Towards Constraints on Dark Energy from Absorption Spectra of Close Quasar Pairs

A comparison between the line of sight power spectrum of absorption in the Lyman-alpha forest and the cross power spectrum between the absorption in neighboring lines of sight offers an evolution-free means to constrain the cosmological constant, or dark energy. Using cosmological simulations, we consider a maximum likelihood method to obtain constraints from this comparison. In our method, measurements of the auto and cross spectra from observations are compared with those from a multi-parameter grid of simulated models of the intergalactic medium. We then marginalize over nuisance parameters to obtain constraints on the cosmological constant. Redshift space distortions due to peculiar velocities and thermal broadening, a potential difficulty in applying this test, are explicitly modeled in our simulations. To illustrate our method, we measure the cross spectrum from a new sample of five close quasar pairs, with separations of 0.5 to 3 arcmin. We attempt to obtain a constraint on Omega_Lambda, but find only weak constraints. An Einstein-de-Sitter cosmology is, however, disfavored by the data at a ~ 2 sigma confidence level. We consider the power of future observations, paying particular attention to the effects of spectral resolution and shot-noise. We find that ~ 50 moderate resolution, FWHM ~ 150 km/s, close separation (~ 30-120 arcsec) pairs should allow a (2 sigma) constraint on Omega_Lambda at the level of 15%, if other modeling parameters are determined through other means. We find that there is a sizeable gain from observing very close, ~ 30 arcsec, separation pairs provided they are observed with high spectral resolution. A sample of ~ 10 such pairs gives similar constraints to the 50 moderate resolution pairs mentioned above.