Extracting cosmological signals from foregrounds in deep mm maps of the sky

TL;DR

This study evaluates a balloon-borne mm/submm survey's ability to extract cosmological signals like CMB, SZ effect, and infrared background from simulated sky maps, demonstrating effective component separation and detection capabilities.

Contribution

It presents a detailed simulation framework and performance analysis for a high-sensitivity balloon-borne survey targeting cosmological signals in mm/submm sky maps.

Findings

CMB power spectrum accurately recovered up to l ~ 3000

Detected 60% of galaxy clusters via SZ effect

Stronger constraints on far infrared background models

Abstract

The high Galactic latitude sky at millimeter and submm wavelengths contains significant cosmological information about the early Universe (in terms of the cosmic microwave background) but also the process of structure formation in the Universe from the far infrared background produced by early galaxies and the Sunyaev-Zeldovich effect in clusters of galaxies. As the Planck mission will produce full sky maps in this frequency range, deeper maps of selected low-foregrounds patches of the sky can produce complementary and important information. Here we analyze the performance of a balloon-borne survey covering a 10^\circ x 10^\circ patch of the sky with a few arcminute resolution and very high pixel sensitivity. We simulate the different components of the mm/submm sky (i.e., CMB anisotropies, SZ effect, radio and infrared sources, far infrared background, and interstellar dust) using current knowledge about each of them. We then combine them, adding detector noise, to produce detailed simulated observations in four observational bands ranging from 130 to 500 GHz. Finally, we analyze the simulated maps and estimate the performance of the instrument in extracting the relevant information about each of the components. We find that the CMB angular power spectrum is accurately recovered up to l ~ 3000. Using the Sunyaev-Zel'dovich effect, most of the galaxy clusters present in our input map are detected (60% efficiency overall). Our results also show that much stronger constrains can be placed on far infrared background models.