Nonlinear backreaction in cosmology

Abstract

In this talk I discuss the effects of nonlinear backreaction of small scale density inhomogeneities in general relativistic cosmology. It has been proposed that in an inhomogeneous universe, nonlinear terms in the Einstein equation could, if properly averaged and taken into account, affect the large scale Friedmannian evolution of the universe. In particular, it was hoped that these terms might mimic a cosmological constant and eliminate the need for dark energy. After reviewing some of these approaches, and some of their flaws, I will describe a perturbative framework (developed with R. Wald) designed to properly take into account these effects. In our framework, we assume that the spacetime metric is "close"---within 1 part in 10^4, except near strong field objects---to a background metric of FLRW symmetry, but we do not assume that the background metric satisfies the Friedmann equation. We also do not require that spacetime derivatives of the metric be close to derivatives of the background metric. This allows for significant deviations in geodesics, and very large curvature inhomogeneities. A priori, this framework also allows for significant backreaction, which would take the form of new effective matter sources in the Friedmann equation. Nevertheless, we prove that if the matter stress-energy tensor satisfies the weak energy condition, then large matter inhomogeneities on small scales cannot produce significant backreaction effects on large scales, and in particular cannot account for dark energy. As I will also review here, with a suitable ‘dictionary,’ Newtonian cosmologies provide excellent approximations to cosmological solutions to Einsteinʼs equation (with dust and a cosmological constant) on all scales. Our results thereby provide strong justification for the mathematical consistency and validity of the LCDM model within the context of general relativistic cosmology. While our rigorous framework makes use of 1-parameter families and weak limits, in this talk I will provide a simple heuristic discussion that places emphasis on the manner in which "averaging" is done, and the fact that one is solving the Einstein equation.