Automated multitone fitting of binary black hole ringdowns

With the ever-increasing accuracy and number of gravitational-wave detections, we will soon be able to perform improved tests of general relativity. With more precise measurements of the ringdown phase of black hole mergers, we will be able to use quasinormal mode (QNM) excitations to test the no-hair theorem and conduct other consistency checks. These checks, however, require that we know what Einstein’s theory predicts for the magnitude of these excitations. This knowledge can only be obtained by extracting these excitations from numerical relativity (NR) simulations, which can be challenging. To simplify this process, we present a Python package that automates the multitone fitting of the ringdown phase in NR simulations of binary black hole mergers. This code uses the variable projection algorithm to iteratively identify unmodeled QNM contributions. We also perform fits to analytic models to establish physically motivated bounds on the stability of arbitrary QNMs against numerical noise. To validate this code, we tested it against 38 toy models that include different combinations of overtones, higher-order modes and retrograde modes, subjected to Gaussian noise with standard deviation ranging from 10^-12 to 10^-4. This new approach enables a more thorough analysis of the ringdown phase in NR waveforms, improving not only our understanding of the reach of first-order perturbation theory, but also of the relationship between binary parameters and the excitation of QNMs.