Tidal signatures in gravitational waves and systematic biases due to waveform inaccuracies

Chandramouli, Rohit Subbarayan

Permalink

https://hdl.handle.net/2142/127311 Copy

Description

Title

Tidal signatures in gravitational waves and systematic biases due to waveform inaccuracies

Author(s)

Chandramouli, Rohit Subbarayan

Issue Date

2024-08-02

Director of Research (if dissertation) or Advisor (if thesis)

Yunes, Nicolás

Doctoral Committee Chair(s)

Witek, Helvi

Committee Member(s)

• Kahn, Yonatan
• Gammie, Charles

Department of Study

Physics

Discipline

Physics

Degree Granting Institution

University of Illinois at Urbana-Champaign

Degree Name

Ph.D.

Degree Level

Dissertation

Keyword(s)

• Gravitational waves
• Waveform Modeling
• Data analysis
• Hierarchical triples
• Tidal dissipation
• Systematic biases

Abstract

There are many avenues of research in the field of gravitational-wave physics and astronomy with rich and challenging problems, which need to be tackled for current and future gravitational-wave observations. In this dissertation, we study three such avenues – (i) probing the astrophysical environment of a binary, (ii) measuring the viscosity of neutron stars, (iii) characterizing systematic biases in tests of General Relativity (GR). We address (i) by considering an intermediate mass black hole binary that is bound to a supermassive black hole within a dense cluster or galactic nuclei. In such hierarchical triple systems, there can be strong tidal effects such as the Kozai-Lidov effect that induces oscillations in eccentricity and inclination. The gravitational waves emitted by the inner binary would carry a direct imprint of the Kozai-Lidov oscillations, which can be detected by space-based detectors such as LISA. We here compute an analytic gravitational waveform that captures the imprint of the Kozai-Lidov oscillations, which appear as corrections to an isolated eccentric binary. Further, we determine the region of parameter space for which the effect is in principle detectable and measurable by LISA. We address (ii) by considering an inspiraling neutron star binary with tidal dissipation caused by viscosity-driven out-of-equilibrium processes. We use a recently developed gravitational waveform model that captures the effect of tidal dissipation (to leading post-Newtonian order) through the dissipative tidal deformability. Using the data from the binary neutron star event GW170817, we perform Bayesian parameter estimation to obtain the first constraint on the dissipative tidal deformability. In turn, we use this constraint to place the first data informed upper bounds on the average bulk (≲ 1031gcm−1s−1) and shear viscosity (≲ 1028gcm−1s−1) of a neutron star. We forecast that these bounds improve by two orders of magnitude with third generation gravitational wave detectors. This opens the door for precision probes of the viscosity-driven out-of-equilibrium dissipative mechanisms inside a neutron star. We address (iii) by studying systematic biases in parameterized tests of GR due to waveform mismodeling of effects within GR. We develop statistical criteria to characterize such false GR deviations and also explicitly connect different statistical measures that are commonly used for assessing systematic biases across gravitational-wave modeling and data analysis. The four regimes of significant systematic bias – Strong Inference of No GR Deviation, Weak Inference of No GR Deviation I & II, Incorrect Inference of GR Deviation – are characterized by the Bayes factor between the ppE model and GR models, and the fitting factor of the ppE model. As a concrete example, we show that when effects of spin-precession are neglected, there are significant biases in the non-GR parameters when the signal is from an edge-on highly precessing source with a signal-to-noise (SNR) of 30. These biases are examples of Strong Inference of No GR Deviation as they are characterized by a significant loss of SNR in using the inaccurate non-GR model, and due to the Bayes factor not strongly disfavoring the GR model. We study the dependence of the biases with the source parameters and SNR and determine for which systems the biases are most significant and likewise for which systems the tests are insensitive to effects of spin-precession. At a higher SNR, we find the biases are examples of Weak Inference of No GR Deviation II as they are characterized by a strong preference for the non-GR model over the GR model, and a significant loss of SNR due to using the non-GR model. Additionally for a toy model, using the linear signal approximation, we show that it is in principle possible for the non-GR parameter to absorb the mismodeled/neglected GR effect, resulting in an Incorrect Inference of a GR Deviation. In such cases, a false GR deviation cannot be distinguished from a true GR deviation, showing the importance of improving the waveform models used to test GR.

Graduation Semester

2024-12

Type of Resource

Thesis

Handle URL

https://hdl.handle.net/2142/127311

Copyright and License Information

Copyright 2024 Rohit Subbarayan Chandramouli

Owning Collections