Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                

Astrophysics with core-collapse supernova gravitational wave signals in the next generation of gravitational wave detectors

Vincent Roma, Jade Powell, Ik Siong Heng, and Raymond Frey
Phys. Rev. D 99, 063018 – Published 26 March 2019
PDFHTMLExport Citation
Abstract
Authors
See Also
Article Text
  • INTRODUCTION
  • GRAVITATIONAL WAVE SIGNALS FROM…
  • THE SUPERNOVA MODEL EVIDENCE EXTRACTOR
  • GRAVITATIONAL WAVE DETECTORS
  • METHOD
  • RESULTS
  • CONCLUSIONS
  • ACKNOWLEDGMENTS
    • References

    Abstract

    The next generation of gravitational wave detectors will improve the detection prospects for gravitational waves from core-collapse supernovae. The complex astrophysics involved in core-collapse supernovae pose a significant challenge to modeling such phenomena. The Supernova Model Evidence Extractor (SMEE) attempts to capture the main features of gravitational wave signals from core-collapse supernovae by using numerical relativity waveforms to create approximate models. These models can then be used to perform Bayesian model selection to determine if the targeted astrophysical feature is present in the gravitational wave signal. In this paper, we extend SMEE’s model selection capabilities to include features in the gravitational wave signal that are associated with g-modes and the standing accretion shock instability. For the first time, we test SMEE’s performance using simulated data for planned future detectors, such as the Einstein Telescope, Cosmic Explorer, and LIGO Voyager. Further to this, we show how the performance of SMEE is improved by creating models from the spectrograms of supernova waveforms instead of their time-series waveforms that contain stochastic features. In third generation detector configurations, we find that about 50% of neutrino-driven simulations were detectable at 100 kpc, and 10% at 275 kpc. The explosion mechanism was correctly determined for all detected signals.

    • Figure
    • Figure
    • Figure
    • Figure
    • Figure
    • Figure
    • Figure
    1 More
    • Received 25 January 2019

    DOI:https://doi.org/10.1103/PhysRevD.99.063018

    © 2019 American Physical Society

    Physics Subject Headings (PhySH)

    1. Research Areas
    Gravitational wave sourcesNovae & supernovae
    1. Techniques
    Gravitational wave detectionGravitational wave detectorsNumerical relativityNumerical simulations in gravitation & astrophysics
    Gravitation, Cosmology & Astrophysics

    Authors & Affiliations

    Vincent Roma1, Jade Powell2, Ik Siong Heng3, and Raymond Frey1

    • 1University of Oregon, Eugene, Oregon 97403, USA
    • 2OzGrav, Swinburne University of Technology, Hawthorn, Melbourne, Victoria 3122, Australia
    • 3University of Glasgow, Physics and Astronomy, Kelvin Building, Glasgow, Lanarkshire G128QQ, United Kingdom

    See Also

    Inferring core-collapse supernova physics with gravitational waves

    J. Logue, C. D. Ott, I. S. Heng, P. Kalmus, and J. H. C. Scargill
    Phys. Rev. D 86, 044023 (2012)

    Article Text (Subscription Required)

    Click to Expand

    References (Subscription Required)

    Click to Expand
    Issue

    Vol. 99, Iss. 6 — 15 March 2019

    Reuse & Permissions
    Access Options
    CHORUS

    Article Available via CHORUS

    Download Accepted Manuscript
    Author publication services for translation and copyediting assistance advertisement

    Authorization Required

    Log In
    Other Options
    ×

    Images

    • Figure 1
      Figure 1

      Top: Spectrogram of a neutrino model gravitational wave signal for a 20  M⊙ progenitor star simulated by Andresen et al. [24]. The g-mode and SASI waveform features have been circled in white and red, respectively. Core bounce occurs at t=1.0  s. Bottom: ASD of the same neutrino model waveform and magnetorotational waveform (R3E2AC) from Scheidegger et al. [22] plotted with projected future detector noise curves. Both waveforms represent a source distance of 50 kpc and use a sky-averaged antenna pattern of .44.

      Reuse & Permissions
    • Figure 2
      Figure 2

      The first three PCs for each catalog. From top to bottom, the catalogs are: neutrino mechanism, magnetorotational mechanism, g-modes, no g-modes, SASI, and no SASI. The principal components represent the most important waveform features for each catalog.

      Reuse & Permissions
    • Figure 3
      Figure 3

      Log Bayes factors for mechanism classification with an increasing number of PCs in simulated aLIGO Gaussian noise. Waveforms with a * were not included when making the PCs. The left figure shows neutrino catalog waveforms, and the right figure shows magnetorotational catalog waveforms. The log Bayes factors stop increasing when an ideal number of PCs is reached. We chose 5 PCs for the neutrino model and 4 for the magnetorotational as there was limited logBS,N improvement beyond those points.

      Reuse & Permissions
    • Figure 4
      Figure 4

      Spectra for LIGO O1 data recolored to future detector sensitivities. Data is shown from 30–2048 Hz, the frequency band used in SMEE’s analysis.

      Reuse & Permissions
    • Figure 5
      Figure 5

      Top row shows injected waveform and waveform reconstruction produced with the neutrino model PCs at 150 kpc in three ET and two Voyager detectors. Bottom row shows Omega scan spectrograms of the data from the detector in the configuration with the highest SNR. Bottom left plot shows the signal is clearly visible by eye for a 10 kpc injection (SNR=120), while the bottom right plot shows almost nothing visible at 150 kpc (SNR=8). This waveform was confidently classified at 150 kpc as corresponding to the neutrino model with g-mode and SASI emission, even though at that distance those features are clearly not visible by eye in the noisy Omega scan.

      Reuse & Permissions
    • Figure 6
      Figure 6

      Minimum detectable SNR for each classification statement. All injections performed in a simulated A+ configuration that included AdVirgo and Kagra. The top two plots both pertain to mechanism classification, and the bottom two are for g-modes and SASI. All results are organized such that positive Bayes values correspond to correct classifications regardless of whether the feature is present or not.

      Reuse & Permissions
    • Figure 7
      Figure 7

      Mechanism classification efficiency. Top plots show results for catalog waveform injections, bottom plots show results for noncatalog injections. Noncatalog injections are considered to be the most realistic test case for a genuine gravitational wave signal from an arbitrary source.

      Reuse & Permissions
    • Figure 8
      Figure 8

      Classification efficiency for g-mode (left) and SASI (right) waveform features. Performance was better for g-mode classification in our tests, but this is also heavily dependent on the energy of the specific waveform. Overall performance was similar to that of neutrino model mechanism classification.

      Reuse & Permissions
    ×

    Sign up to receive regular email alerts from Physical Review D

    Log In

    Cancel
    ×

    Search

    Article Lookup

    Paste a citation or DOI
    Enter a citation
    ×