Planck Constraints on Monodromy Inflation

TL;DR

Planck data are used to test oscillatory modulations in the primordial power spectrum predicted by axion monodromy inflation. The authors implement a grid-based approach to compute the modulated spectrum, varying parameters μ, δn_s, f, and ϕ, and evaluate Planck likelihoods against the LCDM baseline. Their results show no statistically significant evidence for the modulation, with best-fit improvements of $Δχ^2_{eff} \approx 8$–$10$ and Bayes factor $Δ\ln E \approx 1.5$, indicating Planck is consistent with cosmic variance and disfavors the stronger WMAP9 hint. The work emphasizes discrepancies between Planck and WMAP9 for such signals and calls for further theoretical and observational scrutiny of monodromy inflation predictions and potential systematics. It constrains the single-field EFT realization of monodromy inflation and motivates future analyses with additional Planck data and alternative monodromy potentials.

Abstract

We use data from the nominal Planck mission to constrain modulations in the primordial power spectrum associated with monodromy inflation. The largest improvement in fit relative to the unmodulated model has Δχ^2~10 and we find no evidence for a primordial signal, in contrast to a previous analysis of the WMAP9 dataset, for which Δχ^2~20. The Planck and WMAP9 results are broadly consistent on angular scales where they are expected to agree as far as best-fit values are concerned. However, even on these scales the significance of the signal is reduced in Planck relative to WMAP, and is consistent with a fit to the noise associated with cosmic variance. Our results motivate both a detailed comparison between the two experiments and a more careful study of the theoretical predictions of monodromy inflation.

Figures (6)

• Figure 1: Marginalized posterior distributions for inflationary parameters derived from Planck (top) and WMAP (bottom). The WMAP plot reproduces the results of Peiris:2013opa using the methods and notation of the present analysis. The Planck analysis is performed over the full $\ell$ range, combined with the WMAP polarization data on large scales. The shaded region indicates values of $f$ for which the single field effective field becomes strongly coupled and a more careful study of the underlying stringy model may be necessary to see if the predictions of the model change qualitatively.
• Figure 2: Marginalized posterior distributions for inflationary parameters from Planck with $\ell_{\mathrm{max}}$ of 900 (top) and 600 (bottom).
• Figure 3: Marginalized posterior distributions for inflationary parameters for larger values of the axion decay constant. The full angular power spectrum is used in the top panel, while the lower panel has $\ell_{\mathrm{max}}= 600$.
• Figure 4: Difference in $\chi^2_\text{eff}$ between the best-fit point near $f=4.4\times10^{-4}M_{\rm Pl}$ and $f=3.1\times 10^{-3}M_{\rm Pl}$ in the top and bottom half of the Figure derived from the blocks of the CAMspec covariance matrix for the different frequency channels.
• Figure 5: Comparison of data used in CAMspec binned with $\Delta\ell=25$ with the smooth best-fit model (including the model for extragalactic foregrounds). The $100$ GHz data is shown in orange, the $143$ GHz data in red, the $217$ GHz data in blue, and finally the $143\times 217$ cross spectra in green.