Dealing with Dark Energy

Abstract

Discoveries in the last few years have revolutionized our knowledge of the universe and our ideas of its ultimate fate. Measurements of the expansion of the universe show that it is not slowing down under normal gravity but accelerating due to an unknown, gravitationally repulsive "dark energy". This may be a clue to new properties of quantum physics or of gravity beyond Einstein. I present an overview of the puzzles of dark energy and the means for unraveling them through cosmological probes, on both a generally accessible and a technical level. I also highlight the strong benefits of meshing supernova distance and weak lensing methods. Next generation experiments such as the Supernova/Acceleration Probe (SNAP) satellite would measure the supernova distance-redshift relation to high accuracy and map the evolution of structure and dark matter through gravitational lensing. These observations will explore the frontiers of physics and aim to uncover what makes up the still unknown 95% of our universe.

Dark energy in the dark ages

“…5% of that epoch's Hubble time, changes the present growth by more than 10%. This tightly constrains models of oscillating or stochastic dark energy [18,19], which have intermediate epochs of dark age dark energy domination 2 . Even recent transitions, z u > 0.5, would still show 10% deviations for what might be considered short durations ∆ ln a = 0.3 ≪ 1.…”

Separating dark physics from physical darkness: Minimalist modified gravity versus dark energy

“…γ = 0.55 for general relativity and a cosmological constant model); however this will bias the other parameters due to their covariances with γ (this "gravity's bias" was illustrated for the linear growth factor alone, rather than the weak lensing shear power spectrum, in Fig. 5 of [25]). …”

Probing dark energy with supernovae: A concordant or a convergent model?

“…models with w 1 > w 0 + 1 will end in a Big Rip [17]). Models with very exotic w(z) may come from modified gravity [22]. The class of models with w 1 < 0 is roughly excluded at 95%CL, if the strong prior Ω M = 0.27 ± 0.04 is used [2], but is perfectly allowed for higher Ω M values (or larger prior errors).…”

“…This "bias problem" has been mentioned several times in the literature, see e.g. [4,5,6,7]. In this letter, we explore the effect of the Ω M prior on the determination of w(z).…”