CMB Polarization and Theories of Gravitation with Massive Gravitons

We study in this paper three different theories of gravitation with massive gravitons - the modified Fierz-Pauli (FP) model, Massive Gravity and the bimetric theory proposed by Visser - in linear perturbation theory around a Minkowski and a flat FRW background. For the TT tensor perturbations we show that the three theories give rise to the same dynamical equations and to the same form of the Boltzmann equations for the radiative transfer in General Relativity (GR). We then analyze vector perturbations in these theories and show that they do not give the same results as in the previous case. We first show that vector perturbations in Massive Gravity present the same form as found in General Relativity, whereas in the modified FP theory the vector gravitational-wave (GW) polarization modes ($\Psi_{3}$ amplitudes in the Newman-Penrose (NP) formalism) do not decay too fast as it happens in the former case. Rather, we show that such $\Psi_{3}$ polarization modes give rise to an unusual vector Sachs-Wolfe effect, leaving a signature in the quadrupole form $Y_{2,\pm 1}(\theta,\varphi)$ on the CMB polarization. We then derive the details for the Thomson scattering of CMB photons for these $\Psi_{3}$ modes, and then construct the correspondent Boltzmann equations. Based upon these results we then qualitatively show that $\Psi_{3}$-mode vector signatures - if they do exist - could clearly be distinguished on the CMB polarization from the usual $\Psi_4$ tensor modes.

💡 Research Summary

The paper investigates three distinct massive‑graviton theories— a modified Fierz‑Pauli (FP) model, Massive Gravity (MG), and Visser’s bimetric theory— within linear perturbation theory on both Minkowski and flat Friedmann‑Robertson‑Walker (FRW) backgrounds. The authors first review the Newman‑Penrose (NP) formalism, which classifies gravitational‑wave (GW) polarizations into six independent Newman‑Penrose amplitudes: two scalar (Ψ₂, Φ₂₂), two vector (Ψ₃, \barΨ₃), and two tensor (Ψ₄, \barΨ₄) modes. These amplitudes are mapped onto a set of six polarization basis matrices Eᵣ, providing a compact representation of any metric perturbation.

For each massive‑graviton theory the field equations are derived in the weak‑field limit. In the modified FP model a mass term m²(h_{αβ}h^{αβ}−½h²) is added to the linearized Einstein‑Hilbert action. Conservation of the stress‑energy tensor yields the constraint ∂α \bar h^{αβ}=0, which eliminates four gauge degrees of freedom and leaves exactly six propagating modes, matching the NP classification. The resulting equation for the spatial metric perturbation h{ij} is a Klein‑Gordon equation (□+m²)h_{ij}=0. Visser’s bimetric theory introduces a nondynamical background metric f_{μν} and a mass term built from both g_{μν} and f_{μν}. In the weak‑field regime this theory reduces to the same equations as the modified FP model, confirming their equivalence at linear order. Massive Gravity, following the de Rham‑Gabadadze‑Tolley construction, splits the quadratic mass term into five independent pieces, each multiplied by a mass parameter proportional to a common scale m. By appropriate choices of these parameters one recovers the FP dynamics; in general, however, the vector sector behaves differently.

All three theories predict identical dynamics for the transverse‑traceless (TT) tensor sector (Ψ₄, \barΨ₄). Consequently, the tensor Sachs‑Wolfe effect and the associated Boltzmann equations for radiative transfer are exactly the same as in General Relativity (GR). The novelty arises in the vector sector (Ψ₃, \barΨ₃). In Massive Gravity the vector modes decay with the same rate as in GR, rendering them observationally negligible. In contrast, the modified FP model yields vector modes that decay only as a power law (∝1/t) rather than exponentially, provided the graviton mass is sufficiently small. These slowly‑decaying Ψ₃ modes generate an “unusual vector Sachs‑Wolfe effect”: they imprint a quadrupolar pattern proportional to the spherical harmonics Y_{2,±1}(θ,φ) on the Cosmic Microwave Background (CMB) temperature anisotropy and, more importantly, on its polarization.

The authors then compute the Thomson scattering of CMB photons in the presence of these Ψ₃ modes. By projecting the perturbed metric onto the polarization basis and solving the Boltzmann hierarchy, they derive modified evolution equations for the Stokes parameters. The vector modes source an E‑mode polarization pattern with the distinctive Y_{2,±1} angular dependence, whereas the usual tensor modes (Ψ₄) generate the familiar Y_{2,±2} pattern. Because the vector‑induced E‑mode has a different azimuthal structure, it can be disentangled from the tensor contribution in high‑precision polarization data.

A quantitative estimate of the graviton mass limit is provided: m ≈ 10⁻⁶⁶ g ≈ 10⁻²⁹ cm⁻¹. For masses below this threshold the vector modes evolve indistinguishably from massless gravitons, while for larger masses the characteristic Ψ₃ signature becomes suppressed. The paper argues that forthcoming CMB polarization experiments (e.g., CMB‑S4, LiteBIRD, PICO) with exquisite sensitivity to the E‑mode spectrum could detect or constrain such vector signatures, thereby offering a novel test of massive‑graviton theories.

In summary, the work demonstrates that while massive‑graviton theories share the same tensor‑mode phenomenology as GR, they can differ dramatically in the vector sector. The presence of a slowly‑decaying Ψ₃ mode leads to a unique quadrupolar imprint on the CMB polarization, providing a concrete observational target. Detection of this signature would constitute direct evidence for non‑GR graviton dynamics, whereas its absence would tighten bounds on graviton mass and on the viable parameter space of massive‑gravity models.