---
title: Weinberg Gravitational-Wave Damping Formula
url: https://www.emergentmind.com/topics/weinberg-gravitational-wave-damping-formula
type: topic
---

# Weinberg Gravitational-Wave Damping Formula

Searching arXiv for papers directly relevant to Weinberg gravitational-wave damping, including viscous-fluid damping, free-streaming neutrino damping, and later generalizations.
Weinberg gravitational-wave damping formula denotes a class of results in relativistic cosmology in which gravitational-wave tensor perturbations do not propagate as freely oscillating modes once the cosmic medium develops dissipative or anisotropic stress. In the classic viscous-fluid form, shear viscosity adds a friction term to the tensor propagation equation and exponentially attenuates the wave amplitude with damping rate proportional to \(G\eta\), while bulk viscosity does not contribute at leading order to transverse-traceless modes [1803.11397]. In the early-universe free-streaming form associated with Weinberg’s later treatment of collisionless neutrinos, the damping is nonlocal in time: anisotropic stress sourced by decoupled relativistic particles feeds back through an integro-differential kernel and suppresses tensor amplitudes after horizon entry [1207.7285]. Subsequent work has extended these results to viscous cosmologies, self-interacting dark matter, generalized free-streaming species, modified gravity, closed spatial geometry, and dissipative corrections relevant to induced gravitational waves [1803.11397][1307.7571][1903.01502][1409.6958][2503.13670].

## 1. Historical definition and scope

The expression “Weinberg gravitational-wave damping formula” refers to two closely related strands of analysis. One strand concerns tensor perturbations in a viscous medium, where shear viscosity enters the stress-energy tensor and produces friction-like attenuation of gravitational waves during propagation. In the formulation used for cosmological applications, the damping rate is written as \(\beta = 16\pi G\eta\), so that the strain amplitude acquires an exponential suppression \(e^{-\beta D/2}\) over propagation distance \(D\) [1803.11397]. A later paper discussing viscous damping in a local Minkowski treatment uses \(\beta = 8\pi G\eta/c^2\) in the wave equation \(\partial_t^2 h_{ij} + 2\beta \partial_t h_{ij} - \partial_z^2 h_{ij}=0\), with the abstract stating a “damping rate of \(16\pi G\eta\),” corresponding to intensity rather than amplitude damping [2011.09169].

The second strand concerns free-streaming relativistic species, especially neutrinos. In that setting the effective dissipation is not ordinary hydrodynamic viscosity but anisotropic stress generated by the collisionless Boltzmann hierarchy. Weinberg’s tensor-mode equation becomes an integro-differential relation whose kernel is built from spherical Bessel functions, and the observable damping depends on a reduced wave number \(Q\) measuring horizon entry relative to matter-radiation equality [1207.7285]. In the short-wavelength regime, the power suppression approaches \(0.645\), while in the long-wavelength regime relevant to the tensor contribution to the CMB the suppression of the derivative-based observable is \(0.91032\) [1207.7285].

These two strands are conceptually unified by the same principle: tensor modes couple to non-perfect-fluid structure in the cosmic medium. In a hydrodynamic description, the relevant quantity is shear viscosity; in a collisionless description, it is anisotropic stress carried by free-streaming species. A plausible implication is that “Weinberg damping” is best understood as a general tensor attenuation framework rather than a single formula.

## 2. Viscous-medium formulation

In the viscous-fluid description, the background metric is taken to be FRW with transverse-traceless perturbations,
\[
ds^2 = -dt^2 + a^2(t)(\delta_{ij}+h_{ij})dx^i dx^j,
\]
and the imperfect-fluid stress-energy tensor includes shear viscosity \(\eta\) and bulk viscosity. The cosmological application in a viscous universe states that “when a nonzero shear viscosity \(\eta\) is introduced to the fluid energy-momentum tensor, GWs would be dissipated by matter with a damping rate \(\beta \equiv 16\pi G\eta\)” [1803.11397]. The same work emphasizes that the damping depends on shear viscosity rather than bulk viscosity or GW frequency [1803.11397].

The resulting strain formula is written as
\[
h_\alpha = \frac{A(\omega)}{D}\exp\left(\phi_0 + \frac{i\omega D}{a} - \int \frac{i\omega}{a}\,dt \right)e^{-\beta D/2},
\]
so the physically relevant attenuation law is
\[
h \propto \frac{1}{D}e^{-\beta D/2}, \qquad \beta = 16\pi G\eta .
\]
This is the viscous form of the Weinberg damping formula [1803.11397]. The interpretation is direct: the usual \(1/D\) geometric dilution is supplemented by exponential suppression from shear viscosity accumulated along the propagation path.

A complementary derivation in Minkowski background plus a viscous medium writes the tensor equation as
\[
\frac{\partial^2}{\partial t^2}h_{ij} + 2\beta \frac{\partial}{\partial t}h_{ij} - \frac{\partial^2}{\partial z^2} h_{ij}=0,
\]
with \(\beta = 8\pi G\eta/c^2\) in SI units [2011.09169]. Its approximate solution,
\[
h_{ij}\approx h\,e^{-\beta t - i\sqrt{k^2-\beta^2}\,t + ikz},
\]
shows explicitly that amplitude damping is exponential, while intensity damping carries the factor \(2\beta = 16\pi G\eta/c^2\) [2011.09169]. This difference of a factor of two is purely a matter of whether one quotes amplitude damping or energy-flux damping.

The same viscous framework underlies later kinetic-theory work on damping by matter more generally. A unified Boltzmann treatment distinguishes collisional damping and Landau damping, showing that collisions lead to a shear viscosity but also erase anisotropic stress, thereby suppressing damping in most astrophysical settings; damping by intergalactic or interstellar matter is negligible for all but primordial radiation [1707.05192]. In that sense, the viscous Weinberg formula is quantitatively significant mainly for cosmological or extreme-medium applications.

## 3. Effective-distance representation and observational use

A notable reformulation of the viscous result introduces an effective distance
\[
D_{\rm eff} = D\,e^{\beta D/2},
\]
which rewrites the damped strain in the same formal form as the perfect-fluid case [1803.11397]. The significance of this substitution is operational: luminosity-distance posteriors inferred from standard LIGO/Virgo waveform analyses can be reinterpreted directly as constraints on \(D_{\rm eff}\), rather than on the true luminosity distance \(D\) [1803.11397].

Using LIGO/Virgo events together with GW170817, the viscous-universe analysis found no evidence for damping and obtained an upper bound
\[
\beta < 6.75\times 10^{-4}\,{\rm Mpc}^{-1}
\]
at \(95\%\) confidence level [1803.11397]. The same paper notes that a source at luminosity distance \(D\gtrsim 10^4\,{\rm Mpc}\) would be needed to probe dark-matter self-interactions through this damping channel under the assumed hydrodynamic description [1803.11397].

A related 2020 study introduces the notion of a “viscosity redshift”
\[
z_v = e^{\int \beta\,dx}-1,
\]
so that the amplitude suppression factor becomes \((1+z_v)^{-1}\) and is degenerate with luminosity distance in the observed waveform amplitude [2011.09169]. In that treatment, viscosity alters amplitude but not frequency, unlike cosmological, Doppler, and gravitational redshifts, which affect both [2011.09169]. This distinction is useful conceptually because it isolates the phenomenological signature of viscous damping: sources appear farther away in gravitational waves without an accompanying frequency shift.

The same paper applies the formalism to a dark-matter mini-spike around an intermediate-mass black hole, with a density-dependent viscosity \(\eta = k\rho^\lambda\), and concludes that damping can be negligible for \(\lambda<0.6\) or \(r>0.01\,{\rm pc}\), but can become strong for more extreme parameters [2011.09169]. This suggests that environmental rather than truly cosmological viscosity may dominate in some astrophysical scenarios.

## 4. Free-streaming neutrinos and the integro-differential Weinberg equation

The collisionless-neutrino version of Weinberg damping is structurally different. Instead of a local friction term proportional to \(\dot h\), one obtains an integro-differential equation for the tensor transfer function \(\chi\). Stefanek and Repko give the general-wavelength form in terms of a reduced wave number \(Q=\sqrt{2}\,k/k_{\rm EQ}\), where \(k_{\rm EQ}\) is the comoving wave number entering the horizon at equality [1207.7285]. Their \(y\)-space equation is
\[
(1+y)\chi''(y)+\left(\frac{2(1+y)}{y}+\frac{1}{2}\right)\chi'(y)+Q^2\chi(y)
=
-\frac{24f_\nu(0)}{y^2}\int_0^y K(y,y')\frac{d\chi(y')}{dy'}\,dy',
\]
with \(f_\nu(0)=0.40523\) in the radiation era [1207.7285].

The kernel may be written as
\[
K(z)=\frac{1}{15}j_0(z)+\frac{2}{21}j_2(z)+\frac{1}{35}j_4(z),
\]
which makes a spherical-Bessel-series solution natural [1207.7285]. The analytic representation
\[
\chi(s,Q)=\sum_{n=0}^{\infty} a_n(Q) j_n(Qs)
\]
provides a uniform description across all wavelengths [1207.7285].

Two limiting damping results are especially important. For \(Q\to 0\), the relevant derivative ratio at last scattering gives
\[
\left|\frac{\chi'(s_L,0)}{\chi'_0(s_L,0)}\right|^2 = 0.91032,
\]
corresponding to about \(9\%\) suppression in power [1207.7285]. For \(Q\gg 1\), the asymptotic power suppression is
\[
D(Q\gg1)\approx 0.645,
\]
equivalent to roughly \(20\%\) suppression in amplitude [1207.7285]. These results refine the original Weinberg discussion by giving an explicit analytic series valid at all \(Q\).

Generalizations of this free-streaming framework replace standard neutrinos by arbitrary collisionless species. One study derives the anisotropic-stress kernel for general mass and distribution function, showing that nonzero neutrino masses reduce damping for modes with \(k\tau_0\approx100\)–200, and that a nonthermal axion background can produce a distinct phase-dependent damping signature [1307.7571]. Another paper embeds the Einstein-Boltzmann structure into restricted Horndeski gravity, finding that modified-gravity functions \(\delta_{1,2,3}(y)\) alter the effective damping and can maintain non-vanishing free-streaming damping even during matter domination, unlike general relativity [1903.01502].

## 5. Geometry, modified gravity, and later-era extensions

Several works extend the Weinberg damping framework beyond the standard flat radiation-dominated setting. In spatially closed cosmology, the Laplacian spectrum is discrete and curvature modifies both the tensor equation and the neutrino Boltzmann hierarchy. For modes entering the horizon during radiation domination, the closed-universe calculation finds
\[
D_{n,\text{with }\nu}(u\gg1)\simeq 0.4910\,\frac{\sin u}{u},
\]
so the squared amplitude is reduced by about \(76\%\), substantially stronger than the flat-space Weinberg result \(0.8026^2\) [1409.6958]. A subsequent study extending the analysis to matter- and \(\Lambda\)-dominated eras reports much smaller suppression for flat matter-era modes, but strong damping in the closed case, again highlighting the importance of geometry [1608.01472].

A different extension considers expansion itself as an effective damping mechanism in closed de Sitter spacetime. There the amplitude and frequency are extracted from the geodesic deviation equation rather than directly from the tensor Einstein equation, and the standard \(h\propto 1/a\) scaling of subhorizon modes in flat FRW is generalized to curvature- and \(\Lambda\)-dependent functions \(h(t,n,\Lambda)\) and \(\omega(t,n,\Lambda)\) [2105.11690]. The paper argues that the cosmological-constant contribution matters mainly in the early universe, while the dominant late-time effect remains expansion-induced damping [2105.11690]. This suggests that “Weinberg damping” can also encompass geometric redshifting and dilution, not only material dissipation.

Modified-gravity generalizations include \(f(R)\) gravity, where the tensor mode acquires extra friction terms proportional to \(\dot R\) and a scalar polarization appears with its own damping behavior [2302.06402]. In that framework, Landau damping is absent in flat spacetime while collision damping is present, and for \(f(R)=R+\alpha R^2\) the tensor mode with \(m=1\,{\rm eV}\) decays faster than in the massless case, whereas the scalar mode with \(m=1\,{\rm eV}\) suppresses decay [2302.06402]. The general lesson is that the Weinberg idea survives intact, but the effective damping kernel and the propagating degrees of freedom depend strongly on the gravitational theory.

## 6. Current reinterpretations and induced-gravitational-wave applications

Recent work on induced gravitational waves uses the same conceptual machinery but shifts the damping from the tensor sector to the scalar source sector. In radiation domination, finite mean free paths of photons and neutrinos damp scalar perturbations through shear viscosity, with damping scale determined by
\[
k_D^{-2}(\tau)\approx \int_0^\tau d\tilde\tau\,\frac{2\eta(\tilde\tau)}{3a(\rho+P)},
\]
explicitly invoking Weinberg’s 1971 viscosity treatment [2503.13670]. The induced-GW kernel is then computed with an exponentially damped scalar transfer function rather than a damped tensor Green’s function [2503.13670].

The main effects are regularization of the resonant frequency and replacement of the usual far-infrared logarithmic running by a low-frequency tail without logarithmic running [2503.13670]. The paper interprets the damping parameter as an effective finite source lifetime, directly analogous to the finite-width regularization of resonances [2503.13670]. This is not Weinberg’s 2004 neutrino damping formula in the strict sense, but it is a clear descendant of the same dissipative logic.

The paper further notes that direct free-streaming damping of gravitational waves, as in Weinberg’s neutrino calculation, is not included in its main computation [2503.13670]. That separation is useful: it isolates the effect of dissipation in the source sector from dissipation in the propagation sector. A plausible implication is that a fully consistent small-scale induced-GW calculation should incorporate both.

A different modern direction considers microphysical GW-to-EM conversion in a magnetized intergalactic plasma. That mechanism is not part of Weinberg’s original anisotropic-stress formalism, but it can be cast phenomenologically as an effective attenuation coefficient in the GW dispersion relation, with strong frequency dependence at long wavelengths [2202.03830]. This suggests that the broader Weinberg paradigm now includes both hydrodynamic and microscopic damping channels.

## 7. Conceptual interpretation and common misconceptions

A common misconception is that “gravitational-wave damping” always means ordinary Hubble dilution. In fact the literature distinguishes at least three mechanisms. First, expansion alone redshifts subhorizon amplitudes roughly as \(1/a\), which is geometric rather than dissipative [2105.11690]. Second, free-streaming species generate nonlocal anisotropic stress that damps tensor modes even in otherwise standard cosmology [1207.7285]. Third, imperfect-fluid effects such as shear viscosity produce genuine exponential attenuation \(e^{-\beta D/2}\) with \(\beta\propto G\eta\) [1803.11397]. These mechanisms can coexist but are not interchangeable.

A second misconception is that any viscosity contributes equally to tensor damping. The viscous-universe treatments are explicit that bulk viscosity does not attenuate transverse-traceless gravitational waves at leading order; only shear viscosity enters the damping rate [1803.11397][2011.09169].

A third misconception is that damping by matter should be substantial for ordinary astrophysical gravitational waves propagating through galaxies or the intergalactic medium. Kinetic-theory analyses show that both collisional and Landau damping by matter are negligible for essentially all astrophysical sources except primordial radiation, because maximal collisional damping is significant only when the wavelength is comparable to the horizon scale [1707.05192]. This sharply limits the phenomenological importance of the Weinberg damping formula outside early-universe or deliberately exotic settings.

Finally, the various quoted damping factors should not be conflated. The viscous-fluid exponential law refers to amplitude suppression during propagation through a medium [1803.11397]. The \(0.91032\) and \(0.645\) numbers refer to tensor-mode suppression by free-streaming neutrinos in specific \(Q\) limits [1207.7285]. The \(0.4910\) result refers to a spatially closed cosmology, not to flat FRW [1409.6958]. Each number belongs to a distinct physical and geometric regime.

In contemporary usage, the Weinberg gravitational-wave damping formula is therefore best regarded as a family of tensor attenuation laws arising from anisotropic stress or dissipation in cosmological media, with the precise kernel determined by whether the medium is viscous, collisionless, curved-background, or modified-gravity in nature.

Source: https://www.emergentmind.com/topics/weinberg-gravitational-wave-damping-formula