Simulations of the Microwave Sky

We create realistic, full-sky, half-arcminute resolution simulations of the microwave sky matched to the most recent astrophysical observations. The primary purpose of these simulations is to test the data reduction pipeline for the Atacama Cosmology Telescope (ACT) experiment; however, we have widened the frequency coverage beyond the ACT bands to make these simulations applicable to other microwave background experiments. Some of the novel features of these simulations are that the radio and infrared galaxy populations are correlated with the galaxy cluster populations, the CMB is lensed by the dark matter structure in the simulation via a ray-tracing code, the contribution to the thermal and kinetic Sunyaev-Zeldovich (SZ) signals from galaxy clusters, groups, and the IGM has been included, and the gas prescription to model the SZ signals matches the most recent X-ray observations. Regarding the contamination of cluster SZ flux by radio galaxies, we find for 148 GHz (90 GHz) only 3% (4%) of halos have their SZ decrements contaminated at a level of 20% or more. We find the contamination levels higher for infrared galaxies. However, at 90 GHz, less than 20% of clusters with M_{200} > 2.5 x 10^{14} Msun and z<1.2 have their SZ decrements filled in at a level of 20% or more. At 148 GHz, less than 20% of clusters with M_{200} > 2.5 x 10^{14} Msun and z<0.8 have their SZ decrements filled in at a level of 50% or larger. Our models also suggest that a population of very high flux infrared galaxies, which are likely lensed sources, contribute most to the SZ contamination of very massive clusters at 90 and 148 GHz. These simulations are publicly available and should serve as a useful tool for microwave surveys to cross-check SZ cluster detection, power spectrum, and cross-correlation analyses.

Luminous Red Galaxy Halo Density Field Reconstruction and Application to Large Scale Structure Measurements

The complex relationship between the galaxy density field and the underlying matter field limits our ability to extract cosmological constraints from galaxy redshift surveys. Our approach is to use halos rather than galaxies to trace the underlying mass distribution. We identify Fingers-of-God (FOGs) and replace multiple galaxies in each FOG with a single halo object. This removes the nonlinear contributions of satellite galaxies, the one-halo term. We test our method on a large set of high-fidelity mock SDSS Luminous Red Galaxy (LRG) catalogs. We find that the aggressive FOG compression algorithm adopted in the LRG P(k) analysis of Tegmark et al. (2006) leads to a ~10% correction to the underlying matter power spectrum at k = 0.1 h/Mpc and ~40% correction at k=0.2 h/Mpc, thereby compromising the cosmological constraints. In contrast, the power spectrum of our reconstructed halo density field deviates from the underlying matter power spectrum by less than 1% for k less than 0.1 h/Mpc and less than 4% for k less than 0.2 h/Mpc. The reconstructed halo density field also removes the bias in the measurement of the redshift space distortion parameter beta induced by the FOG smearing of the linear redshift space distortions.