Major Detections

1. Introduction — The Dawn of Gravitational Wave Astronomy

For decades, gravitational waves existed only as a theoretical prediction of Einstein's 1916 general relativity. The indirect evidence from the Hulse–Taylor binary pulsar PSR B1913+16, whose orbital decay matched the GR prediction to better than 0.2%, earned the 1993 Nobel Prize and confirmed that gravitational radiation carries energy. But the dream of direct detection required measuring strains of order $h \sim 10^{-21}$ — a displacement a thousand times smaller than the diameter of a proton across LIGO's 4-km arms.

Advanced LIGO achieved this sensitivity by 2015, and the very first observing run (O1) produced a discovery that would reshape astrophysics. The signal, designated GW150914, arrived at the Livingston detector 6.9 ms before Hanford — consistent with a source in the southern sky. The waveform swept upward in frequency from 35 Hz to 250 Hz in about 0.2 seconds, perfectly matching the prediction for an inspiral, merger, and ringdown of two black holes.

The subsequent observing runs O2 (2016–2017) and O3 (2019–2020) expanded the catalog dramatically. The Gravitational-Wave Transient Catalog (GWTC) now contains approximately 90 confident detections: binary black hole (BBH) mergers, binary neutron star (BNS) mergers, and neutron star–black hole (NSBH) systems. Each event encodes information about the source masses, spins, distance, and sky location, extracted through matched filtering against banks of theoretically computed waveform templates.

In the sections that follow, we derive the physical quantities that make each of these detections a precision measurement of fundamental physics.

2. Derivation: GW150914 Parameter Extraction

The waveform of a compact binary inspiral is determined to leading order by a single combination of the component masses called the chirp mass. We now show how to extract this quantity, the final black hole mass, the radiated energy, and the luminosity distance from the observed signal.

2.1 Chirp Mass from the Frequency Evolution

In the Newtonian limit, the orbital frequency of a circular binary evolves according to the balance between gravitational-wave energy loss (quadrupole formula) and the orbital binding energy. The gravitational-wave frequency $f = 2f_{\rm orb}$ obeys the chirp equation:

where the chirp mass is defined as:

This equation can be rearranged to solve directly for the chirp mass. Measuring $f$ and$\dot{f}$ at any point on the inspiral waveform:

For GW150914, the observed (redshifted) frequency near $f \approx 40\,\text{Hz}$ and the corresponding $\dot{f}$ yield a detector-frame chirp mass of $\mathcal{M}_c^{\rm det} \approx 30.4\,M_\odot$. Converting to the source frame using $\mathcal{M}_c = \mathcal{M}_c^{\rm det}/(1+z)$ with $z \approx 0.09$:

This single number, combined with the mass ratio $q = m_2/m_1$ extracted from higher-order post-Newtonian corrections to the phase evolution, determines the individual component masses: $m_1 \approx 36\,M_\odot$ and $m_2 \approx 29\,M_\odot$.

2.2 Final Black Hole Mass from the Ringdown

After merger, the remnant black hole settles into a Kerr solution by radiating quasi-normal modes (QNMs). The dominant $(\ell, m, n) = (2, 2, 0)$ mode has a frequency and damping time that depend only on the mass $M_f$ and spin $a_f$ of the final hole (the no-hair theorem). The ringdown frequency is approximately:

For GW150914, the observed ringdown frequency $f_{\rm ring} \approx 250\,\text{Hz}$combined with the damping time yields $M_f \approx 62\,M_\odot$ and$a_f \approx 0.67$. Note that the total initial mass was$m_1 + m_2 \approx 65\,M_\odot$, so approximately $3\,M_\odot$ was radiated as gravitational waves.

2.3 Radiated Energy

The energy carried away by gravitational waves during the inspiral, merger, and ringdown is simply the mass deficit:

This energy was released in approximately 0.2 seconds. The peak gravitational-wave luminosity therefore was:

This exceeds the combined electromagnetic luminosity of all stars in the observable universe by a factor of roughly 50. For a brief instant, GW150914 was the most powerful event in the universe since the Big Bang.

2.4 Luminosity Distance from the Amplitude

The gravitational-wave strain amplitude for a circular binary at leading (Newtonian) order is:

Since $\mathcal{M}_c$ is known from the frequency evolution and $f$ is measured directly, the observed amplitude $h$ determines the luminosity distance $d_L$. Solving:

For GW150914, the best-fit analysis yields $d_L = 410^{+160}_{-180}\,\text{Mpc}$, corresponding to $z \approx 0.09$. The large uncertainty arises because the amplitude depends on the unknown inclination angle of the binary orbital plane with respect to the line of sight; a face-on binary appears about twice as loud as an edge-on one.

3. Derivation: GW170817 and Multi-Messenger Astronomy

On August 17, 2017, LIGO and Virgo detected a long-duration inspiral signal from two neutron stars merging in the galaxy NGC 4993, at a distance of $\sim 40\,\text{Mpc}$. Just 1.7 seconds after the GW signal ended, the Fermi and INTEGRAL satellites detected a short gamma-ray burst (GRB 170817A). This was followed by optical, infrared, X-ray, and radio counterparts — the first multi-messenger observation with gravitational waves.

3.1 Speed of Gravity from the GW–GRB Delay

The gravitational waves and gamma rays traveled $D \approx 40\,\text{Mpc} \approx 1.3 \times 10^{26}\,\text{m}$from the source. They arrived within $\Delta t = 1.7\,\text{s}$ of each other. The travel time for light is:

If the GW speed differs from $c$ by a fractional amount $\delta = v_{\rm GW}/c - 1$, then the arrival time difference due to this speed difference alone would be $\delta T = |\delta| \cdot T$. Requiring this to be less than the observed delay (conservatively, accounting for the possibility that the GRB was emitted with some intrinsic delay of order 10 seconds):

A more refined analysis, allowing for a plausible GRB emission delay of $(-10\,\text{s}, +1.7\,\text{s})$, gives the celebrated bound:

This single measurement eliminated vast classes of modified gravity theories (including many scalar-tensor, vector-tensor, and dark energy models) that predicted $v_{\rm GW} \neq c$.

3.2 Tidal Deformability and the Neutron Star Equation of State

Unlike black holes, neutron stars are tidally deformed by the gravitational field of their companion during the late inspiral. The tidal deformability $\Lambda$ parameterizes how much the star is distorted. The induced quadrupole moment is:

where $\mathcal{E}_{ij}$ is the external tidal field and $\lambda$ is the tidal deformability parameter. The dimensionless tidal deformability is defined as:

where $k_2$ is the Love number and $R$ is the stellar radius. The tidal effects enter the gravitational-wave phase at 5PN (post-Newtonian) order through a mass-weighted combination $\tilde{\Lambda}$:

The GW170817 analysis constrains $\tilde{\Lambda} = 300^{+420}_{-230}$ (90% credible interval), ruling out very stiff equations of state. This corresponds to neutron star radii of $R \approx 11\text{--}13\,\text{km}$, providing one of the tightest constraints on the dense matter equation of state.

3.3 Kilonova Lightcurve from r-Process Heating

The neutron-rich material ejected during the merger undergoes rapid neutron capture (r-process) nucleosynthesis, forming heavy elements beyond iron. As these radioactive nuclei decay, they power an optical/infrared transient called a kilonova. The basic model, due to Li & Paczyński (1998) and Metzger et al. (2010), gives the heating rate from radioactive decay as:

where $\epsilon_0 \sim 2 \times 10^{10}\,\text{erg}\,\text{g}^{-1}\,\text{s}^{-1}$and $\alpha \approx 1.3$. The ejecta expand at velocity $v_{\rm ej} \sim 0.1\text{--}0.3\,c$and become optically thin at the diffusion timescale:

where $\kappa$ is the opacity, which depends critically on the composition. Lanthanide-rich ejecta (from dynamical tidal tails) have $\kappa \sim 10\,\text{cm}^2/\text{g}$, producing a red, long-lived kilonova peaking after $\sim 1$ week. Lanthanide-poor polar ejecta (from neutrino-driven winds) have $\kappa \sim 1\,\text{cm}^2/\text{g}$, producing a blue, fast-fading component peaking after $\sim 1$ day. The total bolometric luminosity follows:

The kilonova AT 2017gfo associated with GW170817 matched this model beautifully, with$M_{\rm ej} \approx 0.05\,M_\odot$. Spectroscopic analysis revealed the presence of strontium and other heavy elements, confirming that NS mergers are a major (possibly dominant) site of r-process nucleosynthesis in the universe.

4. Derivation: Tests of General Relativity

Every detected gravitational wave event provides a test of general relativity in the strong-field, highly dynamical regime — a regime inaccessible to Solar System or binary pulsar experiments. The LIGO/Virgo/KAGRA collaboration has developed a systematic framework for performing these tests.

4.1 Parameterized Post-Einsteinian Framework

The inspiral waveform phase can be expanded in post-Newtonian (PN) orders. In the parameterized post-Einsteinian (ppE) framework, each PN coefficient is allowed to deviate from its GR value:

where $\delta\hat{\varphi}_k$ are the fractional deviations at each PN order$k$. GR predicts $\delta\hat{\varphi}_k = 0$ for all $k$. Using GW150914 and subsequent events, the LVK collaboration has bounded all tested coefficients to be consistent with zero at the $\sim 10\text{--}100\%$ level, with the tightest constraints coming from the lower PN orders (which accumulate the most phase over the inspiral).

4.2 Testing the No-Hair Theorem from Ringdown

The no-hair theorem states that a stationary black hole in GR is completely characterized by its mass $M$ and spin $a$ (and charge, which is astrophysically negligible). Therefore, all QNM frequencies and damping times are functions of $(M, a)$ alone. If we can measure two or more QNM overtones, we can test whether they are mutually consistent with a single Kerr black hole. The QNM spectrum is:

where the quality factor $Q = \pi f \tau$ relates the frequency to the damping time $\tau$. From the dominant $(2,2,0)$ mode, we extract one pair$(f_{220}, Q_{220})$ and determine $(M_f, a_f)$. If a subdominant mode (e.g., $(3,3,0)$ or $(2,2,1)$) is also measurable, its frequency and damping must be predicted by the same $(M_f, a_f)$. Any deviation would signal a departure from Kerr geometry — evidence for exotic compact objects or beyond-GR physics.

Current tests with GW150914 (and the louder event GW150914-like events in GWTC-3) show consistency with the no-hair theorem at the $\sim 20\%$ level. Future detectors will push this to $\sim 1\%$.

4.3 Area Increase Theorem

Hawking's area theorem (1971) states that in classical GR, the area of a black hole event horizon can never decrease. For a Kerr BH, the horizon area is:

where $a_* = a/M = cJ/(GM^2)$ is the dimensionless spin. For GW150914, the sum of the initial horizon areas is:

The final area $A_f$ is computed from $(M_f, a_{*f})$. The LIGO/Virgo analysis of GW150914 found $A_f/A_{\rm initial} = 1.5 \pm 0.4$, confirming$A_f > A_{\rm initial}$ at the 97% confidence level — the first observational test of Hawking's area theorem.

4.4 Graviton Mass Bound from Dispersion

In GR, gravitational waves propagate at the speed of light because the graviton is massless. A massive graviton would introduce frequency-dependent dispersion: lower-frequency components would travel slower. The dispersion relation for a massive particle is:

This modifies the gravitational-wave phase. For a signal sweeping from frequency $f_1$to $f_2$, the accumulated phase difference between a massless and massive graviton, over a distance $D$, is:

where $h_P$ is Planck's constant (not the strain!). The absence of any detectable dispersion in the combined GWTC-3 catalog bounds the Compton wavelength of the graviton $\lambda_g = h_P/(m_g c) > 1.27 \times 10^{13}\,\text{km}$, or equivalently:

This is the most stringent dynamical bound on the graviton mass, improving on Solar System bounds by many orders of magnitude.

5. Derivation: Population Properties and the Mass Function

With a catalog of $\sim 90$ events from GWTC-1, GWTC-2, and GWTC-3, we can move beyond individual detections and study the population of compact binary mergers. The key observables are the merger rate density, the mass distribution, the spin distribution, and their evolution with redshift.

5.1 Merger Rate Density

The observed number of events in an observing run is related to the intrinsic merger rate density $\mathcal{R}(z)$ (mergers per comoving volume per proper time) by:

where $dV_c/dz$ is the comoving volume element, the factor $(1+z)^{-1}$converts from source-frame to detector-frame time, and $p_{\rm det}(z)$ is the detection probability (which depends on the masses, spins, and detector sensitivity). The comoving volume element in a flat $\Lambda$CDM cosmology is:

The inferred local BBH merger rate from GWTC-3 is $\mathcal{R}_{\rm BBH}(z=0) = 17.9\text{--}44\,\text{Gpc}^{-3}\,\text{yr}^{-1}$, and the BNS merger rate is $\mathcal{R}_{\rm BNS}(z=0) = 10\text{--}1700\,\text{Gpc}^{-3}\,\text{yr}^{-1}$. The BBH rate increases with redshift roughly as $\mathcal{R}(z) \propto (1+z)^\kappa$ with $\kappa \approx 2.7$, broadly consistent with the cosmic star formation rate.

5.2 The Mass Function and Mass Gaps

The primary mass distribution $dN/dm_1$ reveals rich structure. The population is commonly modeled as a mixture of a truncated power law and a Gaussian peak:

where $S(m_1)$ is a smoothing function at the low-mass end, $\mathcal{W}$is a window function tapering the distribution at the high-mass end, and the Gaussian component with fraction $\lambda_p$ captures a possible peak in the mass spectrum near $\sim 35\,M_\odot$. The GWTC-3 analysis finds $\alpha \approx 3.4$ and a Gaussian peak near $\mu_m \approx 34\,M_\odot$.

Two notable features in the mass spectrum stand out:

5.3 Formation Channels

The mass and spin distributions encode information about the formation pathway. The two main channels are:

• ●Isolated binary evolution: Two massive stars in a binary undergo common envelope evolution, producing aligned spins ($\chi_{\rm eff} > 0$) and a mass ratio $q$ clustered near unity. The mass function is set by stellar winds and the initial mass function.
• ●Dynamical assembly: In dense stellar environments (globular clusters, nuclear star clusters, AGN disks), BHs pair up through gravitational encounters. This produces isotropic spin orientations ($\chi_{\rm eff}$ distributed symmetrically around zero) and allows hierarchical mergers that populate the pair-instability mass gap.

The GWTC-3 population analysis suggests that both channels contribute, with the effective spin distribution showing a preference for small positive values ($\langle\chi_{\rm eff}\rangle \approx 0.06$), and a tail toward negative $\chi_{\rm eff}$ suggesting a dynamical subpopulation.

6. Applications

7. Historical Context

The path to gravitational wave detection spans over a century of theoretical development and five decades of experimental effort.

8. Python Simulation

The following simulation generates two plots. First, a GW150914-like chirp waveform showing the inspiral, merger, and ringdown phases. The inspiral phase uses the leading-order post-Newtonian frequency evolution (chirp equation), while the ringdown is modeled as a damped sinusoid at the QNM frequency. Second, a simulated mass distribution inspired by GWTC-3, illustrating the lower mass gap ($2.5\text{--}5\,M_\odot$) and the pair-instability gap ($\sim 50\text{--}120\,M_\odot$).