Investigating dark energy experiments with principal components
Abstract
We use a principal component approach to contrast different kinds of probes of dark energy, and to emphasize how an array of probes can work together to constrain an arbitrary equation of state history w(z). We pay particular attention to the role of the priors in assessing the information content of experiments and propose using an explicit prior on the degree of smoothness of w(z) that is independent of the binning scheme. We also show how a figure of merit based on the mean squared error probes the number of new modes constrained by a data set, and use it to examine how informative various experiments will be in constraining the evolution of dark energy.
Keywords
Cite
@article{arxiv.astro-ph/0510293,
title = {Investigating dark energy experiments with principal components},
author = {Robert G. Crittenden and Levon Pogosian and Gong-Bo Zhao},
journal= {arXiv preprint arXiv:astro-ph/0510293},
year = {2010}
}
Comments
A significantly expanded version with an added PCA for weak lensing, a new detailed discussion of the correlation prior proposed in this work, and a new discussion outlining the differences between the Bayesian and the frequentist approaches to reconstructing w(z). Matches the version accepted to JCAP. 8 pages, 2 figures