The observational status of Galileon gravity after Planck

Abstract

We use the latest CMB data from Planck, together with BAO measurements, to constrain the full parameter space of Galileon gravity. We constrain separately the three main branches of the theory known as the Cubic, Quartic and Quintic models, and find that all yield a very good fit to these data. Unlike in ΛCDM, the Galileon model constraints are compatible with local determinations of the Hubble parameter and predict nonzero neutrino masses at over 5σ significance. We also identify that the low-l part of the CMB lensing spectrum may be able to distinguish between ΛCDM and Galileon models. In the Cubic model, the lensing potential deepens at late times on sub-horizon scales, which is at odds with the current observational suggestion of a positive ISW effect. Compared to ΛCDM, the Quartic and Quintic models predict less ISW power in the low-l region of the CMB temperature spectrum, and as such are slightly preferred by the Planck data. We illustrate that residual local modifications to gravity in the Quartic and Quintic models may render the Cubic model as the only branch of Galileon gravity that passes Solar System tests.

Breaking the Vainshtein screening in clusters of galaxies

“…Moreover, we do not believe that the distances might help to constrain such parameters: cosmological geometrical probes are generally very weak when used to compare ΛCDM (GR) with other alternative models; we have no observational errors on these distances which, thus, in principle, could re-scale in a completely free and un-physical way. Finally, in [13] it is shown that the Hubble function H(z) derived from their galileon model can differ from the expected ΛCDM behaviour for 5% in the redshift range covered by our data; a variation which is smaller than present observational errors and dispersion and, thus, still not detectable in a statistically valid way nowadays. It is worth to stress that results from [13] are obtained using only Planck CMB and some BAO data, which very likely pinpoint high redshifts regime behaviour quite well, but not equally well the lower one.…”

“…Finally, in [13] it is shown that the Hubble function H(z) derived from their galileon model can differ from the expected ΛCDM behaviour for 5% in the redshift range covered by our data; a variation which is smaller than present observational errors and dispersion and, thus, still not detectable in a statistically valid way nowadays. It is worth to stress that results from [13] are obtained using only Planck CMB and some BAO data, which very likely pinpoint high redshifts regime behaviour quite well, but not equally well the lower one. The fit would have surely benefit (and maybe reduced the deviation from the baseline ΛCDM background) from considering two further elements: SNeIa, which are well known to play a complementary role with respect to BAO in fixing many cosmological parameters; and by applying a prior on H 0 from independent observations, given that present errors on H 0 are ∼ 2% [90], which is less than half the deviation from ΛCDM depicted in the same [13].…”

Beyond the cosmological standard model

“…Cosmological parameter analyses to test MG theories have been performed by several authors [340,884,895,896,903,[908][909][910][911][912][913]. Specific studies to determine the cosmological impact of galileon theories and to constrain model parameters appear in [265,[914][915][916][917][918][919][920].…”

Speed of gravitational waves and the fate of scalar-tensor gravity

“…However, no distant GW-EM event will possibly be observed if c g is modified significantly, since the delay between both signals will be much larger than the monitoring time around the GW detection. This is the case of cosmic acceleration models without a cosmological constant such as covariant Galileons [52,62], for which |c g /c − 1| ∼ 10 − 100% (see [57] and Fig. 1 of Ref.…”