Pith. sign in

REVIEW 4 major objections 6 minor 2 cited by

Vector dark matter generates gravitational waves through broken isotropy

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

T0 review · glm-5.2

2026-07-05 01:48 UTC pith:7RSOUUHD

load-bearing objection GW spectrum from VFDM: careful derivation, but the main numerical feature may be beyond the perturbative order used the 4 major comments →

arxiv 2604.21080 v3 pith:7RSOUUHD submitted 2026-04-22 astro-ph.CO gr-qc

Cosmological Gravitational Waves from Ultralight Vector Dark Matter

Tom\'as Ferreira Chase , Diana L\'opez Nacir This is my paper
classification astro-ph.CO gr-qc
keywords ultralight vector dark mattergravitational wavesBianchi I cosmologyscalar-vector-tensor mixingstochastic gravitational wave backgroundanisotropic perturbationsProca field cosmologyWeinberg adiabatic mode
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

If dark matter is an ultralight vector (spin-1) field rather than a scalar, its homogeneous background points in a fixed spatial direction, breaking the universe's isotropy at the fundamental level. This paper argues that the resulting anisotropic geometry — a Bianchi I spacetime rather than the usual isotropic FRW — forces scalar, vector, and tensor perturbation sectors to mix at linear order, something that never happens in standard cosmology. The mixing means ordinary density perturbations (scalar sector) act as a source for gravitational waves (tensor sector), producing a stochastic gravitational-wave background that would not exist if dark matter were a scalar field. The authors derive the full linear perturbation equations in the anisotropic background, show that Weinberg's adiabatic mode construction survives the anisotropy at leading order through cancellations between vector-field and metric-shear terms, and implement the coupled system numerically in a modified version of CLASS. The predicted present-day gravitational-wave spectrum is anisotropic — its amplitude depends on the angle between the dark-matter field direction and each Fourier mode via a factor of sin²(γ_k) — and exhibits characteristic peaks near the Jeans scale at the time the vector field begins oscillating. The spectrum's shape and amplitude depend on the vector boson mass, with lighter masses producing more gravitational-wave power. For the masses considered, the signal lies below current detector sensitivities but within reach of future experiments.

Core claim

The central mechanism is the scalar-vector-tensor (SVT) mixing induced by a background vector dark matter field in a Bianchi I spacetime. In standard isotropic cosmology, scalar (density), vector (vorticity), and tensor (gravitational wave) perturbations evolve independently at linear order — this decoupling is a direct consequence of spatial isotropy. A homogeneous vector field with a fixed spatial direction breaks that isotropy, and the paper shows that the broken symmetry couples the sectors through the metric shear tensor σ_ij and the vector field's anisotropic stress. The key calculational result is that scalar perturbations source tensor modes through this coupling, generating a calcul

What carries the argument

Bianchi I metric with metric shear σ_ij; SVT (scalar-vector-tensor) mixing at linear order in perturbation theory; Weinberg adiabatic mode construction extended to anisotropic backgrounds; class.VFDM (modified CLASS Boltzmann solver) for numerical evolution

Load-bearing premise

The perturbative treatment of the metric shear (keeping only linear order in σ_ij) is self-consistent for computing the gravitational-wave spectrum. The authors themselves note that the backreaction of tensor modes on the scalar sector is of the same order as the linear shear terms they neglect, and that second-order corrections are needed for reliable predictions at high frequencies.

What would settle it

If second-order corrections in the metric shear substantially alter the spectrum shape or amplitude — particularly at high frequencies where relativistic modes contribute — the quantitative predictions of this paper would need revision. Conversely, a measurement of the predicted anisotropic spectrum with peaks at the Jeans scale for a given mass would confirm the mechanism.

Watch this falsifier. Get emailed when new claim-graph text bears on it.

If this is right

  • A detection of an anisotropic stochastic gravitational-wave background with a sin²(γ_k) angular dependence would point specifically toward vector dark matter and away from scalar alternatives.
  • The gravitational-wave spectrum's peak location encodes the vector boson mass through the Jeans scale at oscillation onset, offering a potential mass-measurement channel complementary to structure-formation constraints.
  • If the primordial scalar-tensor cross-correlation is nonzero (as some inflationary models generating vector fields might predict), the gravitational-wave signal would be amplified beyond what the paper computes with zero primordial tensors.
  • Extending the calculation to second order in the metric shear — which the authors flag as needed for precision on small scales — could shift the high-frequency portion of the spectrum and might reveal additional observational signatures.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

4 major / 6 minor

Summary. This paper studies the production of a stochastic gravitational-wave (GW) background from ultralight vector (spin-1) dark matter (VFDM) in a Bianchi I spacetime. The homogeneous background vector field breaks spatial isotropy, inducing mixing between scalar, vector, and tensor (SVT) perturbation sectors. The authors derive the linear perturbation equations, construct super-horizon adiabatic initial conditions via the Weinberg construction (verifying explicit cancellations between vector-field and shear-tensor contributions), and implement the tensor-sector evolution in a modified version of CLASS (class.VFDM). The main numerical output is the present-day GW abundance spectrum (Fig. 2), which exhibits peaks near the Jeans scale at oscillation onset and an anisotropic sin^2(gamma_k) dependence. The public code is provided.

Significance. The paper addresses a well-motivated question—distinguishing the spin of ultralight dark matter through cosmological observables—and extends prior work (Refs. [12, 13]) by the same authors to the tensor sector. The derivation of the perturbation equations and the Weinberg adiabatic mode construction in a Bianchi I background with SVT mixing is carefully executed, with cancellations verified explicitly. The provision of reproducible code (class.VFDM, Ref. [22]) and falsifiable spectral predictions (Fig. 2, with detector sensitivity comparisons) are concrete strengths. The anisotropic sin^2(gamma_k) dependence is a distinctive, testable signature. However, the central quantitative claim—the GW spectrum—rests on a perturbative truncation whose self-consistency for the dominant spectral features is questionable, as detailed below.

major comments (4)
  1. Sec. IV, discussion of Fig. 2 and the zoom in Fig. 1: The first peak in the GW spectrum is explicitly attributed by the authors to the discrepancy between the numerical solution and the leading-order Weinberg adiabatic mode. The text states: 'the numerical solution is different from the Weinberg solution... and the difference is an effect of second order in |sigma_ij|.' The equations of motion (Eq. 32) and the VFDM fluid variables (Eqs. 21) retain only zeroth-order terms in |sigma_ij| in the stress-energy tensor (with the Einstein tensor kept to linear order). If the leading-order solution cancels exactly (as demonstrated by the Weinberg construction in Sec. III), the residual signal generating the first peak is formally second order in |sigma_ij|, but the code does not include the second-order source terms that would self-consistently produce this signal. The authors should clarify: (a)
  2. Sec. IV: Photon and neutrino sources for tensor modes are dropped 'for simplicity.' The standard free-streaming damping of tensor modes by neutrinos (and photons before recombination) is a well-known effect that suppresses the GW spectrum at frequencies entering the horizon during radiation domination. Since the predicted spectrum in Fig. 2 has features precisely in the frequency range where this damping is relevant, the omission is load-bearing for the quantitative amplitude. The authors should either include these sources or provide a quantitative estimate of the expected suppression.
  3. Sec. IV, Fig. 2: The spectrum is shown only for transversal modes (gamma_k = pi/2), with the statement that the general case is multiplied by sin^2(gamma_k). However, the angular dependence of the scalar-tensor mixing and the vector-field perturbation sources (Eqs. 21, 33) involves both cos(gamma_k) and sin(gamma_k) factors in a non-trivial way. A simple sin^2(gamma_k) overall factor should be verified explicitly, or the qualification should be stated more precisely.
  4. Sec. IV, end of section and Sec. V: The authors acknowledge that the backreaction of tensor modes on the scalar sector is of the same order as terms linear in |sigma_ij| that are neglected, and that second-order corrections in sigma_ij are needed for precision on small scales. Given that the high-frequency increase in the spectrum is attributed to relativistic modes (small scales), the quantitative reliability of this portion of Fig. 2 is unclear. The authors should delineate which frequency range of the predicted spectrum is robust at the stated perturbative order and which is not.
minor comments (6)
  1. Ref. [22] (code repository) appears as an unresolved placeholder in the reference list. Please verify the link is correct and functional.
  2. Eq. (37) and surrounding text: the notation for the transfer functions (script T) is introduced somewhat abruptly. A brief sentence defining each symbol would improve readability.
  3. Fig. 1: the axis labels and legend are difficult to read in the multi-panel layout. Please increase font sizes and clarify what each curve represents.
  4. Sec. II, Eq. (4): the solution for sigma_ij is stated without derivation in the main text (deferred to Appendix A). A brief forward-reference or a one-line summary of the perturbative assumption would help if stated upfront.
  5. Appendix C, Eqs. (C4)-(C5): the averaging parameters theta_tol = 0.1 and theta_th = 100 are stated without justification. A brief comment on the sensitivity of results to these values would be useful.
  6. The abstract states the spectrum is obtained from class.VFDM but does not mention the key caveats (neglected photon/neutrino sources, perturbative order limitations). Consider adding a brief qualifier or adding a footnote.

Simulated Author's Rebuttal

4 responses · 0 unresolved

We thank the referee for a careful and constructive report. The referee raises important questions about (1) the self-consistency of the perturbative truncation for the first peak in the GW spectrum, (2) the omission of photon and neutrino damping sources, (3) the angular dependence of the spectrum, and (4) the robustness of the high-frequency portion of the spectrum. We address each point below. We find that comments (1), (2), and (4) are partially or fully valid and require revisions to the manuscript text to clarify the perturbative order and limitations. Comment (3) is addressed by an explicit verification of the angular dependence.

read point-by-point responses

  1. Referee: Sec. IV, discussion of Fig. 2 and the zoom in Fig. 1: The first peak in the GW spectrum is attributed to the discrepancy between the numerical solution and the leading-order Weinberg adiabatic mode, which is formally second order in |sigma_ij|, but the code does not include second-order source terms. The authors should clarify this self-consistency issue.

    Authors: The referee is correct that the first peak in the GW spectrum arises from a residual that is formally second order in |sigma_ij|, while the equations implemented in the code retain only zeroth-order terms in |sigma_ij| in the stress-energy tensor (with the Einstein tensor kept to linear order). We acknowledge that this is a genuine limitation of the current perturbative treatment. To clarify: the code evolves the full (untruncated) tensor equation of motion, Eq. (32), including all terms shown. The Weinberg adiabatic mode provides the initial conditions and the super-horizon analytic approximation. The discrepancy between the numerical solution and the Weinberg mode for the lightest masses (visible in the zoom of Fig. 1) arises because the numerical solution captures effects that the leading-order Weinberg construction does not—specifically, subleading corrections in |sigma_ij| that enter through the full evolution equations but are not fully self-consistent at the level of the source terms in the stress-energy tensor. We agree that a fully self-consistent treatment of this peak would require including second-order source terms in |sigma_ij|. We will revise the manuscript to state this limitation explicitly and to qualify the first peak as a feature whose existence is robust (it arises from the known incompleteness of the leading-order Weinberg approximation) but whose quantitative amplitude is not fully self-consistent at the stated perturbative order. revision: partial

  2. Referee: Sec. IV: Photon and neutrino sources for tensor modes are dropped 'for simplicity.' The standard free-streaming damping by neutrinos and photons suppresses the GW spectrum at frequencies entering the horizon during radiation domination. The omission is load-bearing for the quantitative amplitude. The authors should either include these sources or provide a quantitative estimate of the expected suppression.

    Authors: The referee is correct that the omission of photon and neutrino anisotropic stress sources is load-bearing for the quantitative amplitude of the GW spectrum in the frequency range where free-streaming damping is relevant. We will address this in two ways. First, we will add a quantitative estimate of the expected suppression. The standard damping of tensor modes by free-streaming neutrinos suppresses the GW amplitude by a factor of approximately (1 - R_nu/2)^2 ~ 0.8 in the radiation-dominated era for modes entering the horizon before matter-radiation equality, where R_nu ~ 0.4 is the neutrino fraction of radiation. For modes entering during matter domination, the suppression is negligible. Second, we will include the neutrino and photon tensor sources in a forthcoming update to class.VFDM. For the revised manuscript, we will add a paragraph in Sec. IV providing this quantitative estimate and clearly stating that the spectrum in Fig. 2 represents an upper bound on the GW abundance in the relevant frequency range, with the actual spectrum suppressed by the factor quoted above. revision: partial

  3. Referee: Sec. IV, Fig. 2: The spectrum is shown only for transversal modes (gamma_k = pi/2), with the statement that the general case is multiplied by sin^2(gamma_k). However, the angular dependence involves both cos(gamma_k) and sin(gamma_k) factors in a non-trivial way. A simple sin^2(gamma_k) overall factor should be verified explicitly.

    Authors: We have verified the angular dependence explicitly. The key observation is that the background vector field is decomposed as A_L = cos(gamma_k) * c * tau and A_T = sin(gamma_k) * c * tau (see Eqs. A9). The tensor perturbation h_+ is sourced by the transversal component of the vector field perturbation delta_A_{t1} and by the shear component sigma_+, both of which are proportional to A_T^2 and hence to sin^2(gamma_k). Specifically, sigma_+ = -c_T^2 / (2 m_P^2 a^2) (Eq. A9c), where c_T = sin(gamma_k) * c. The source terms in the tensor equation (Eq. 33a) and the mass term (Eq. 32a) both involve A_{t1} and sigma_+, which carry the sin^2(gamma_k) factor. The longitudinal component A_L does not source h_+ directly; it enters only through the constraint equation (Eq. 11) and the scalar sector, which decouple from the tensor sector for the adiabatic mode (as shown by the cancellations in Sec. III). For the h_x polarization, the background has no component along e_2 by construction, so sigma_x = 0 and h_x = 0. Therefore, the overall sin^2(gamma_k) factor is exact for the GW spectrum sourced by the VFDM mixing. We will add a brief derivation of this result in the revised manuscript to make the verification explicit. revision: yes

  4. Referee: Sec. IV, end of section and Sec. V: The backreaction of tensor modes on the scalar sector is of the same order as terms linear in |sigma_ij| that are neglected, and second-order corrections are needed for precision on small scales. The quantitative reliability of the high-frequency portion of Fig. 2 is unclear. The authors should delineate which frequency range is robust and which is not.

    Authors: The referee is correct that the high-frequency portion of the spectrum, attributed to relativistic modes on small scales, is not quantitatively reliable at the stated perturbative order. We will revise the manuscript to delineate the robust and non-robust frequency ranges. Specifically: (1) The low-frequency portion of the spectrum (frequencies below the Jeans scale at oscillation onset, corresponding to the first peak) is robust in terms of the existence of the feature, though its quantitative amplitude is affected by the second-order issues discussed in response to comment 1. (2) The second peak, corresponding to the Jeans scale k_J = a_osc * m * H_osc, is a robust qualitative feature determined by the background dynamics. (3) The high-frequency increase in the spectrum (frequencies above the Jeans scale) is not quantitatively reliable, as we explicitly acknowledge in the manuscript: 'on such small scales it becomes necessary to include second-order corrections in cosmological perturbation theory.' We will add a clear statement in Sec. IV and Sec. V delineating these ranges, and we will shade or mark the non-robust region in Fig. 2. We will also add a statement that the spectrum should be understood as an order-of-magnitude estimate in the high-frequency regime, pending the inclusion of next-to-leading-order corrections in |sigma_ij|. revision: yes

Circularity Check

0 steps flagged

No significant circularity; self-citations provide building blocks but the GW spectrum is independently derived

full rationale

The paper's derivation chain proceeds from the Proca action (Eq. 1) through the Bianchi I Einstein equations to obtain SVT-coupled perturbation equations (Eqs. 12, 21, 22, 32), uses Weinberg's adiabatic mode construction (external citations [14, 15]) for initial conditions, and numerically integrates the tensor perturbation equations to obtain the GW spectrum (Fig. 2). The constant c is fixed by the observed DM abundance (Eq. A8), an external input. The self-citations [12, 13] provide the background evolution and scalar-sector CLASS implementation, but these are prior derivations from the same first principles (Proca action + Einstein equations), not fits to the target GW quantity. The tensor-sector equations and their coupling to the scalar sector are derived in this paper. The sin²(γ_k) anisotropy factor arises geometrically from the vector field decomposition (Eqs. 6-7), not from a fit. The skeptic's concern about the first peak being a second-order effect in |σ_ij| beyond the truncation order is a correctness/consistency issue, not circularity—the authors are transparent about this limitation and do not hide it. No step in the derivation chain reduces to its inputs by construction.

Axiom & Free-Parameter Ledger

4 free parameters · 4 axioms · 1 invented entities

The axiom ledger is modest. The Proca action is a standard field theory. The key free parameter is the mass m, which is scanned rather than fitted. The constant c is fixed by the observed DM abundance—an external benchmark. The perturbative treatment of σ_ij is the main ad hoc assumption, acknowledged by the authors. No new particles or forces beyond the Proca field are invented.

free parameters (4)
  • m (vector field mass) = 10^{-22} eV and 10^{-25} eV shown
    The mass of the Proca field is a free parameter of the model, not derived. Results are shown for two representative values.
  • c (background field amplitude) = Fixed by Eq. (A8) using observed Ω_DM,0
    Integration constant of the background vector field, fixed by matching to the observed dark matter abundance. This is an external input, not a fit to GW data.
  • w+ (primordial tensor amplitude) = 0
    Set to zero to isolate the GW signal from SVT mixing. This is a choice of initial condition, not a fit.
  • θ_tol, θ_th (averaging parameters) = 0.1, 100
    Numerical parameters controlling the smoothing of oscillatory functions in the CLASS implementation (Eqs. C4-C5). These are computational choices, not physical parameters.
axioms (4)
  • domain assumption The dark matter is entirely described by a minimally coupled Proca field (Eq. 1).
    The Proca action with mass m is the starting point. No coupling to other sectors beyond gravity is included.
  • domain assumption The vector field was produced in a sufficiently coherent and homogeneous state that the classical field approximation is valid (Sec. I).
    The classical field treatment requires this production assumption; specific inflationary models that generate it are deferred to future work.
  • ad hoc to paper Metric shear σ_ij can be treated perturbatively (linear order) for the masses considered.
    The entire perturbative framework relies on σ_ij being small. The authors note this is valid for the masses considered but acknowledge next-to-leading order corrections are needed for precision (Sec. IV, end).
  • domain assumption Weinberg's adiabatic mode construction holds at leading order in anisotropies for Bianchi I with VFDM (Sec. III).
    The construction is shown to work via explicit cancellations, but relies on the anisotropic stress scaling as k² in the k→0 limit.
invented entities (1)
  • Ultralight vector (spin-1) dark matter field independent evidence
    purpose: Dark matter candidate
    The model makes falsifiable predictions: an anisotropic matter power spectrum (Ref. [12]) and now a GW background with a specific mass-dependent spectrum. These are testable against observations. The field itself is not directly detected, but the model is constrained by CMB and large-scale structure data.

pith-pipeline@v1.1.0-glm · 23878 in / 3077 out tokens · 296896 ms · 2026-07-05T01:48:40.210974+00:00 · methodology

0 comments
read the original abstract

We compute the abundance of cosmological gravitational waves produced during the evolution of an ultralight vector (spin-1) dark matter field. A homogeneous background vector field breaks spatial isotropy, requiring a Bianchi I geometry and inducing a mixing between the scalar, vector, and tensor perturbation sectors. We derive the perturbation equations in this background and show that, as a consequence of this mixing, scalar perturbations act as a source of tensor modes, generating a stochastic GW background. The production and cosmological evolution of these gravitational waves are implemented in \texttt{class.VFDM}, a modified version of \texttt{CLASS}, from which we obtain the present-day spectrum.

Figures

Figures reproduced from arXiv: 2604.21080 by Diana L\'opez Nacir, Tom\'as Ferreira Chase.

Figure 1
Figure 1. Figure 1: Numerical solution for the tensor perturbation amplitude, for the transversal Fourier modes ( [PITH_FULL_IMAGE:figures/full_fig_p009_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Gravitational wave abundance at present time (shadowed regions), calculated as defined in Eq. ( [PITH_FULL_IMAGE:figures/full_fig_p010_2.png] view at source ↗

Review history (2 revisions) →

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 2 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Isotropic universes with a preferred direction

    astro-ph.CO 2026-08 conditional novelty 7.0

    A tuned vector-field EFT can have an exactly isotropic FLRW background while hiding a preferred direction that reappears in perturbations as direction-dependent propagation and scalar–tensor mixing.

  2. Spin-1 Ultralight Dark Matter under Cosmological Scrutiny: Mass Constraints from CMB and Distance Probes

    astro-ph.CO 2026-07 conditional novelty 6.0

    Cosmological data place a lower bound near 10⁻²⁴ eV on spin-1 ultralight dark matter and predict a CMB anisotropy signature that may be detectable when the vector field is a minor dark-matter component.

Reference graph

Works this paper leans on

38 extracted references · 38 canonical work pages · cited by 2 Pith papers

  1. [1]

    Scalar sector For a perturbed energy momentum tensorδT µν, scalar effective fluid variables can be defined as: δρ=�δT 0 0 ,(20a) δP= 1 3 δT i i ,(20b) (ρ+P)θ=ik iδT 0 i ,(20c) (ρ+P)δΣ ∥ =� �ˆkiˆkj � 1 3 γj i � δΣi j ,(20d) whereδΣ ij =δT ij �δT kkδij/3 is the anisotropic stress. For the VFDM, the fluid variables in Newtonian gauge are given by δρA = 1 a4 ...

    work page

  2. [2]

    The Weinberg construction of the adiabatic mode ex- cludes vector perturbations

    Vector sector The Einstein vector equations can be obtained by pro- jectingδG 0i in the directions ˆe1 and ˆe2, thus obtaining respectively � m2 m2 P A2 t� + k2 2 � V1 = m2At� m2 P δA0 +S V ,(28) k 2 V2 =�2iσ v� h× � i a2m2 P A′ LδAt� ,(29) whereS V is SV =�ik 1 � 2 � (3� 2 � � ′)σv� +�σ ′ v� � ψ(30) �ikσ v� ϕ�ik σv� � ψ′ . The Weinberg construction of th...

    work page

  3. [3]

    Tensor sector The tensor sector is obtained by projecting Einstein equations with Λil jm =P ilPjm � 1 2 P i j P l m ,(31) whereP ij = (γ ij � ˆkiˆkj). The equations of motion for the two polarizations of the tensor perturbations are then calculated by contracting the resulting equations in the directions ˆe1 i ˆe1 j and ˆe1 i ˆe2 j. By doing this on both ...

    work page

  4. [4]

    spatial metric

    Therefore, by settingw × = 0, we have thath × = 0 for the whole evolution. This behavior is confirmed by our numerical results, although we do not show it for simplicity, as the amplitude remains negligible at all times. In Fig. 2 we present the gravitational wave abundance at present-day for different masses of the vector field and forγ k =π/2. We note t...

    work page

  5. [5]

    Scalar sector Now we present the scalar sector of Einstein’s equa- tions in Bianchi I background with VFDM. The Einstein tensor at linear order in the metric shear can be written as δG0(S) 0 = 2k2η� �h ′ + 3σ∥η′ + 2σ+h′ + ,(B5a) +iσ v� E1 ˆkiδG0(S) i = 2k2η′ �3k 2σ∥η�k 2σ+h+ ,(B5b) δGi(S) i = 3h′′ + 6�h ′ �6k 2η+ 18σ +h′ + (B5c) + 9ikσv� E′ 1 , δG(S) ∥ = ...

    work page

  6. [6]

    For the vector sector we consider that the VFDM is the only species that sources the vector modes

    Vector sector We present the vector sector in synchronous gauge for completeness, although it is not implemented in CLASS. For the vector sector we consider that the VFDM is the only species that sources the vector modes. As in newto- nian gauge, the Einstein vector equations can be obtained by projectingδG 0i in the directions ˆe1 and ˆe2, thus ob- taini...

    work page

  7. [7]

    Tensor sector The tensor perturbations are gauge invariant, so their equations of motion in synchronous gauge are still given by Eqs. (32). The sourcesH + andA +, however, need to be written in terms of the variables in synchronous gauge, as H+ = σ+ � η′′ + σ+ 2 h′ �2 � (1� � ′ � 2 )σ+ + σ′ + � � η′ (B10) + � � −1(σ′′ + + (k2 �2� ′)σ+)� � −2(� ′′σ+ + 2� ′...

    work page

  8. [8]

    This implementation is per- formed in Synchronous gauge, and is based on Ref

    Background In this appendix, we describe the implementation of the VFDM model in CLASS, including the mixing of the scalar and tensor sector. This implementation is per- formed in Synchronous gauge, and is based on Ref. [12]. In here, we present the equations that need to be modified to include the tensor sector. We start by reviewing the background imple...

    work page

  9. [9]

    The tensor perturbations are only affected by the ˆe 1 compo- nents of the vector field

    Linear perturbations Now we present the modifications carried out in the linear regime to include the tensor perturbations. The tensor perturbations are only affected by the ˆe 1 compo- nents of the vector field. We therefore focus on the im- plementation of the transversal Fourier modes, since the longitudinal ones where unmodified with respect to [12]. ...

    work page

  10. [10]

    Planck 2018 results. vi. cosmological parameters,

    P. collaboration, “Planck 2018 results. vi. cosmological parameters,”Astronomy & Astrophysics, vol. 641, p. A6, sep 2020

    work page 2018

  11. [11]

    Ultra-light dark matter,

    E. G. M. Ferreira, “Ultra-light dark matter,”The As- tronomy and Astrophysics Review, vol. 29, sep 2021

    work page 2021

  12. [12]

    Axion cosmology,

    D. J. Marsh, “Axion cosmology,”Physics Reports, vol. 643, pp. 1–79, jul 2016

    work page 2016

  13. [13]

    Vector Coherent Oscillation Dark Mat- ter,

    K. Nakayama, “Vector Coherent Oscillation Dark Mat- ter,”JCAP, vol. 10, p. 019, 2019

    work page 2019

  14. [14]

    Misalignment mechanism for a mass-varying vector boson,

    K. Kaneta, H.-S. Lee, J. Lee, and J. Yi, “Misalignment mechanism for a mass-varying vector boson,”JCAP, vol. 09, p. 017, 2023

    work page 2023

  15. [15]

    Viable vector coherent oscillation dark matter,

    N. Kitajima and K. Nakayama, “Viable vector coherent oscillation dark matter,”JCAP, vol. 07, p. 014, 2023

    work page 2023

  16. [16]

    A Lower Bound on Dark Matter Mass,

    M. A. Amin and M. Mirbabayi, “A Lower Bound on Dark Matter Mass,”Phys. Rev. Lett., vol. 132, no. 22, p. 221004, 2024

    work page 2024

  17. [17]

    Axion dark matter, soli- tons and the cusp–core problem,

    D. J. E. Marsh and A.-R. Pop, “Axion dark matter, soli- tons and the cusp–core problem,”Mon. Not. Roy. As- tron. Soc., vol. 451, no. 3, pp. 2479–2492, 2015

    work page 2015

  18. [18]

    Strong Bound on Canoni- cal Ultralight Axion Dark Matter from the Lyman-Alpha Forest,

    K. K. Rogers and H. V. Peiris, “Strong Bound on Canoni- cal Ultralight Axion Dark Matter from the Lyman-Alpha Forest,”Phys. Rev. Lett., vol. 126, no. 7, p. 071302, 2021

    work page 2021

  19. [19]

    Isotropy theorem for cosmological vector fields,

    J. A. R. Cembranos, C. Hallabrin, A. L. Maroto, and S. J. N. Jareno, “Isotropy theorem for cosmological vector fields,”Phys. Rev. D, vol. 86, p. 021301, 2012

    work page 2012

  20. [20]

    Isotropy theorem for arbitrary-spin cosmolog- ical fields,

    J. A. R. Cembranos, A. L. Maroto, and S. J. N´ u˜ nez Jare˜ no, “Isotropy theorem for arbitrary-spin cosmolog- ical fields,”JCAP, vol. 03, p. 042, 2014

    work page 2014

  21. [21]

    Cosmological perturbations with ultralight vector dark matter fields: Numerical implementation in class,

    T. F. Chase, M. Leizerovich, D. L´ opez Nacir, and S. Lan- dau, “Cosmological perturbations with ultralight vector dark matter fields: Numerical implementation in class,” Phys. Rev. D, vol. 111, no. 10, p. 103520, 2025

    work page 2025

  22. [22]

    Ultralight vector dark 15 matter, anisotropies, and cosmological adiabatic modes,

    T. F. Chase and D. L´ opez Nacir, “Ultralight vector dark 15 matter, anisotropies, and cosmological adiabatic modes,” Phys. Rev. D, vol. 109, no. 8, p. 083521, 2024

    work page 2024

  23. [23]

    Adiabatic modes in cosmology,

    S. Weinberg, “Adiabatic modes in cosmology,”Physical Review D, vol. 67, jun 2003

    work page 2003

  24. [24]

    Damping of tensor modes in cosmology,

    S. Weinberg, “Damping of tensor modes in cosmology,” Phys. Rev. D, vol. 69, p. 023503, 2004

    work page 2004

  25. [25]

    Theory of cosmological perturbations in an anisotropic uni- verse,

    T. S. Pereira, C. Pitrou, and J.-P. Uzan, “Theory of cosmological perturbations in an anisotropic uni- verse,”Journal of Cosmology and Astroparticle Physics, vol. 2007, pp. 006–006, sep 2007

    work page 2007

  26. [26]

    Cosmological perturba- tion theory in the synchronous and conformal Newtonian gauges,

    C.-P. Ma and E. Bertschinger, “Cosmological perturba- tion theory in the synchronous and conformal Newtonian gauges,”Astrophys. J., vol. 455, pp. 7–25, 1995

    work page 1995

  27. [27]

    Gravitational waves from stochastic relativistic sources: Primordial turbulence and magnetic fields,

    C. Caprini and R. Durrer, “Gravitational waves from stochastic relativistic sources: Primordial turbulence and magnetic fields,”Phys. Rev. D, vol. 74, p. 063521, 2006

    work page 2006

  28. [28]

    Gravitational waves from self-ordering scalar fields,

    E. Fenu, D. G. Figueroa, R. Durrer, and J. Garcia- Bellido, “Gravitational waves from self-ordering scalar fields,”JCAP, vol. 10, p. 005, 2009

    work page 2009

  29. [29]

    Relativistic viscous effects on the primordial gravitational waves spectrum,

    N. Mir´ on-Granese, “Relativistic viscous effects on the primordial gravitational waves spectrum,”JCAP, vol. 06, p. 008, 2021

    work page 2021

  30. [30]

    The cosmic linear anisotropy solving system (class). part ii: Approxima- tion schemes,

    D. Blas, J. Lesgourgues, and T. Tram, “The cosmic linear anisotropy solving system (class). part ii: Approxima- tion schemes,”Journal of Cosmology and Astroparticle Physics, vol. 2011, no. 07, p. 034, 2011. [22]���������������������������������������

    work page 2011

  31. [31]

    Imprint of ultralight vector fields on gravitational wave propagation,

    A. D. Miravet and A. L. Maroto, “Imprint of ultralight vector fields on gravitational wave propagation,”Phys. Rev. D, vol. 103, no. 12, p. 123546, 2021

    work page 2021

  32. [32]

    Gravitational-wave sensitivity curves,

    C. J. Moore, R. H. Cole, and C. P. L. Berry, “Gravitational-wave sensitivity curves,”Class. Quant. Grav., vol. 32, no. 1, p. 015014, 2015

    work page 2015

  33. [33]

    Improved Constraints on Primor- dial Gravitational Waves using Planck, WMAP, and BI- CEP/Keck Observations through the 2018 Observing Season,

    P. A. R. Adeet al., “Improved Constraints on Primor- dial Gravitational Waves using Planck, WMAP, and BI- CEP/Keck Observations through the 2018 Observing Season,”Phys. Rev. Lett., vol. 127, no. 15, p. 151301, 2021

    work page 2018

  34. [34]

    Constraints on primordial gravitational waves from the Cosmic Mi- crowave Background,

    T. J. Clarke, E. J. Copeland, and A. Moss, “Constraints on primordial gravitational waves from the Cosmic Mi- crowave Background,”JCAP, vol. 10, p. 002, 2020

    work page 2020

  35. [35]

    Tensor perturbations in inflationary models as a probe of cos- mology,

    M. S. Turner, M. J. White, and J. E. Lidsey, “Tensor perturbations in inflationary models as a probe of cos- mology,”Phys. Rev. D, vol. 48, pp. 4613–4622, 1993

    work page 1993

  36. [36]

    Gravitational wave experiments and early universe cosmology,

    M. Maggiore, “Gravitational wave experiments and early universe cosmology,”Phys. Rept., vol. 331, pp. 283–367, 2000

    work page 2000

  37. [37]

    xPand: An algorithm for perturbing homogeneous cosmologies,

    C. Pitrou, X. Roy, and O. Umeh, “xPand: An algorithm for perturbing homogeneous cosmologies,”Class. Quant. Grav., vol. 30, p. 165002, 2013

    work page 2013

  38. [38]

    xPert: Computer algebra for metric perturbation theory,

    D. Brizuela, J. M. Martin-Garcia, and G. A. Mena Maru- gan, “xPert: Computer algebra for metric perturbation theory,”Gen. Rel. Grav., vol. 41, pp. 2415–2431, 2009

    work page 2009