Cyberinfrastructure for Cosmology and Line-of-Sight Projection in Optical Galaxy Clusters.

Abstract

Upcoming wide-area sky surveys such as the Dark Energy Survey (DES) offer the power to test the source of cosmic acceleration by placing extremely precise constraints on existing cosmological model parameters. These observational surveys will employ multiple tests based on statistical signatures of galaxies and larger-scale structures such as clusters of galaxies. Simulations of large-scale structure provide the means to maximize the power of sky survey tests by characterizing key sources of systematic uncertainties. This dissertation explores two subjects motivated by these facts. First, it explores how grid-aware cyberinfrastructure needs to be utilized in current and upcoming simulation campaigns that support large-area sky surveys. Second, it shows how line-of-sight projection plays into cosmological analysis based on galaxy cluster counts in the same wide-area sky surveys. In the first part, an Apache Airavata-enabled grid-aware application workflow for managing simulations is described. Results pertaining to efficiency in producing N-body simulations are reported. In the second part, bias in cosmological parameter estimates caused by incorrectly assuming a Gaussian (projection-free) mass--observable relation when the true relation is non-Gaussian due to projection is explored. Projection tends to skew the mass--observable relation of galaxy clusters by creating a small fraction of severely blended systems, those for which the measured observable property of a cluster is strongly boosted relative to the value of its primary host halo. A model motivated by optical cluster-finding applied to the Millennium Simulation is introduced for projection and Fisher information matrix parameter bias forecasts are produced for a DES-like sky survey. The model predicts significant biases in the dark energy density and equation of state parameters. The model additional predicts an increase in uncertainties in dark energy parameters to a factor of about two larger than forecast uncertainties. Additionally, new parameters used to characterize the model degrade uncertainties in the dark energy parameters. Motivated by this result, this dissertation also contains preliminary results for a new projection model meant to reduce bias in cluster analysis based on redMaPPer identified clusters for the DES.