Multi-Scale Pipeline for the Search of String-Induced CMB Anisotropies

TL;DR

This paper introduces a multi-scale edge-detection pipeline combining curvelet and Canny algorithms to detect cosmic string imprints on CMB maps, achieving high sensitivity in simulations with varying noise levels.

Contribution

The authors develop a novel multi-scale detection method that enhances cosmic string signatures in CMB data, improving detection thresholds compared to previous techniques.

Findings

Detects cosmic strings with Gμ as low as 4.3×10⁻¹⁰ in noiseless maps.

Detection threshold increases to Gμ ≈ 1.2×10⁻⁷ with realistic noise levels.

Uses statistical tools to quantify deviations caused by cosmic strings in simulated CMB maps.

Abstract

We propose a multi-scale edge-detection algorithm to search for the Gott-Kaiser-Stebbins imprints of a cosmic string (CS) network on the Cosmic Microwave Background (CMB) anisotropies. Curvelet decomposition and extended Canny algorithm are used to enhance the string detectability. Various statistical tools are then applied to quantify the deviation of CMB maps having a cosmic string contribution with respect to pure Gaussian anisotropies of inflationary origin. These statistical measures include the one-point probability density function, the weighted two-point correlation function (TPCF) of the anisotropies, the unweighted TPCF of the peaks and of the up-crossing map, as well as their cross-correlation. We use this algorithm on a hundred of simulated Nambu-Goto CMB flat sky maps, covering approximately $10\%$ of the sky, and for different string tensions $G\mu$. On noiseless sky maps with an angular resolution of $0.9'$, we show that our pipeline detects CSs with $G\mu$ as low as $G\mu\gtrsim 4.3\times 10^{-10}$. At the same resolution, but with a noise level typical to a CMB-S4 phase II experiment, the detection threshold would be to $G\mu\gtrsim 1.2 \times 10^{-7}$.