The diffuse supernova neutrino background: an update with modern population synthesis and core-collapse simulations

TL;DR

This work addresses predicting the diffuse supernova neutrino background (DSNB) by combining state-of-the-art CCSN neutrino spectra from multi-dimensional simulations with binary population synthesis to model a binary-affected progenitor population. A key step is mapping the CO-core mass $M_{ m CO}$ to the pre-collapse compactness $\xi_{2.5}$, then deriving neutrino flux parameters for each flavor as functions of $\xi_{2.5}$; black-hole-forming collapses are explicitly included, extending beyond previous 1D treatments. The main results show that black-hole formation enhances the DSNB high-energy tail by up to about 50% for $E \,>\, 30$–$40$ MeV, while binary evolution affects the overall flux moderately (up to ~15% depending on the model). The framework is designed for easy updates as new simulations and population-synthesis models become available, making the DSNB a more powerful probe of both core-collapse physics and the evolution of binary stars; however, uncertainties in the overall core-collapse rate and neutrino flavor conversions remain limiting factors.

Abstract

We present a new, state-of-the-art computation of the Diffuse Supernova Neutrino Background (DSNB), where we use neutrino spectra from multi-dimensional, multi-second core collapse supernova simulations - including both neutron-star and black-hole forming collapses - and binary evolution effects from modern population synthesis codes. Large sets of numerical results are processed and connected in a consistent manner, using two key quantities: the mass of the star's Carbon-Oxygen (CO) core at an advanced pre-collapse stage - which depends on binary evolution effects - and the compactness parameter, which is the main descriptor of the post-collapse neutrino emission. The method enables us to model the neutrino emission of a very diverse, binary-affected population of stars, which cannot unambiguously be mapped in detail by existing core collapse simulations. We find that including black hole-forming collapses enhances the DSNB by up to 50% at energies greater than 30-40 MeV. Binary evolution effects can change the total rate of collapses and generate a sub-population of high core mass stars that are stronger neutrino emitters. However, the net effect on the DSNB is moderate - up to a 15% increase in flux - due to the rarity of these super-massive cores and to the relatively modest dependence of the neutrino emission on the CO core mass. The methodology presented here is suitable for extensions and generalizations, and therefore it lays the foundation for modern treatments of the DSNB.

Figures (7)

• Figure 1: Schematic overview of DSNB flux calculation procedure. The input parameters for stellar evolution simulations are specified at zero-age main sequence (ZAMS) and include the mass $M$, the mass ratio $q$, the orbital period $P$ or separation $a$, the eccentricity $e$ and the metallicity, $Z$. The simulations typically terminate at C or Ne ignition, and provide the CO-core mass, $\mathrel{{M_{\rm CO} }}$. Using stellar profiles given by other stellar evolution models (with known $\mathrel{{M_{\rm CO} }}$), core-collapse simulations are performed and provide neutrino spectrum, which is well described by the compactness, $\xi_{M_{\rm e}}$. In our work, we follow Horiuchi:2020jnc and map CO-core mass and compactness $\xi_{M_{\rm e}}$ to connect the two sets of stellar models, and ultimately obtain the neutrino emission of an evolved population of stars. See text for details.
• Figure 2: An illustration of the main elements used to compute the population-averaged flux for redshift $z=0$ (and therefore the DSNB, after accounting the evolution with $z$, see Sec. \ref{['sec:DSNBresults']}). (a) The stellar population is described by its distribution in bins of $\mathrel{{M_{\rm CO} }}$. Shown here is the distribution obtained using BPASS including binary evolution effects (Sec. \ref{['sec:BinaryEffects']}). For each star of given $\mathrel{{M_{\rm CO} }}$, the neutrino emission is modeled considering that: (b) the compactness parameter ($\mathrel{{\xi_{2.5} }}$) is a function of $\mathrel{{M_{\rm CO} }}$ (here numerical results, as well as an interpolating curve, are shown), and (c) the parameters describing the neutrino luminosity and spectrum (for each flavor; here results for $\mathrel{{\bar{\nu}}_e}$ are shown) can be described by simple functions of $\mathrel{{\xi_{2.5} }}$. The shaded area indicates the region ($2.0<\mathrel{{M_{\rm CO} }}/M_\odot <3.1$) where black hole formation was found in the simulations of Vartanyan:2023zlb. We note that, in the range $\mathrel{{\xi_{2.5} }}>0.37$ the curves in (c) are an extrapolation; see Fig. \ref{['fig:nuparam']} for details.
• Figure 3: Neutrino parameters as functions of $\mathrel{{\xi_{2.5} }}$, from the results of Vartanyan:2023zlb, for runs that extend to $t\geq 3.5$ s post-bounce. For each flavor we show the total energy emitted (over the entire burst, obtained by extrapolation), the average energy, and the rms energy. The filled markers in color are for neutron-star-forming collapses; for these, functional fits are shown (curves of matching color). The empty markers along the vertical dashed line refer to black-hole-forming collapses.
• Figure 4: Overview of BPASS results, for single and binary stars (dashed and solid lines respectively). In (a) and (c), $N$ indicates the number of stars in a bin of either ZAMS mass ($M$) or CO core mass ($\mathrel{{M_{\rm CO} }}$). Panel (b) shows a density plot, where each pixel represents a bin in the $\mathrel{{M_{\rm CO} }}-M$ parameter space, for binary stars. The colors represent the number of stars in each bin, see legend (numbers are not integer due to some integration or averaging procedures involved in BPASS PattonPrivComm). The gray color indicates $N=0$ (empty bin). The cyan dashed line shows the dependence of $\mathrel{{M_{\rm CO} }}$ on $M$ for single stars. The total number of stars is $N_{\ast}=24038.5$ and $N_{\ast}=22745.6$ for single and binary stars respectively. Note that we restrict to the interval $4.1 \leq M/\mathrel{{M_{\odot} }} \leq 65$ to ensure consistency between the single and binary sets.
• Figure 5: The same as Fig. \ref{['fig:BPASSoverview']}, for the fiducial Kinugawa:2014zha model. The total number of stars in this data set is $N_{\ast}=41066$ and $N_{\ast}=47529$ for single and binary stars respectively.