Searches for primordial black holes using gravitational waves

by Marc Andrés Carcasona (IFAE)

Thursday 27 Mar 2025, 15:00 → 18:00 Europe/Madrid

IFAE Seminar Room + Zoom (Hybrid)

Thesis supervisor: Lluïsa-Maria Mir Martínez

Thesis committee: Frédérique Marion (LAPP), Michele Maggiore (University of Geneva) and Matthew Evans (MIT)

This thesis investigates three key aspects of gravitational wave experiments: the mitigation of scattered light noise, the application of machine learning techniques to data analysis, and dedicated searches for primordial black holes. First, it addresses scattered light noise by detailing the use of instrumented baffles for Virgo during O5 and proposing an initial layout for baffles in the main arms of the Einstein Telescope. Additionally, it introduces a detailed discussion of the baffle diffracted aided reflection noise from the beampipe. Second, it introduces a novel approach employing convolutional neural networks for both the detection and parameter estimation of gravitational wave signals, significantly reducing the computing cost. The downside of this approach is a lower sensitivity and a higher parameter uncertainty. Finally, it analyzes the data from the LIGO-Virgo-KAGRA collaboration to constrain the abundance of PBHs. By searching for highly mass-asymmetric binaries with a subsolar component, continuous gravitational waves from the inspiral of very light compact objects, and through population studies of binary black holes, the thesis establishes new limits on the PBH abundance and their potential contribution to the dark matter content of the Universe.