A theoretical framework for the detection of point-sources in Cosmic Microwave Background maps

TL;DR

This paper develops a statistical framework for evaluating point-source detection methods in CMB maps, providing analytical insights into their performance and limitations, and proposing a simple, effective detection algorithm for future experiments.

Contribution

It offers a comprehensive analytical analysis of detection techniques based on Neyman-Pearson and SNR criteria, filling a gap in the statistical understanding of their properties.

Findings

Analytical characterization of detection method performance

Comparison of techniques under different criteria

Development of a simple, effective detection algorithm

Abstract

The detection of point-sources in microwave maps is a critical step in the analysis of the Cosmic Microwave Background (CMB) data. If not properly removed, these sources have adverse effects on the estimation of the power-spectrum and/or the test of Gaussianity of the CMB component. The techniques to extract point sources from an observed image used up to now do not show a general consensus about their real performance and properties. Their characteristics have been studied essentially through numerical simulations based on semi-empirical models of the CMB and the Galactic foreground. Such models often have different levels of sophistication and/or are based on different physical assumptions (e.g. the number of Galactic components and level of the noise). Moreover, the application of a given technique to a set of data (either simulated or experimental) requires the tuning of one or more parameters that unavoidably is a subjective operation. Hence, a reliable comparison is difficult. What is missing is a statistical analysis of the properties of the proposed methodologies. This is the aim of the present paper. The statistical properties of the detection techniques in the context of two different criteria, i.e. the Neyman-Pearson criterion and the maximization of the signal-to-noise ratio, are analyzed through an analytical approach. Various examples, techniques and the conditions under which they work satisfactorily are shown together with their limits and implementation details. Exploiting some a priori information, it is possible to develop simple algorithms with performances similar to those of more sophisticated but complex techniques. In this respect, a detection algorithm, useful in future ground-based experiments as, for example the Atacama Large Millimeter/submillimeter Array, is presented.