Professor Pedro Ferreira

Abstract:

A possible solution to the dark energy problem is that Einstein's theory of general relativity is modified. A suite of models have been proposed that, in general, are unable to predict the correct amount of large scale structure in the distribution of galaxies or anisotropies in the Cosmic Microwave Background. It has been argued, however, that it should be possible to constrain a general class of theories of modified gravity by focusing on properties such as the growing mode, gravitational slip and the effective, time varying Newton's constant. We show that assuming certain physical requirements such as stability, metricity and gauge invariance, it is possible to come up with consistency conditions between these various parameters. In this paper we focus on theories which have, at most, 2nd derivatives in the metric variables and find restrictions that shed light on current and future experimental constraints without having to resort to a (as yet unknown) complete theory of modified gravity. We claim that future measurements of the growth of structure on small scales (i.e. from 1-200 h^{-1} Mpc) may lead to tight constraints on both dark energy and modified theories of gravity.

Abstract:

We present a comprehensive investigation of cosmological constraints on the class of vector field formulations of modified gravity called Generalized Einstein-Aether models. Using linear perturbation theory we generate cosmic microwave background and large-scale structure spectra for general parameters of the theory, and then constrain them in various ways. We investigate two parameter regimes: a dark-matter candidate where the vector field sources structure formation, and a dark-energy candidate where it causes late-time acceleration. We find that the dark matter candidate does not fit the data, and identify five physical problems that can restrict this and other theories of dark matter. The dark energy candidate does fit the data, and we constrain its fundamental parameters; most notably we find that the theory's kinetic index parameter $n_{\mathrm{ae}}$ can differ significantly from its $\Lambda$CDM value.