A Linear and Quadratic Time-Frequency Analysis of Gravitational Waves from Core-Collapse Supernovae

TL;DR

This paper introduces advanced linear and quadratic time-frequency analysis techniques to better identify and interpret gravitational wave signals from core-collapse supernovae, revealing new multimodal features and polarization behaviors.

Contribution

It develops and applies sophisticated TFA methods, including polarization analysis, to improve detection and understanding of GW signals from CCSN simulations, highlighting new spectral modes and polarization states.

Findings

Clear separation of multimodal GW signatures

Identification of a new oscillation mode at 600-700 Hz

Correlation of polarization states with PNS oscillations

Abstract

Recent core-collapse supernova (CCSN) simulations have predicted several distinct features in gravitational-wave (GW) spectrograms, including a ramp-up signature due to the g-mode oscillation of the proto-neutron star (PNS) and an excess in the low-frequency domain (100-300 Hz) potentially induced by the standing accretion shock instability (SASI). These predictions motivated us to perform a sophisticated time-frequency analysis (TFA) of the GW signals, aimed at preparation for future observations. By reanalyzing a gravitational waveform obtained in a three-dimensional general-relativistic CCSN simulation, we show that both the spectrogram with an adequate window and the quadratic TFA separate the multimodal GW signatures much more clearly compared with the previous analysis. We find that the observed low-frequency excess during the SASI active phase is divided into two components, a stronger one at 130 Hz and an overtone at 260 Hz, both of which evolve quasi-statically during the simulation time. We also identify a new mode whose frequency varies from 700 to 600 Hz. Furthermore, we develop the quadratic TFA for the Stokes I, Q, U, and V parameters as a new tool to investigate the GW circular polarization. We demonstrate that the polarization states that randomly change with time after bounce are associated with the PNS g-mode oscillation, whereas a slowly changing polarization state in the low-frequency domain is connected to the PNS core oscillation. This study demonstrates the capability of the sophisticated TFA for diagnosing the polarized CCSN GWs in order to explore their complex nature.