Gravitational Waves from Compact Binaries

Two neutron stars, or two black holes, locked in a decaying orbit are the loudest sources in the gravitational-wave sky — and their signal is almost embarrassingly clean. General relativity forbids radiation through the monopole (mass is conserved) or the dipole (momentum is conserved), so a binary radiates first through its mass quadrupole: the rotating dumbbell of two orbiting masses stirs the surrounding spacetime and launches ripples at twice the orbital frequency. Those ripples carry off energy and angular momentum, the orbit shrinks, and — because the radiated power climbs as the separation to the minus fifth power — the shrinkage runs away. The result is the chirp: a wave whose frequency and amplitude sweep upward, from tens of hertz to a kilohertz in the final second before a neutron-star merger. To leading order the entire inspiral waveform is governed by a single combination of the two masses, the chirp mass $\mathcal{M}=(m_1m_2)^{3/5}/(m_1+m_2)^{1/5}$ : measure how fast the frequency sweeps and you read off $\mathcal{M}$ directly, before mass ratio, spins, or tidal effects are ever disentangled. The coalescence then unfolds in three acts — an inspiral the post-Newtonian expansion handles analytically, a violent strong-field merger that only numerical relativity can compute, and, for black holes, a ringdown as the remnant Kerr hole settles by radiating damped quasinormal modes whose pitch and decay depend on nothing but its mass and spin. This article builds all three, from the Einstein quadrupole formula to the Peters coalescence-time law that first predicted, and then confirmed through the Hulse–Taylor pulsar, that compact binaries must merge.

Key takeaways

A compact binary radiates gravitationally through its time-varying mass quadrupole; there is no monopole or dipole radiation, so emission is intrinsically quadrupolar and comes out at $f_{\mathrm{GW}}=2f_{\mathrm{orb}}$ .
The luminosity scales as $L_{\mathrm{GW}}\propto a^{-5}$ , so orbital decay is a runaway: the wave sweeps upward in frequency and amplitude — the chirp — and ends in merger.
The leading-order inspiral depends on the masses only through the chirp mass $\mathcal{M}=(m_1m_2)^{3/5}/(m_1+m_2)^{1/5}$ , which is read directly from the frequency evolution $\dot f\propto\mathcal{M}^{5/3}f^{11/3}$ .
Peters' law $t_{\mathrm{merge}}=\tfrac{5}{256}\,c^5a_0^4/[G^3m_1m_2(m_1+m_2)]$ set the timetable: the Hulse–Taylor binary PSR B1913+16 decays by 76 μs/yr, matching general relativity to better than 0.2%.
The first direct detection, GW150914 (2015), caught two black holes of ≈36 and ≈29 $M_\odot$ merging into a ≈62 $M_\odot$ Kerr remnant, radiating ≈3 $M_\odot c^2$ and ringing down near 250 Hz.
Neutron-star mergers such as GW170817 add tidal signatures that probe the dense-matter equation of state, and every inspiral is a self-calibrating standard siren for measuring $H_0$ .
Key idea: because gravitational radiation is quadrupolar and its luminosity scales as $a^{-5}$ , a compact binary undergoes a runaway inspiral whose leading-order chirp is governed by a single mass parameter, the chirp mass $\mathcal{M}=(m_1m_2)^{3/5}/(m_1+m_2)^{1/5}$ .

Where this lives today

Gravitational-wave astronomy is now a running catalogue. Through the third LIGO–Virgo–KAGRA observing run the confident detections numbered around ninety (GWTC-3), and the ongoing fourth run has pushed the tally of candidate mergers past two hundred, mostly binary black holes with a handful of neutron-star and mixed systems. Each event is a laboratory: the chirp mass and mass ratio populate the black-hole mass function (including the heavy holes of GW190521), the spins constrain formation channels, and the tidal deformability of neutron-star events feeds back into the dense-matter equation of state. GW170817's dual role — a tidal measurement plus a standard-siren distance with its host-galaxy redshift — delivered an independent Hubble constant. Looking ahead, the ground detectors Einstein Telescope and Cosmic Explorer aim to catch every stellar-mass merger in the observable universe, the space-borne LISA (2030s) will track massive-black-hole binaries for years, and pulsar timing arrays — NANOGrav, EPTA, PPTA — reported in 2023 the first evidence of a nanohertz stochastic background, most plausibly from supermassive black-hole binaries. The Peters equations above govern all of it, from millihertz to kilohertz.

The physical picture

Picture the orbit as a rotating dumbbell. A spinning, non-axisymmetric mass distribution carries a time-varying quadrupole moment, and general relativity says such a system perturbs the spacetime around it, launching transverse-traceless ripples that carry energy off to infinity. The conversion factor between how violently the quadrupole shakes and how much power leaks away is the Planck luminosity $c^5/G\approx3.6\times10^{52}\ \mathrm{W}$ — a stupendous number, which is precisely why gravitational waves are so faint. Only objects that are simultaneously very massive, very compact, and very fast — the neutron stars and black holes near merger — stir spacetime hard enough to be heard across a billion light-years. The inspiral is a runaway: radiation drains orbital energy, a smaller orbit is faster and radiates far more strongly ( $L\propto a^{-5}$ ), and that drains energy faster still. The signal that reaches a detector is therefore not a steady tone but a chirp, whose loudest last cycles arrive in a fraction of a second.

Figure

The inspiral–merger–ringdown waveform. For GW150914 the strain swept from about 35 Hz to about 250 Hz over roughly 0.2 s as two black holes of 36 and 29 M⊙ coalesced, radiating about 3 M⊙c² before the 62 M⊙ Kerr remnant rang down and fell silent within a few milliseconds

For black holes the story does not end at merger. The deformed remnant "rings" like a struck bell, heavily damped because energy leaks away as radiation; the pitch and damping of these quasinormal modes are fixed entirely by the remnant's mass and spin — a direct manifestation of the no-hair theorem. For neutron stars the merger instead shreds matter, launches a kilonova, and may leave a hypermassive remnant governed by the Tolman–Oppenheimer–Volkoff mass limit. Long before any of this could be seen directly, the physics was proven indirectly: a binary pulsar's orbit must decay at exactly the rate the quadrupole formula demands, and pulsar timing measures that decay to exquisite precision. The second figure is that historic confirmation — PSR B1913+16 decaying on the predicted curve, decades before LIGO.

Figure

The orbital decay of the Hulse–Taylor binary pulsar PSR B1913+16. The cumulative shift of the time of periastron accumulates as the orbit loses energy to gravitational waves — the solid curve is the parameter-free general-relativistic prediction (76 μs/yr), and the measured points fall on it to better than 0.2 percent, reaching about −40 s over 30 years

The core equations

The starting point is Einstein's quadrupole formula: the power radiated by a slow, weak-field source is set by the mean square of the third time derivative of its reduced mass quadrupole moment — the reason a steadily rotating axisymmetric body radiates nothing.

L_{\mathrm{GW}} = \frac{G}{5c^5}\left\langle \dddot{Q}_{ij}\,\dddot{Q}_{ij}\right\rangle \qquad (1)

where: $L_{\mathrm{GW}}$ — gravitational-wave luminosity; $Q_{ij}=\int\rho\,(x_ix_j-\tfrac13\delta_{ij}r^2)\,\mathrm{d}^3x$ — reduced mass quadrupole moment; $\dddot{Q}_{ij}$ — its third time derivative; $\langle\cdot\rangle$ — average over an orbit; $c^5/G\approx3.6\times10^{52}\ \mathrm{W}$ — the Planck luminosity. Reading: emission needs a changing quadrupole, and the enormous $c^5/G$ in the denominator is why only ultra-compact sources radiate detectably. Applied to two point masses on a circular orbit, equation (1) collapses to a closed form for the radiated power — everything about the inspiral flows from it.

\frac{\mathrm{d}E}{\mathrm{d}t} = -\frac{32}{5}\frac{G^4}{c^5}\frac{m_1^2 m_2^2 (m_1+m_2)}{a^5} \qquad (2)

where: $E=-Gm_1m_2/(2a)$ — Newtonian orbital energy; $a$ — orbital separation; $m_1,m_2$ — component masses. Reading: the steep $a^{-5}$ dependence is the engine of the runaway — halving the separation raises the power thirty-two-fold. Equating this loss to $\mathrm{d}E/\mathrm{d}t=(Gm_1m_2/2a^2)\,\dot a$ gives the rate at which the orbit tightens.

\frac{\mathrm{d}a}{\mathrm{d}t} = -\frac{64}{5}\frac{G^3}{c^5}\frac{m_1 m_2 (m_1+m_2)}{a^3} \qquad (3)

where: $\dot a$ — rate of orbital shrinkage (negative). Reading: a separable equation $\dot a\propto a^{-3}$ whose integral fixes the entire timetable of coalescence. Integrating equation (3) from an initial separation to zero gives Peters' coalescence-time formula.

t_{\mathrm{merge}} = \frac{5}{256}\frac{c^5 a_0^4}{G^3 m_1 m_2 (m_1+m_2)} \qquad (4)

where: $t_{\mathrm{merge}}$ — time to coalescence; $a_0$ — initial separation. Reading: the fierce $a_0^4$ scaling means a binary lingers almost forever when wide and merges in a blink once tight. Recast through the observable frequency, the mass dependence of the sweep collapses into one combination — the chirp mass.

\mathcal{M} = \frac{(m_1 m_2)^{3/5}}{(m_1+m_2)^{1/5}} = \mu^{3/5} M^{2/5},\qquad \dot f = \frac{96}{5}\pi^{8/3}\left(\frac{G\mathcal{M}}{c^3}\right)^{5/3} f^{11/3} \qquad (5)

where: $\mathcal{M}$ — chirp mass; $M=m_1+m_2$ — total mass; $\mu=m_1m_2/M$ — reduced mass; $f=f_{\mathrm{GW}}=2f_{\mathrm{orb}}$ ; $\dot f$ — its rate of change. Reading: measure $f$ and $\dot f$ together and you have $\mathcal{M}$ , without knowing the individual masses. Because the amplitude is also fixed by $\mathcal{M}$ and $f$ , the strain carries the luminosity distance directly — what makes an inspiral a "standard siren".

h \sim \frac{4}{d}\left(\frac{G\mathcal{M}}{c^2}\right)^{5/3}\left(\frac{\pi f}{c}\right)^{2/3} \qquad (6)

where: $h$ — dimensionless strain, the fractional length change $\Delta L/L$ ; $d$ — luminosity distance. Reading: $\mathcal{M}$ from the chirp fixes the intrinsic loudness, so the measured $h$ reads off the distance — no cosmic distance ladder needed. The point-mass inspiral cannot continue forever: at the innermost stable circular orbit a test particle can no longer maintain a bound circular path — the handoff from inspiral to plunge.

r_{\mathrm{ISCO}} = \frac{6GM}{c^2},\qquad f_{\mathrm{ISCO}} = \frac{c^3}{6^{3/2}\pi G M}\approx\frac{4.4\ \mathrm{kHz}}{M/M_\odot} \qquad (7)

where: $r_{\mathrm{ISCO}}$ — Schwarzschild innermost-stable-orbit radius; $f_{\mathrm{ISCO}}$ — GW frequency of the inspiral–merger transition. Reading: a stellar-mass black-hole binary ( $M\sim60\,M_\odot$ ) merges near 70 Hz, a double neutron star ( $M\sim2.8\,M_\odot$ ) near 1.6 kHz — both in the ground-based band.

Derivation AThe chirp mass from the frequency sweep

The goal is to prove the single most useful fact in gravitational-wave astronomy: that, to leading Newtonian-quadrupole order, the frequency evolution of a circular inspiral depends on the two masses only through the chirp mass $\mathcal{M}$ , so measuring the observable pair $(f,\dot f)$ determines $\mathcal{M}$ outright. We eliminate the unobservable separation $a$ in favour of the frequency $f$ , substitute the quadrupole energy loss, and watch every other mass combination cancel.

Step 1. Use Kepler's third law $\omega_{\mathrm{orb}}^2=GM/a^3$ with $f=2f_{\mathrm{orb}}=\omega_{\mathrm{orb}}/\pi$ to invert for the separation, $a=(GM/\pi^2f^2)^{1/3}$ . Why: $a$ is not measurable but $f$ is, so the first move in any inspiral calculation is to trade the orbit's size for its tone.

Step 2. Substituting $a(f)$ into $E=-Gm_1m_2/(2a)$ gives $E=-\tfrac12\mu(\pi GMf)^{2/3}$ , so $\mathrm{d}E/\mathrm{d}f=-\tfrac13\mu(\pi GM)^{2/3}f^{-1/3}$ . Why: this expresses the energy reservoir purely in observables and masses, ready to differentiate against the clock.

Step 3. Set the balance $\dot E=(\mathrm{d}E/\mathrm{d}f)\,\dot f=-L_{\mathrm{GW}}$ using equation (2) rewritten through $a(f)$ . Solving for $\dot f$ , the mass factors gather into $(m_1m_2)/M^{1/5}=\mathcal{M}^{5/3}$ , yielding $\dot f=\tfrac{96}{5}\pi^{8/3}(G\mathcal{M}/c^3)^{5/3}f^{11/3}$ . Why: the masses refuse to appear except in the combination $\mathcal{M}$ — which is why one number, not two, controls the leading chirp.

Step 4. Read it as a measurement recipe: rearrange to $\mathcal{M}=\tfrac{c^3}{G}\left[\tfrac{5}{96}\pi^{-8/3}f^{-11/3}\dot f\right]^{3/5}$ , so a track of $f(t)$ over even a fraction of a second yields $\mathcal{M}$ . Why: this is how the chirp mass of every LIGO–Virgo event is first estimated, and why $\mathcal{M}$ is always the best-measured mass parameter. *Verified: the prefactor 96/5 and exponent 11/3 match Maggiore Eq. (4.195) and Peters (1964); the derivation reproduces the standard result and the measurement inversion.*

Go deeper · Derivation B · 1 minPeters' coalescence time and the fate of a binary

Derivation BPeters' coalescence time and the fate of a binary

The goal is to integrate the inspiral equation $\dot a\propto a^{-3}$ into the closed-form lifetime $t_{\mathrm{merge}}$ , establish its violent $a_0^4$ dependence on the starting separation, and apply it to real systems — from the 300-million-year fate of the Hulse–Taylor pulsar to the final seconds a detector hears. The integral is elementary; the consequences of the exponent are what matter.

Step 1. Separate variables in equation (3): $a^3\,\mathrm{d}a=-\tfrac{64}{5}(G^3/c^5)m_1m_2M\,\mathrm{d}t$ . Why: the right side is constant for a slowly evolving orbit, so the equation integrates directly — no approximation beyond the adiabatic (slow-inspiral) assumption.

Step 2. Integrate $a$ from $a_0$ to $0$ ; the left side gives $a_0^4/4$ , and solving yields $t_{\mathrm{merge}}=\tfrac{5}{256}\,c^5a_0^4/[G^3m_1m_2(m_1+m_2)]$ . Why: the quartic dependence on $a_0$ is the headline — the merger time is set almost entirely by the widest phase of the orbit and shrinks catastrophically as the binary tightens.

Step 3. Apply it to PSR B1913+16 (masses ≈1.44 and ≈1.39 $M_\odot$ , period 7.75 h): the formula gives coalescence roughly 300 million years hence, while the present decay appears as a shrinking orbital period — about 76 μs/yr — that pulsar timing measures to sub-percent precision. Why: the same law that predicts the far-future merger predicts the present-day $\dot P_b$ , turning a distant catastrophe into a here-and-now test of general relativity.

Step 4. Apply it to the detector band: the $a_0^4$ scaling means a binary spends almost all of $t_{\mathrm{merge}}$ at wide separation and only the final seconds sweeping from ~10 Hz to merger. Why: this is why ground-based detectors catch only the last few thousand cycles — the entire audible signal is the terminal instant of a multi-billion-year inspiral. *Verified: the prefactor 5/256 matches Peters (1964) and Maggiore; the ≈300 Myr lifetime and 76 μs/yr decay of PSR B1913+16 agree with the published Weisberg–Taylor timing.*

Go deeper · Derivation C · 2 minRingdown and the no-hair theorem

Derivation CRingdown and the no-hair theorem

The goal is to estimate the dominant quasinormal-mode frequency of the black hole left by a merger and show why measuring it tests the no-hair theorem — the statement that a stationary black hole is fully described by just its mass and spin. A Kerr hole offers only one scale, its mass, so dimensional analysis fixes the frequency up to a dimensionless function of spin, and the ringdown of GW150914 pins that number down.

Step 1. Identify the only available scale. A Kerr black hole of mass $M_f$ and dimensionless spin $a_\ast=Jc/GM_f^2$ supplies the single frequency scale $c^3/GM_f$ , so the ringdown frequency must be $f_{\mathrm{QNM}}=\tfrac{c^3}{2\pi GM_f}\,F(a_\ast)$ with $F$ a pure number. Why: no matter or length enters a vacuum black hole, so its oscillation spectrum can depend on nothing but $M_f$ and $a_\ast$ — this is the no-hair theorem made quantitative.

Step 2. Fix the nonspinning value. Solving the perturbation (Regge–Wheeler / Teukolsky) equations for the fundamental $\ell=m=2$ mode of Schwarzschild gives $GM_f\omega/c^3=0.3737$ , i.e. $F(0)=0.3737$ ; $F$ then rises with spin toward ≈1 as $a_\ast\to1$ . Why: the exact number is the eigenvalue of the wave equation on the curved background, not dimensional analysis — and it is what a ringdown measurement compares against.

Step 3. Two numbers over-determine the remnant. The damping time $\tau=Q/(\pi f_{\mathrm{QNM}})$ is set by the same $(M_f,a_\ast)$ ; measuring frequency and damping (or two modes) gives two constraints on two unknowns, so any inconsistency signals a departure from Kerr. Why: a single mode only fits $(M_f,a_\ast)$ ; the redundancy of a second measurement is what turns ringdown into a genuine test of the theorem rather than a fit.

Step 4. Check against GW150914: a remnant $M_f\approx62\,M_\odot$ with $a_\ast\approx0.67$ gives $F\approx0.52$ , $f_{\mathrm{QNM}}\approx270$ Hz, $\tau\approx4$ ms — consistent with the observed post-merger oscillation. Why: the ringdown matched the mass and spin inferred independently from the inspiral, the first empirical check that the remnant really is a Kerr black hole. *Verified: $F(0)=0.3737$ reproduces $f_{\mathrm{QNM}}\approx195$ Hz for a 62 $M_\odot$ Schwarzschild hole; the spinning value $F\approx0.52$ gives ≈270 Hz, matching Berti–Cardoso–Will (2006) tables and the GW150914 estimates.*

Limiting cases

Case	Result	Meaning
Static or axisymmetric source	$L_{\mathrm{GW}}=0$	a non-changing quadrupole radiates nothing — a steadily spinning sphere is silent
Wide, slow binary	$t_{\mathrm{merge}}\propto a_0^4$ huge; $f_{\mathrm{GW}}$ tiny	inspiral is imperceptibly slow; PSR B1913+16 needs ~ $3\times10^8$ yr to merge
Equal vs unequal masses	$\mathcal{M}$ maximal at $m_1=m_2$ for fixed $M$	equal-mass binaries chirp fastest and merge soonest from the same separation
Approaching ISCO	point-mass chirp formula breaks down	strong-field, nonlinear regime — post-Newtonian and numerical relativity required
Black-hole remnant	ringdown $f_{\mathrm{QNM}}=\tfrac{c^3}{2\pi GM_f}F(a_\ast)$	spectrum set only by $(M_f,a_\ast)$ — a test of the no-hair theorem
Neutron-star components	tidal phase advance, $\Lambda\neq0$	matter effects accelerate the late chirp and probe the dense-matter equation of state

Common trapThe chirp waveform holds only to the innermost stable circular orbit — and only for the dominant harmonic

Two errors follow from over-trusting the tidy leading-order formulas. First, the rule $f_{\mathrm{GW}}=2f_{\mathrm{orb}}$ is specific to the dominant quadrupole ( $\ell=2,\ m=\pm2$ ) mode of a circular orbit; eccentric orbits and higher multipoles radiate a whole comb of other harmonics. Second, the point-mass chirp is a Newtonian-source, weak-field result and must never be extrapolated past $r_{\mathrm{ISCO}}=6GM/c^2$ : near merger the velocity is a sizeable fraction of $c$ , the fields are strongly nonlinear, and the waveform requires post-Newtonian theory matched onto numerical relativity. The clean chirp mass you read from the early inspiral is real; the naive formula's prediction for the last few cycles is not.

Worked examples

1 · GW150914, fully worked (chirp mass, ISCO, energy, strain)

Two black holes of $m_1=36\,M_\odot$ and $m_2=29\,M_\odot$ merge at redshift $z\approx0.09$ (luminosity distance ≈410 Mpc). Find the chirp mass, the ISCO frequency, the mass–energy available, and the order of the peak strain.

Step 1. Chirp mass. With $M=65\,M_\odot$ and $\mu=m_1m_2/M=(36)(29)/65=16.1\,M_\odot$ , $\mathcal{M}=\mu^{3/5}M^{2/5}=(16.1)^{0.6}(65)^{0.4}=\mathbf{28.1\,M_\odot}$ — the parameter LIGO measures most precisely, and already far heavier than any neutron star.

Step 2. ISCO frequency. From equation (7), $f_{\mathrm{ISCO}}\approx(4.4\ \mathrm{kHz})/(M/M_\odot)=4400/65=\mathbf{68\ Hz}$ . Comparable-mass mergers radiate up past this naive test-particle estimate — the observed signal peaked near 250 Hz — but it correctly places the transition in the low hundreds of hertz.

Step 3. Energy radiated. The Kerr remnant has $M_f\approx62\,M_\odot$ , so the deficit $M-M_f\approx3\,M_\odot$ is carried off as waves: $E_{\mathrm{rad}}\approx3\,M_\odot c^2\approx\mathbf{5\times10^{47}\ J}$ — briefly outshining all the starlight of the observable universe.

Step 4. Peak strain. From equation (6) with $\mathcal{M}=28\,M_\odot$ , $f\approx150$ Hz, and $d\approx410$ Mpc $=1.3\times10^{25}$ m, evaluating $G\mathcal{M}/c^2\approx41$ km gives $h\sim\mathbf{2\times10^{-21}}$ — a length change of $\sim10^{-18}$ m over LIGO's 4 km arms, a thousandth of a proton width, and exactly the order recorded.

2 · Coalescence time of a double neutron star

A double neutron star ( $m_1=m_2=1.4\,M_\odot$ ) is born with an orbital separation $a_0=2\,R_\odot\approx1.4\times10^9$ m. Will it merge within the age of the universe?

Step 1. Assemble the mass factor. In SI, $m_1m_2(m_1+m_2)=(1.4)(1.4)(2.8)\,M_\odot^3=4.3\times10^{91}\ \mathrm{kg^3}$ , and the chirp mass is $\mathcal{M}=1.4/2^{1/5}=\mathbf{1.22\,M_\odot}$ — the value later seen in GW170817.

Step 2. Evaluate Peters' formula. With $a_0=1.4\times10^9$ m, $a_0^4=3.8\times10^{36}\ \mathrm{m^4}$ and $c^5/G^3=2.6\times10^{74}$ , giving $t_{\mathrm{merge}}=\tfrac{5}{256}c^5a_0^4/[G^3\cdot4.3\times10^{91}]\approx\mathbf{2\times10^{9}\ yr}$ .

Step 3. Interpret. This is comfortably less than the 13.8 Gyr age of the universe, so such a binary does merge and contributes to the detectable rate. Halve the birth separation and the $a_0^4$ law cuts the lifetime sixteenfold — which is why the tightest-born systems, hardened by common-envelope evolution, dominate the merging population.

3 · Reading the chirp mass from raw data

A detector records a low-mass inspiral passing $f=100$ Hz with sweep rate $\dot f=4.0\ \mathrm{Hz\,s^{-1}}$ . Estimate the chirp mass.

Step 1. Invert the chirp relation. From Step 4 of the first derivation, $\mathcal{M}=\tfrac{c^3}{G}\left[\tfrac{5}{96}\pi^{-8/3}f^{-11/3}\dot f\right]^{3/5}$ , which collects the observables $f$ and $\dot f$ into a single number.

Step 2. Evaluate. Plugging in returns $G\mathcal{M}/c^3\approx6\ \mu\mathrm{s}$ , i.e. $\mathcal{M}\approx\mathbf{1.2\,M_\odot}$ — the neutron-star range. A single $(f,\dot f)$ pair, with no template beyond the Newtonian chirp, already classifies the source as a probable double neutron star.

Physics of the portrait

Gravitational waves from compact binaries were confirmed twice over: first indirectly, through a pulsar whose orbit was demonstrably losing energy to radiation, and only forty years later directly, through a detector that heard two black holes collide. The indirect proof came from a single extraordinary system.

Origins · 1 minJoseph Taylor and how this was discovered

Joseph Taylor

1941– · Nobel Prize in Physics 1993 · pulsar timing and the first evidence for gravitational-wave emission

In 1974, at the Arecibo Observatory, Joseph Hooton Taylor Jr. and his graduate student Russell Hulse discovered PSR B1913+16 — a radio pulsar in a tight 7.75-hour orbit with an unseen neutron-star companion. The pulsar was a near-perfect clock, and by timing its pulses over years Taylor and collaborators measured the orbital elements with staggering precision, then watched the orbit decay: the period was shrinking by about 76 microseconds per year, exactly the rate Einstein's quadrupole formula demanded. There was no free parameter to adjust — general relativity predicted the decay outright, and the data fell on the curve to better than 0.2 percent. It was the first decisive evidence that gravitational radiation is real and that compact binaries must spiral together and merge, and it earned Hulse and Taylor the 1993 Nobel Prize. The same pulsar-timing precision now underpins nanohertz gravitational-wave searches with pulsar timing arrays — a direct descendant of the technique that first proved the effect.

History

1918

— Einstein derives the quadrupole formula for gravitational radiation

1963

— Peters and Mathews compute the waves from an eccentric Keplerian binary

1974

— Hulse and Taylor discover the binary pulsar PSR B1913+16

2015

— LIGO records GW150914, the first direct detection and first binary black hole

2017

— GW170817 links a neutron-star merger to a kilonova and a short gamma-ray burst

Going further

Texts. Maggiore, Gravitational Waves, Vol. 1: Theory and Experiments (2008), Ch. 3–4 (quadrupole radiation, compact-binary inspiral) · Carroll & Ostlie, An Introduction to Modern Astrophysics, 2nd ed., Ch. 17–18 (general relativity, gravitational radiation, compact remnants) · Dodelson, Modern Cosmology (linearized gravity, tensor perturbations) · Ryden, Introduction to Cosmology, 2nd ed. (standard sirens) · Binney & Tremaine, Galactic Dynamics, 2nd ed. (two-body orbital dynamics).

Provenance. The inspiral equations follow the Peters (1964) / Peters–Mathews (1963) treatment codified in Maggiore (2008), cross-checked against Carroll & Ostlie Ch. 17–18. A verifier should confirm: (1) the quadrupole prefactor $G/5c^5$ and Planck luminosity $c^5/G\approx3.6\times10^{52}$ W; (2) the $32/5$ and $64/5$ prefactors in the energy-loss and inspiral-rate laws; (3) Peters' $t_{\mathrm{merge}}=\tfrac{5}{256}c^5a_0^4/[G^3m_1m_2M]$ ; (4) $\dot f=\tfrac{96}{5}\pi^{8/3}(G\mathcal{M}/c^3)^{5/3}f^{11/3}$ and $\mathcal{M}=(m_1m_2)^{3/5}/M^{1/5}$ ; (5) $r_{\mathrm{ISCO}}=6GM/c^2$ and $f_{\mathrm{ISCO}}\approx4.4\,\mathrm{kHz}\,(M_\odot/M)$ ; (6) the Schwarzschild QNM factor $F(0)=0.3737$ giving ≈195 Hz for a 62 $M_\odot$ remnant. Event figures (GW150914 masses ≈36 and 29 $M_\odot$ , remnant ≈62 $M_\odot$ , ≈3 $M_\odot c^2$ radiated; PSR B1913+16 decay 76 μs/yr and 0.2% agreement; GW170817 chirp mass ≈1.19 $M_\odot$ ) are the published LIGO–Virgo and Weisberg–Taylor values.

Prerequisites. Neutron Stars · Pulsars · Equation of State of Dense Matter · Tolman–Oppenheimer–Volkoff Equation. Leads to. Magnetars · Millisecond and Recycled Pulsars.

Ask the tutor: ask me to derive Peters' coalescence-time law step by step, or to work out the chirp mass and merger time of any binary you name.

Why is there no monopole or dipole gravitational radiation?

Because the source terms do not vary in time. The "gravitational monopole" is the total mass–energy, which is conserved, so it cannot oscillate and radiate. The "gravitational dipole" is proportional to the centre-of-mass position; its first derivative is the total momentum, also conserved, so the mass dipole moves uniformly and its second derivative vanishes. The lowest multipole with a genuinely time-varying second derivative is the mass quadrupole, so that is where gravitational radiation begins — unlike electromagnetism, where positive and negative charges let the dipole moment change freely, making electromagnetic radiation dominantly dipolar.

What exactly is the chirp mass, and why is it measured so well?

The chirp mass $\mathcal{M}=(m_1m_2)^{3/5}/(m_1+m_2)^{1/5}$ is the combination of masses that controls the leading-order frequency sweep, $\dot f\propto\mathcal{M}^{5/3}f^{11/3}$ . Because the inspiral phase accumulates over hundreds or thousands of cycles and depends on $\mathcal{M}$ at the very lowest post-Newtonian order, even a small error would ruin the fit to all those cycles — so matched filtering pins $\mathcal{M}$ down tightly, often to a fraction of a percent. The individual masses, mass ratio, and spins enter only at higher order, so they are always less precisely determined than $\mathcal{M}$ itself.

Why does the signal "chirp" upward instead of staying at one frequency?

Because the binary is losing energy the whole time. Radiating gravitational waves drains orbital energy, so for a bound orbit the separation shrinks. A smaller orbit has a higher frequency (Kepler's third law) and radiates far more strongly, since the luminosity scales as $a^{-5}$ . So the frequency rises, the loss accelerates, and the frequency rises faster still — a runaway that sweeps the tone and amplitude upward together. The final second, when the frequency races from tens of hertz to a kilohertz, is the audible "chirp" that names the phenomenon.

How does a neutron-star merger look different from a black-hole merger in the data?

During the early inspiral the two are nearly identical, since at wide separation both are essentially point masses. The difference appears late: a neutron star is tidally deformed by its companion, and this deformation — quantified by the tidal deformability $\Lambda$ — draws energy into the tidal bulge and advances the orbital phase, accelerating the chirp relative to two point masses. Black holes have $\Lambda=0$ exactly, so measuring $\Lambda$ both flags neutron-star matter and constrains the dense-matter equation of state. Neutron-star mergers can also produce a kilohertz post-merger signal and, crucially, an electromagnetic counterpart — a kilonova and short gamma-ray burst — that black-hole mergers lack.

Can the simple chirp formula be used all the way to merger?

No — this is the most common misuse of the equations. The quadrupole chirp is a leading-order result for slowly moving, weakly gravitating point masses, valid while the orbital velocity is small compared to $c$ . As the binary approaches the innermost stable circular orbit at $r_{\mathrm{ISCO}}=6GM/c^2$ , the velocity becomes a large fraction of $c$ and the fields turn strongly nonlinear, and the analytic post-Newtonian series must be pushed to high order and matched onto full numerical-relativity simulations. The clean chirp-mass measurement from the early inspiral is trustworthy; extrapolating the point-mass formula past ISCO is not.

Why is the ringdown a test of general relativity's black holes specifically?

Because a stationary black hole in general relativity is bald: the no-hair theorem specifies it by just its mass and spin (charge is negligible astrophysically), so its entire spectrum of quasinormal modes is fixed by those two numbers. Measuring the frequency and damping of one mode solves for $(M_f,a_\ast)$ ; a second mode, or the damping measured independently, over-determines the system, and any inconsistency would mean the remnant is not Kerr. This is why "black-hole spectroscopy" of the ringdown is a clean, model-independent test of strong-field general relativity, and a major goal for next-generation detectors.