Taylor binary pulsar, discovered in 1974, shrinks year by year at exactly the rate predicted if the pair is losing energy as gravitational waves. That convinced most physicists the waves are real. Detecting them directly needed sources far more violent: the final moments before two neutron stars or black holes merge, when they orbit each other hundreds of times a second at a sizeable fraction of the speed of light. Possible sources also include core-collapse supernovae and the very early universe, though neither has yet been detected.
Detectors made of light
The Laser Interferometer Gravitational-Wave Observatory, LIGO, has two sites in the United States, at Hanford, Washington, and Livingston, Louisiana. Each is an L-shaped instrument with two 4 km vacuum arms. Laser light bounces between mirrors hung at the ends, then the two beams are recombined. A passing wave lengthens one arm while shortening the other, changing how the beams combine. Rainer Weiss proposed the design in 1972, and Weiss, Barry Barish and Kip Thorne shared the 2017 Nobel Prize in Physics for LIGO. Virgo, with 3 km arms near Pisa in Italy, and KAGRA, underground in Japan, joined the network, and a detector in India is being built.
Worked example: how small is the signal?
A gravitational wave with strain \(h = 1.0 \times 10^{-21}\) passes through one of LIGO's 4.0 km arms. How much does the arm's length change? 1. Rearrange the strain: \(\Delta L = hL\). 2. Substitute:
3. Compare: a proton is about \(1.7 \times 10^{-15}\) m across, so the change is about \(\frac{1.7 \times 10^{-15}}{4.0 \times 10^{-18}} \approx 400\) times smaller than a proton. 4. Why two sites: earthquakes, traffic and other local shaking affect each site differently, while a real wave appears at both, within the light-travel time between them.
The first detection
On 14 September 2015 both LIGO detectors recorded the same signal, now called GW150914. It was a chirp: a wave whose frequency and strength rose over a fraction of a second as two black holes of about 36 and 29 solar masses (one solar mass, \(M_\odot\), is about \(1.99 \times 10^{30}\) kg) spiralled together, merged and rang down into a single black hole of about 62 solar masses. The merger happened about 1.3 billion light-years away, so the waves had travelled for about 1.3 billion years. The signal matched computer solutions of Einstein's equations closely, a striking test of general relativity in the strongest gravity yet probed.
Worked example: the energy of GW150914
In GW150914, black holes of about 36 and 29 solar masses formed one of about 62 solar masses. How much energy did the waves carry, and what does the timing say? 1. Mass lost: \(36 + 29 - 62 = 3\) solar masses. 2. Energy:
3. That is about 5% of the starting mass, radiated mostly in a fraction of a second, briefly out-powering all the stars in the observable universe. 4. Timing: the sites are about 3,000 km apart, so a wave at speed \(c\) needs at most \(\frac{3.0 \times 10^{6}}{3.00 \times 10^{8}} = 0.010\) s to cross between them. The signal reached Livingston about 7 ms before Hanford, within that limit, and the delay helped show where in the sky it came from.
Neutron stars collide
On 17 August 2017 LIGO and Virgo detected GW170817, two neutron stars of about 1.1 to 1.6 solar masses spiralling together, a signal lasting about 100 seconds. With three detectors the source could be located, and about 1.7 seconds after the merger the Fermi satellite recorded a short gamma-ray burst from the same direction. Telescopes then found a kilonova in the galaxy NGC 4993, about 130 million light-years away, glowing with freshly made heavy elements. The event confirmed, as "Binary star evolution and gamma-ray bursts" described, that such mergers cause short bursts. The near-simultaneous arrival also showed that gravitational waves travel at the speed of light to about one part in a thousand trillion.
Hundreds of mergers
Detections quickly became routine. The first three observing runs, from 2015 to 2020, found about 90 events. The fourth run, from May 2023 to 18 November 2025, was by far the richest: by late 2025 the collaboration's catalogue held 218 confident detections, with about 170 more candidates from the end of the run still being analysed, roughly 400 in all. Most are black-hole pairs; a few involve neutron stars, including black hole-neutron star pairs first reported in 2021. The heaviest so far, GW231123, announced in 2025, formed a black hole of about 225 solar masses. Many black holes seen are heavier than those in X-ray binaries, and how they formed is debated. A six-month interim run was planned to begin around November 2026.
Other wavebands for gravity
LIGO-type detectors hear waves of tens to thousands of cycles per second. Slower waves need bigger instruments. In 2023 pulsar timing arrays, which watch dozens of millisecond pulsars for tiny shared shifts in their timing, announced evidence for a background of waves with periods of years, probably from pairs of supermassive black holes in distant galaxies. In space, the European Space Agency adopted the LISA mission in 2024: three spacecraft millions of kilometres apart, planned for launch in the mid-2030s, to hear merging supermassive black holes. Together these will make gravitational waves a full branch of astronomy, like the electromagnetic spectrum.
Common mistakes to avoid
First, confusing gravitational waves with gravity waves, ripples in water or air driven by gravity. Second, thinking gravitational waves travel faster than light: GW170817 showed they travel at c. Third, assuming detectors measure a large movement: the change is far smaller than a proton. Fourth, thinking one detector can locate a source: several are needed to triangulate it. Fifth, assuming every merger is seen in light: black-hole pairs usually give no light, while neutron-star mergers can.
Key takeaways
Gravitational waves are ripples in spacetime from accelerating masses, travelling at c and measured by strain, h = ΔL/L, about 10⁻²¹ at Earth. The shrinking Hulse-Taylor pulsar orbit gave indirect proof. LIGO's 4 km interferometers, with Virgo and KAGRA, measure changes far smaller than a proton. GW150914 in 2015 was the first detection: two black holes merging and radiating about 3 solar masses as waves. GW170817 in 2017 linked a neutron-star merger to a gamma-ray burst and a kilonova. By late 2025 there were roughly 400 detections and candidates; pulsar timing arrays and LISA extend the search to slower waves.
Flashcards
Flashcard
A ripple in spacetime from accelerating masses
Test your understanding
Question 1 of 5
A gravitational wave of strain 2.0 × 10⁻²¹ passes through a detector arm 3.0 km long. By how much does the arm's length change?
Question 2 of 5
In a merger, black holes of 30 and 25 solar masses form one of 52 solar masses. How much mass was converted into gravitational waves?
Question 3 of 5
Why does LIGO need two widely separated detectors?
Question 4 of 5
What did GW170817 show that the first black-hole detections could not?
Question 5 of 5
Which statement about gravitational-wave detections is correct as of late 2025?
