On the cosmological backreaction for large distance modifications of gravity

TL;DR

This paper investigates whether large-scale cosmological modifications of gravity, which include nonlinear mechanisms to recover General Relativity near matter sources, lead to significant backreaction effects, finding that certain symmetries can suppress such effects.

Contribution

It demonstrates that in some modified gravity theories with shift or Galilean symmetries, the expected large backreaction from nonlinearities can be small, challenging previous assumptions.

Findings

Shift symmetry can suppress large backreaction effects.

DGP-like and f(G) models exhibit small averaged nonlinear effects.

Large small-scale nonlinearities do not necessarily imply large large-scale backreaction.

Abstract

Every theory that modifies gravity at cosmological distances and that is not already ruled out by the Solar system observations must exhibit some nonlinear mechanism that turns off the modification close to a compact matter source. Given this nonlinearity, one might expect such a theory to show a large gravitational backreaction, i.e. an order one influence of the small scale inhomogeneities on the large scale evolution of the Universe. We argue that this is not necessarily the case. If the dominant nonlinear terms in the equations obey a shift symmetry, the averaged effect of the nonlinearities can be small, although the effect on small scales is large. This happens for DGP (-like) modifications and so called f(G) (or Gauss-Bonnet) models. For both type of models the shift symmetry is part of the larger "Galilean" symmetry.