Aristotle University Of Thessaloniki (AUTH)
Robust data analysis methods in gravitational wave astronomy
Abstract
dc:descriptionGravitational Wave astronomy has made remarkable progress since the first detection of GW150914, with over 300 detections, to date. Future ground- and space-based observatories, including third-generation detectors (Einstein Telescope (ET), Cosmic Explorer (CE)) and missions like the Laser Interferometer Space Antenna (LISA), will achieve unprecedented sensitivity. However, these advancements introduce new challenges in data analysis, particularly due to non-Gaussian noise, transient glitches, and overlapping signals. This thesis addresses these challenges by developing a novel statistical framework based on heavy-tailed distributions to improve parameter estimation under noise uncertainties. The effectiveness of this framework is demonstrated through benchmarking on both synthetic and real data, with applications to the LISA Pathfinder (LPF) mission and Bayesian parameter estimation of Supermassive Black Hole Binaries (SMBHB)s, Binary Black Hole (BBH)s, and Ultra Compact Galactic Binaries (UCB)s. Additionally, the framework is applied to the detection of the Stochastic Gravitational-Wave Back- ground (SGWB) from astrophysical and cosmological sources, showcasing its robustness in handling non-Gaussian noise and non-stationary features. A key scientific application explored in this work is the post-merger phase of Binary Neutron Star (BNS) mergers. Using wavelet-based signal reconstruction, specifically the BayesWave pipeline, the detectability of rotational instabilities in BNS remnants is assessed. The results demonstrate the potential of future detectors (CE, ET, and a proposed 25-km interferometer) to probe neutron star physics through post-merger signals. This work introduces and compares two distinct statistical approaches for future Gravitational Wave (GW) data analysis, each tailored to different scientific goals. For post-merger signal analysis, the BayesWave pipeline is evaluated as a tool for signal reconstruction. From a broader data analysis perspective, a robust statistical framework based on heavy-tailed distributions is proposed to enhance signal reconstruction, improve noise estimation, and refine data characterization for a variety of sources.
Degree
thesis:*- Grantor dc:publisher
- Aristotle University Of Thessaloniki (AUTH)
- Year dc:date
- 2025
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Sasli, Argyro
- Σασλή, Αργυρώ
Subjects
dc:subject × 21- Ανάλυση δεδομένων
- Βαρυτικά κύματα
- Στοχαστικά βαρυτικά κύματα
- Ανακατασκευή σήματος
- Εκτίμηση παραμέτρων
- Ανάλυση με κυματίδια
- Data analysis
- Gravitational waves
- Wavelet analysis
- Parameter estimation
- Signal reconstruction
- Φυσική
- Φυσικές Επιστήμες
- Αστρονομία και Αστροφυσική
- Μαθηματικά
- Στατιστική και Πιθανότητες
- Physical Sciences
- Natural Sciences
- Astronomy and Astrophysics
- Mathematics
- Statistics and Probability
Rights
- Language dc:language
- eng
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/59253