The physics of gravitational waves

The notes aim to provide a first-principles, physics-focused treatment of gravitational waves within general relativity, suitable for graduate students. Motivated by the rapid growth of observational data following the first direct detections by LIGO-Virgo (and later KAGRA), the purpose is to build solid theoretical foundations—propagation, generation, energy content, detector response, and data analysis—rather than cataloging astrophysical sources. Core questions include: what are the true gravitational degrees of freedom, how do GWs propagate on flat and curved backgrounds, how are they generated by sources (e.g., binaries), how is their energy flux defined in GR, how do detectors respond, and how are signals extracted statistically from noisy data.

The work synthesizes classical results and modern formulations: linearized GR, Lorenz/harmonic gauge, and the scalar–vector–tensor decomposition and Bardeen gauge-invariant variables; post-Newtonian theory (Poisson & Will; Will), standard PN gauge; the Landau–Lifshitz pseudotensor and relaxed Einstein equations; Regge–Wheeler and Zerilli formalisms for Schwarzschild perturbations; Teukolsky’s separation for Kerr perturbations; connections between quasinormal modes and photon rings (Schutz & Will; Ferrari & Mashhoon; Berti et al.); detector response formalism and pattern functions (Maggiore; LIGO instrumentation papers); matched filtering, SNR, and Fisher matrix methods. It also references observational milestones (GW150914), effective-one-body approaches for inspiral-merger-ringdown modeling, and tests beyond GR (polarization content).

• Linear perturbation theory on flat spacetime: Expand gμν=ημν+hμν, adopt Lorenz (harmonic) gauge, derive linearized Einstein equations. Plane-wave solutions demonstrate two transverse polarizations (h+ and h×) and propagation at light speed. Use residual gauge to reach transverse-traceless (TT) form.
• Linear perturbations on curved backgrounds: In Lorenz gauge derive wave equation with curvature coupling (□hαβ+2Rα β μ ν hμν=0), showing scattering by curvature and counting of physical degrees of freedom via gauge constraints.
• Scalar–vector–tensor (SVT) decomposition on flat space: Decompose metric perturbations into scalars (φ,γ,H,λ), vectors (βi,εi), and TT tensor hTTij. Construct gauge-invariant Bardeen variables. Einstein equations split into Poisson-like (non-propagating) equations for scalars/vectors and wave equation for TT modes.
• Generation of GWs I (linear theory): Retarded Green’s function solution for TT metric; projection via transverse-traceless projector; standard quadrupole derivation using stress-energy conservation in flat space, yielding hTTij≈(2G/(c4r))Pijkl Q̈kl. Caveats noted for compact binaries.
• Post-Newtonian (PN) expansion: Systematic expansion in 1/c; metric ansatz with scalar potentials φ,ψ, vector ωi, and TT tensor χij; derive Einstein equations in Poisson and standard PN gauges through 1PN, identifying radiation reaction at 2.5PN. Present 1PN metric and PN potentials via Green integrals.
• Generation of GWs II (relaxed Einstein equations): Use harmonic coordinates and Hμν=ημν−√−g gμν; write fully nonlinear equations □η Hμν=−16π τμν, with τμν including matter and gravitational-field contributions (Landau–Lifshitz pseudotensor). Show ∂μτμν=0 encodes dynamics. Far-zone solution and conservation lead to a rigorous quadrupole formula with τtt sourcing the quadrupole, resolving linear-theory inconsistencies.
• Local flatness and coordinates: Prove local flatness theorem (Riemann normal coordinates); construct Fermi normal coordinates along an accelerated worldline; interpret apparent forces and relation to detector frames.
• GW stress-energy: Derive nonlocal, averaged stress-energy tensor of GWs (Isaacson) TαβGW=(1/32π)⟨∇αhTTρσ∇βhTTρσ⟩; emphasize averaging scale separation (λ≪L) and nonlocality due to equivalence principle.
• Application to compact stars: TOV equations; gravitational mass vs baryonic plus internal energy; difference equals binding/self-energy (Newtonian limit reproduces Uself).
• Binary inspiral and merger: Use quadrupole scaling, energy balance, and PN fluxes; define chirp mass; discuss geodesics, ISCO and photon ring in Schwarzschild/Kerr; role of spin (Lense–Thirring) and orbital hang-up; qualitative merger and ringdown picture.
• Black hole perturbations and QNMs: For Schwarzschild, decompose into odd/even sectors (Regge–Wheeler/Zerilli equations) and boundary conditions yielding damped QNMs. For Kerr, use Teukolsky formalism with spin-weighted spheroidal harmonics; QNM spectrum depends only on M and a (no-hair tests).
• Detector response: Low-frequency approximation via geodesic deviation and TT gauge, arm-length perturbation δL/L=(1/2)hTTij uiu j; pattern functions F+,F× and detector tensor; general frequency response via Killing symmetries and transfer function T(f)=sin(πfτ)/(πfτ). Phase response and dependence on orientation and frequency.
• Data analysis framework: Stationary Gaussian noise with power spectral density S(f). Inner product (A|B)=4 Re∫0∞ A*(f)B(f)/S(f) df; matched filtering via Wiener optimal filter; optimal SNR ρ2=(h|h). Stationary phase approximation for inspirals yields |h(f)|∝M5/6 c f−7/6/DL and SNR scaling; Fisher matrix for parameter uncertainties.

• Degrees of freedom: In GR, gravitational radiation carries two transverse, traceless tensor polarizations; scalar and vector perturbations are non-propagating potentials satisfying Poisson-like equations.
• Propagation: Linear GWs satisfy the wave equation and propagate at light speed; on curved backgrounds, wave propagation is influenced by curvature (scattering and inside-lightcone effects).
• Generation (rigorous): The relaxed Einstein equations with τμν (matter plus gravitational pseudotensor) justify the standard quadrupole formula hTTij=(2G/(c4r))Pijkl Q̈kl with Qij=∫τtt xixj d3x at leading PN order; using Tμν alone in linear theory leads to errors (e.g., factor-of-two discrepancy fixed by gravitational-field terms in τij).
• PN structure: Conservative corrections appear at even PN orders (1PN, 2PN...), while radiation reaction enters at 2.5PN (O(c−5)), consistent with the quadrupole luminosity ĖGW=(32/5)(G4/c5)(m1 2 m2 2 M/r5)∝v10/c5.
• GW energy and nonlocality: The GW stress-energy tensor is meaningful only after averaging over several wavelengths (geometric-optics/Isaacson result TαβGW=(1/32π)⟨∇αhTTρσ∇βhTTρσ⟩), reflecting the equivalence principle.
• Coordinates and frames: Construction of Riemann and Fermi normal coordinates clarifies local flatness and the observer’s laboratory (apparent forces for non-geodesic motion) and underpins detector response calculations.
• Detector response: Low-frequency phase shift Δφ∝[F+ h+ + F× h×]; full response includes transfer function T(f)=sin(πfτ)/(πfτ), causing high-frequency roll-off. Pattern functions encode sky position, polarization, and detector geometry.
• Inspiral scaling and detectability: For circular binaries, GW strain scales as h≈4 μ M2/3 Ω2/3 /(c4 D) (restoring units), with fGW=2forb; typical compact binaries at 10–400 Mpc produce strains ∼10−22–10−21 around 100 Hz. SNR threshold for detection is typically ≳8.
• Strong-field dynamics: In Schwarzschild, ISCO at r=6M; in Kerr, ISCO and photon-ring radii/frequencies depend on spin and orbit orientation (prograde vs retrograde), impacting inspiral length and ringdown. Orbital hang-up with aligned spins yields more cycles and higher radiative efficiency (up to ∼10% of total mass radiated in GWs for high aligned spins).
• Ringdown and tests of GR: Post-merger signals are superpositions of damped QNMs; in Kerr, mode spectrum depends only on (M,a), enabling no-hair consistency checks when multiple modes are measured.
• Data analysis: Optimal matched filtering weights by 1/S(f); inspiral Fourier-domain amplitude |h(f)|∝M5/6 c f−7/6/DL; Fisher information approximates parameter uncertainties at high SNR. Examples provided for ground-based and space-based PSDs and typical bandwidths.

The notes unify multiple complementary approaches—linear perturbation theory, PN expansion, and fully nonlinear relaxed equations—to provide a coherent and rigorous understanding of gravitational-wave generation and propagation. They clarify the physical content of GR (two tensor polarizations), the Newtonian-like role of scalar/vector perturbations in GR, and show how a correct far-zone GW solution arises only when including gravitational-field contributions to the source via τμν, thereby resolving inconsistencies of naive linear theory. The framework directly informs data analysis: detector response functions link theory to measured phase shifts; PSD-driven matched filtering and the stationary-phase inspiral model enable optimal searches and parameter inference. Strong-field insights from geodesic dynamics and BH perturbation theory explain qualitative features of observed signals—chirps culminating near an effective ISCO and damped ringdowns tied to the photon ring—supporting astrophysical interpretations such as the BH nature of GW150914. Together, these results underpin current GW astronomy and testing of GR in the radiative, highly dynamical regime.

The work provides a self-contained, first-principles treatment of gravitational-wave physics in GR: identifying the radiative degrees of freedom, deriving propagation on flat and curved backgrounds, rigorously establishing the quadrupole formula via the relaxed Einstein equations and PN theory, elucidating the nonlocal stress-energy of GWs, and connecting theory to detection through accurate detector response and data-analysis formalisms. It further bridges inspiral dynamics with strong-field BH perturbation theory, explaining merger-ringdown phenomenology and enabling GR tests (e.g., no-hair). Future directions include extending beyond-GR frameworks (additional polarizations and dipole radiation), higher-order PN and resummation (effective-one-body) improvements, tighter integration with numerical relativity for full waveforms, refined instrument modeling (transfer functions and non-Gaussian noise), and multi-band/multi-messenger synergies to probe fundamental physics and dense-matter microphysics.

• The presentation is pedagogical lecture notes rather than a peer-reviewed empirical study; many results are known but summarized and rederived for clarity.
• Approximations are essential: linearization, geometric-optics averaging, PN expansions (valid for v≪c and weak fields), and far-zone assumptions; numerical-relativity input is only qualitatively addressed.
• Astrophysical modeling (populations, environments, microphysics) is intentionally minimal; detector systematics and non-Gaussian/non-stationary noise are only briefly treated.
• Some derivations rely on specific gauges; although observables are gauge-invariant, practical implementations can be gauge-dependent. Exercises indicate further steps not fully worked out in the text.