Modern Physics · Advanced Topics

Gravitational Waves

Ripples in spacetime curvature propagating at c — predicted by Einstein in 1916, first directly detected by LIGO in 2015. Gravitational wave astronomy has opened a new observational window on black holes, neutron stars, and the Big Bang.

PrerequisitesGeneral relativity (Ch. GR) · Tensor calculus (Ch. TC) · Electromagnetic waves (Ch. W) · Special relativity (Ch. 21)
Learning Goals
  • Derive the linearized Einstein equations in Lorenz gauge and identify the two GW polarizations.
  • Apply the quadrupole formula to compute GW strain amplitude and radiated power for a binary system.
  • Extract the chirp mass from an observed df/dt and use it to estimate the source distance.
  • Explain the noise sources limiting LIGO sensitivity and how squeezed light surpasses the SQL.
  • Describe multi-messenger astronomy with GW170817 and how standard sirens constrain H0H_{0}.

GW.1 Linearized General Relativity

In the weak-field limit, write gμν=ημν+hμνg_{\mu\nu} = \eta_{\mu\nu} + h_{\mu\nu} with ∣hμν∣≪1|h_{\mu\nu}| \ll 1. The Einstein equations linearize. In Lorenz (harmonic) gauge ∂μhˉμν=0\partial^\mu \bar h_{\mu\nu} = 0(where hˉμν=hμν−12ημνh\bar h_{\mu\nu} = h_{\mu\nu} - \tfrac{1}{2}\eta_{\mu\nu} h is the trace-reversed perturbation):

□hˉμν=−16πGc4Tμν(linearized Einstein equations)\Box \bar h_{\mu\nu} = -\frac{16\pi G}{c^4} T_{\mu\nu} \qquad \text{(linearized Einstein equations)}(GW.1)

In vacuum (Tμν=0T_{\mu\nu} = 0), this is a wave equation: □hˉμν=0\Box \bar h_{\mu\nu} = 0, giving waves propagating at speed c. Residual gauge freedom → transverse-traceless (TT) gauge:hTTh^{\text{TT}} has only spatial components, hii=0h_{ii} = 0 (traceless), and kihij=0k^i h_{ij} = 0 (transverse). Two independent polarizations: plus (+) and cross (×).

Definition GW.1 — GW Polarizations
A GW propagating in the z-direction in TT gauge:hμν=h+eμν+cos⁡(kz−ωt)+h×eμν×cos⁡(kz−ωt+ϕ)h_{\mu\nu} = h_+ e^+_{\mu\nu}\cos(kz - \omega t) + h_\times e^\times_{\mu\nu}\cos(kz - \omega t + \phi)where e+=diag(0,1,−1,0)/2e^+ = \text{diag}(0, 1, -1, 0)/2 and e×e^\times has off-diagonal ±1/2 in xy. The plus polarization stretches x while compressing y (and vice versa), oscillating at fGWf_{\text{GW}} (twice the orbital frequency). The cross polarization does the same but rotated by 45°.

GW.2 Generation: The Quadrupole Formula

GWs are produced by changing quadrupole moments (not monopole/dipole — those are conserved by mass/momentum conservation). The leading-order emission:

h+,h×∼2Grc4 I¨ij(quadrupole formula, Iij=reduced quadrupole moment)h_+, h_\times \sim \frac{2G}{rc^4}\,\ddot I_{ij} \qquad \text{(quadrupole formula, } I_{ij} = \text{reduced quadrupole moment)}(GW.2)
PGW=G5c5⟨I...ij I...ij⟩(radiated power, three dots = third time derivative)P_{\text{GW}} = \frac{G}{5c^5}\left\langle \dddot I_{ij}\, \dddot I_{ij} \right\rangle \qquad \text{(radiated power, three dots = third time derivative)}(GW.3)

The factor G/c5=2.8×10−53 W−1G/c^5 = 2.8\times 10^{-53}\ \text{W}^{-1} makes GW emission extremely weak for laboratory sources. Only compact astrophysical objects (NS, BH) emit detectable GWs.

Theorem GW.1 — Inspiral and Merger
A binary system of masses m1,m2m_{1}, m_{2} (chirp mass Mc=(m1m2)3/5/(m1+m2)1/5\mathcal{M}_c = (m_1 m_2)^{3/5}/(m_1+m_2)^{1/5}) loses energy to GWs and spirals inward. The frequency evolves as:dfdt=965π8/3(GMcc3)5/3f11/3\frac{df}{dt} = \frac{96}{5}\pi^{8/3}\left(\frac{G\mathcal{M}_c}{c^3}\right)^{5/3} f^{11/3}The strain amplitude:h∼4r(GMcc2)5/3(πfc)2/3h \sim \frac{4}{r}\left(\frac{G\mathcal{M}_c}{c^2}\right)^{5/3}\left(\frac{\pi f}{c}\right)^{2/3}At merger (Schwarzschild radius contact): fISCO≈c363/2πGMtotal≈4400 Hz×M⊙Mtotalf_{\text{ISCO}} \approx \frac{c^3}{6^{3/2}\pi G M_{\text{total}}} \approx 4400\ \text{Hz} \times \frac{M_\odot}{M_{\text{total}}}.
Example GW.1 — GW150914 — First Direct Detection

LIGO detected GW150914 on September 14, 2015. Reconstruct key parameters from the observed signal: f sweeps 35→150Hz35 \to 150 Hz in ~0.2 s, peak strain h≈10−21h \approx 10^{-21}, estimated distance ~410 Mpc.

Chirp mass from df/dt:Atf≈100HzAt f \approx 100 Hz the measured local slope is df/dt≈3300Hz/sdf/dt \approx 3300 Hz/s (the sweep accelerates, so the average (150−35)/0.2≈575Hz/s150-35)/0.2 \approx 575 Hz/s badly underestimates it). Mc=(c3/G)×[(5/96)π−8/3f−11/3(df/dt)]3/5≈28.3M⊙M_{c} = (c^{3}/G) \times [(5/96) \pi^{-8/3} f^{-11/3} (df/dt)]^{3/5} \approx 28.3 M_\odot. This is the best-measured parameter from the inspiral phase (fractional uncertainty ~1%).
Mass scale from high-frequency cutoff:The observed high-frequency merger/ringdown near 150 Hz is not a clean Schwarzschild ISCO frequency, but it sets the compact-object mass scale. Detailed waveform matching gives initial black-hole masses ≈36M⊙\approx 36 M_\odot and 29M⊙29 M_\odot, with final mass ~62M⊙62 M_\odot. About 3M⊙c2≈5×1047J3 M_\odot c^{2} \approx 5\times 10^{47} J was radiated in GWs over ~0.2 s. Power: ~3×1048W—3\times 10^{48} W — briefly outshining all stars in the observable universe combined.
Distance from strain:h 4(GM)5/3(πf)2/3/(c4r)h ~ 4(GM)^{5/3}(\pi f)^{2/3}/(c^{4} r). Atf=100Hz,M=28M⊙:h≈10−21→r≈410MpcAt f = 100 Hz, M = 28 M_\odot: h \approx 10^{-21} \to r \approx 410 Mpc. LIGO arm length L=4kmL = 4 km; length change ΔL=hL/2≈10−21×4000m/2=2×10−18m=2am\Delta L = h L/2 \approx 10^{-21} \times 4000m/2 = 2\times 10^{-18} m = 2 am (attometers). 1/500 of a proton radius.
Significance:Signal-to-noise ratio SNR≈24SNR \approx 24. Detection significance: > 5.1σ5.1\sigma (false alarm rate < 1 per 200,000 yr). Confirmed as two black holes — first direct observation of a binary black hole merger, and first observation of black holes with masses M>20M⊙M > 20 M_\odot. Nobel Prize in Physics 2017 to Weiss, Barish, Thorne.

GW.3 LIGO Interferometer

LIGO uses a modified Michelson interferometer with 4 km arms. A GW stretches one arm and squeezes the other: ΔL = h L/2.

Noise sources (from high to low frequency): 1. Seismic noise (below ~10 Hz): ground vibrations. Mitigated by 4-stage pendulum isolation. 2. Thermal noise (10–200 Hz): Brownian motion of mirror coatings and suspensions. Fused silica fibers, low-loss coatings (SiO₂/Ta₂O₅), low temperature prototypes. 3. Quantum noise (above ~100 Hz): photon shot noise. Standard quantum limit:hSQL=1L8ℏmω2h_{\text{SQL}} = \frac{1}{L}\sqrt{\frac{8\hbar}{m\omega^2}} — trades off shot noise against radiation pressure noise. Overcome by squeezed light injection (LIGO O3: ~3 dB of squeezing).

Power recycling: mirror between laser and BS reflects light back → builds up 200 kW of intracavity power (from 20 W laser input). Signal recycling: mirror at dark port tunes the detector's frequency response. Advanced LIGO sensitivity (O4): h ~ 3×10⁻²⁴/√Hz at 100 Hz, horizon distance ~200 Mpc for binary neutron star mergers.

GW.4 Sources and Multi-Messenger Astronomy

Binary black holes (BBH): most numerous detections (~100+ by O3 end). No EM counterpart (BH mergers don't produce photons).Binary neutron stars (BNS): GW170817 (2017) — first BNS detection. Simultaneous gamma-ray burst GRB170817A detected 1.7 s after merger by Fermi. Multi-messenger observation confirmed: NS mergers = short GRBs + kilonovae (r-process nucleosynthesis — gold, platinum produced).

Hubble constant from GW: "standard siren" — GW gives absolute distance (no distance ladder), EM gives redshift. GW170817: H0=70−8+12H_0 = 70^{+12}_{-8} km/s/Mpc. With more events: will resolve Hubble tension model-independently.

Pulsar timing arrays (PTAs): millisecond pulsars as a GW detector. Nanohertz GW background (f ~ 1–100 nHz) from supermassive black hole binaries. First evidence (NANOGrav 2023, ~3–4σ) for GW background — new frequency window.LISA (2030s): space-based, 2.5 Mkm arms, millihertz band — targets SMBH mergers, extreme mass-ratio inspirals, stochastic GW background from inflation.

Definition GW.2 — Common Traps
  • Gravitational waves are strain waves: detectors measure fractional length change, not a force meter reading.
  • Monopole and dipole radiation are absent: gravitational radiation begins at quadrupole order.
  • Polarizations are transverse tidal patterns: plus and cross describe stretching directions.
  • Signal frequency tracks orbital motion: inspiral chirps upward as the orbit shrinks.
Exercises — GW.1–GW.4 Gravitational Waves
1.Describe how the plus polarization of a gravitational wave deforms a ring of test masses. What is the antenna pattern F_+ for a Michelson interferometer?
Straightforward
2.
Derive the gravitational wave luminosity (Peters formula) for an equal-mass circular binary. Compute the merger timescale for the Hulse-Taylor pulsar.
Myr
Intermediate
3.
Explain how binary neutron star mergers serve as "standard sirens" for measuring H0H_{0}. Use GW170817 data (DL=40Mpc,z=0.0099D_{L} = 40 Mpc, z = 0.0099) to estimate H0H_{0} and compare to other measurements.
km/s/Mpc
Intermediate
4.Derive the standard quantum limit (SQL) for a gravitational wave detector by balancing shot noise and radiation pressure noise. How does squeezed light injection help LIGO surpass the SQL?
Challenging
Key Takeaways
  • Linearized GR:□hˉμν=−16πG/c4TμνGR: \Box \bar{h}_\mu\nu = -16\pi G/c^{4} T_\mu\nu. TT gauge: two polarizations h+,hh_+, h_× propagating at c.
  • Quadrupole formula: h (2G/rc4)×d2I/dt2h ~ (2G/rc^{4}) \times d^{2}I/dt^{2}. Power P (G/5c5)(⋯I)2—P ~ (G/5c^{5})(\cdots I)^{2} — extremely weak for lab sources.
  • Inspiral: chirp mass M from df/dt. Strain h (GM)5/3/(c4r)h ~ (GM)^{5/3}/(c^{4} r). Merger at fISCO≈4400Hz×M⊙/Mtotalf_{ISCO} \approx 4400 Hz \times M_\odot/M_{\text{total}}.
  • GW150914: BBH merger, 3M⊙c23 M_\odot c^{2} radiated, ΔL 2am(2×10−18m).>5.1σ\Delta L ~ 2 am (2\times 10^{-18} m). > 5.1\sigma. Nobel 2017.
  • LIGO noise: seismic (<10 Hz), thermal (10-200 Hz), quantum (shot + radiation pressure). SQL 1.8×10−24/HzSQL ~ 1.8\times 10^{-24} /\sqrt Hz at 100 Hz.
  • GW170817 (BNS): confirmed NS mergers = short GRBs + kilonovae. H0=70km/s/MpcH_{0} = 70 km/s/Mpc as "standard siren".