Multimessenger asteroseismology of core-collapse supernovae

Abstract

We investigate correlated gravitational wave and neutrino signals from rotating core-collapse supernovae with simulations. Using an improved mode identification procedure based on mode function matching, we show that a linear quadrupolar mode of the core produces a dual imprint on gravitational waves and neutrinos in the early post-bounce phase of the supernova. The angular harmonics of the neutrino emission are consistent with the mode energy around the neutrinospheres, which points to a mechanism for the imprint on neutrinos. Thus, neutrinos carry information about the mode amplitude in the outer region of the core, whereas gravitational waves probe deeper in. We also find that the best-fit mode function has a frequency bounded above by ∼ 280 Hz, and yet the mode's frequency in our simulations is almost double, due to the use of Newtonian hydrodynamics and a widely used pseudo-Newtonian gravity approximation. This overestimation is particularly important for the analysis of gravitational wave detectability and asteroseismology, pointing to serious limitations of pseudo-Newtonian approaches for these purposes, possibly even resulting in excitation of incorrect modes. In addition, mode frequency matching (as opposed to mode function matching) could be resulting in mode misidentification in recent work. Lastly, we evaluate the prospects of a multimessenger detection of the mode using current technology. The detection of the imprint on neutrinos is most challenging, with a maximum detection distance of ∼ 1 kpc using the IceCube Neutrino Observatory. The maximum distance for detecting the complementary gravitational wave imprint is ∼ 5 kpc using Advanced LIGO at design sensitivity.

Consistent perturbative modeling of pseudo-Newtonian core-collapse supernova simulations

“…The ensuing frequency mismatches generated by the use of different equations may result in mode misidentification during a mode frequency matching procedure, particularly due the absence of the lapse function in the hydrodynamic fluxes in the simulations. * jwestern@email.arizona.edu In [8,12] a mode function matching procedure was followed instead. This entails comparing the mode functions computed perturbatively with the velocity data in the simulation.…”

“…The mode function matching procedure produced convincing mode identification despite the use of perturbative equations that are not consistent with the simulation, because neighboring mode functions have distinct enough morphology that the best-fitting mode function is clearly superior to the next-best-fitting one (provided the mode's excitation is large enough with respect to stochastic or nonlinear motions). A frequency mismatch between the best-fitting mode functions and the simulation frequencies of order ∼ 15% was observed in [8,12], which is large enough to have caused a mode misidentification via mode frequency matching. During targeted modeling of the next galactic core-collapse supernova, this would have produced incorrect inferences about the source.…”

“…It is more expensive because the entire fluid data in the system must be saved with sufficient temporal cadence such that the spectral resolution allows a clean Fourier extraction of individual mode activity. In [8,12] axisymmetric simulations were performed, which alleviates the storage issue, but one wishes to identify modes in fully 3D simulations as well. Large searches of the CCSN progenitor parameter space would be hampered by the need to perform mode function matching.…”

“…Previous studies indicate that mode frequencies are systematically shifted with respect to general relativity (see e.g. [22]), and overestimated in particular [8,12,22], but one cannot know for sure without directly identifying the excited modes in each case (e.g. by mode function matching).…”

“…by mode function matching). The pitfalls found in [8,12] in using pseudo-Newtonian simulations to study mode frequencies were anticipated clearly in [22].…”