The Gravitational-Wave Power Gap in Core-Collapse Supernovae: Insights from 60 Axisymmetric Simulations

Abstract

We analyse the gravitational-wave emission from 60 two-dimensional core-collapse supernova simulations. The models cover a range of progenitors and equations of state. We focus on the narrow frequency interval in the gravitational-wave spectrum where the emitted power is strongly suppressed (the power gap) and how its central frequency relates to the physical properties of the simulations. We find that the power-gap frequency exhibits strong and systematic correlations with the properties of the inner core of the forming neutron star, for example the sound speed, suggesting that the gap encodes information about the behaviour of matter at extreme densities. We further examine how well several mechanisms proposed in the literature account for the presence and evolution of the gap in our simulations. Finally, we explore a scenario in which the gap arises from destructive interference between a narrow oscillation mode and a broadband background signal, demonstrating that such an interaction can produce a sharp minimum in the emitted gravitational-wave power.

Figures (12)

• Figure 1: Schematic of GW emission from core-collapse supernovae in the time–frequency domain. Yellow indicates the high-frequency ridge, the grey background represents the haze, and the narrow horizontal feature where emission is suppressed marks the power gap.
• Figure 2: Average shock radius ($R_\mathrm{sh}$) as a function of time after bounce for all 60 models. Each panel shows the 10 EOS variations for a given progenitor, the progenitor name is indicated in the upper left corner. Each EOS is represented by one distinct line colour.
• Figure 3: GW strains ($h_+$) multiplied by the distance to the observer ($D$) (left panels) and corresponding spectrograms (right panels) for a representative subset of our models. Each row shows one model, indicated in the top right of the waveform panel. Time is given in seconds after bounce. The colourbar is logarithmic and each spectrogram has been normalised by the same value, so that they can be directly compared.
• Figure 4: Example of the masking procedure described in section \ref{['sec:gws']}. The figure shows the spectrograms for s15_DD2 with and without masking. The left panel shows the original spectrogram and the right panel shows the spectrogram with the main emission ridge masked out according to Eq. \ref{['eq:hazemask']}. The spectrogram where the haze is masked is the complement of the spectrogram shown in the right panel. Time is given in seconds after bounce and the colour scale is logarithmic.
• Figure 5: Ratio of the GW energy contained in the main ridge and the energy contained in the haze. Grey horizontal lines guide the eye to ratios of 0.2 and 0.5. Each dot corresponds to one simulation, where the model name is given on the x-axis above or below the plot.