Constraining primordial tensor features with the anisotropies of the cosmic microwave background

Abstract

It is commonly assumed that the stochastic background of gravitational waves on cosmological scales follows an almost scale-independent power spectrum, as generically predicted by the inflationary paradigm. However, it is not inconceivable that the spectrum could have strongly scale-dependent features, generated, e.g., via transient dynamics of spectator axion-gauge fields during inflation. Using the temperature and polarisation maps from the Planck and BICEP/Keck datasets, we search for such features, taking the example of a log-normal bump in the primordial tensor spectrum at CMB scales. We do not find any evidence for the existence of bump-like tensor features at present, but demonstrate that future CMB experiments such as LiteBIRD and CMB-S4 will greatly improve our prospects of determining the amplitude, location and width of such a bump. We also highlight the role of delensing in constraining these features at angular scales ℓ ≳ 100.

Measuring the spectrum of primordial gravitational waves with CMB, PTA and laser interferometers

“…In this work, we also check that all the parameter choices are consistent with observational constraints on the axion-SU(2) model from the analysis of current CMB datasets [115]. However, as also noted in [115], the shape of the tensor power spectrum is very weakly constrained by the Planck and BICEP/Keck data: the upper limits on the model parameters are prior dominated and strongly affected by degeneracies.…”

“…This allows, in principle, for a larger signal from the axion-SU(2) model at these scales. Second, we update the forecast with new bounds on the SU(2) parameter space from backreaction [92] and the current upper limits from the Planck and BICEP data [31,32,35,115]. Third, we show results for the full covariance matrix from realistic LiteBIRD simulations, instead of the simplified Fisher approach of [90].…”

“…10, instead, we compare ∆χ 2 for different choices of k p that can be probed by CMB observations, using the full covariance matrix. The white area in both figures represents the portion of parameter space excluded by current upper limits on the tensor-to-scalar ratio by Planck and BICEP/Keck data at each scale [31,32,35,115,126]. Note that we assume r * = 5r vac for all k p choices, except for k p = 5 × 10 −3 Mpc −1 for which we assume r * = r vac , following Ref.…”

LiteBIRD science goals and forecasts. A case study of the origin of primordial gravitational waves using large-scale CMB polarization

“…Figure 1 also shows for reference the cosmic variance-only (including primordial and lensing B-mode variance) and total LiteBIRD ±1 σ binned error bars (including foreground residuals) as gray and blue regions, respectively. In this work, we also check that all the parameter choices are consistent with observational constraints on the axion-SU(2) model from the analysis of current CMB datasets [115]. However, as also noted in [115], the shape of the tensor power spectrum is very weakly constrained by the Planck and BICEP/Keck data: the upper limits on the model parameters are prior dominated and strongly affected by degeneracies.…”

Primordial feature constraints from BOSS + eBOSS

“…There is already a plethora of analyses using WMAP [32][33][34][35][36][37][38][39] and Planck [2,8,[40][41][42][43][44] datasets. Forecasts for future CMB surveys are present in recent works [45][46][47]. Gravitational waves are one of the main byproducts of inflation; the presence of PF could also affect the stochastic gravitational wave background [48][49][50][51][52][53], and [54] showed a forecast for LISA.…”