Exploring the physics behind the non-thermal emission from star-forming galaxies detected in γ rays

Abstract

Context. Star-forming galaxies emit non-thermal radiation from radio to γ rays. Observations show that their radio and γ-ray luminosities scale with their star formation rates, supporting the hypothesis that non-thermal radiation is emitted by cosmic rays produced by their stellar populations. However, the nature of the main cosmic-ray transport processes that shape the emission in these galaxies is still poorly understood, especially at low star formation rates. Aims. Our aim is to investigate the main mechanisms of global cosmic-ray transport and cooling in star-forming galaxies. The way they contribute to shaping the relations between non-thermal luminosities and star formation rates could shed light onto their nature, and allow us to quantify their relative importance at different star formation rates. Methods. We developed a model to compute the cosmic-ray populations of star-forming galaxies, taking into account their production, transport, and cooling. The model is parametrised only through global galaxy properties, and describes the non-thermal emission in radio (at 1.4 GHz and 150 MHz) and γ rays (in the 0.1−100 GeV band). We focused on the role of diffusive and advective transport by galactic winds, either driven by turbulent or thermal instabilities. We compared model predictions to observations, for which we compiled a homogeneous set of luminosities in these radio bands, and updated those available in γ rays. Results. Our model reproduces reasonably well the observed relations between the γ-ray or 1.4 GHz radio luminosities and the star formation rate, assuming a single power-law scaling of the magnetic field (with index β = 0.3) and winds blowing either at Alfvenic speeds (∼tens of km s−1, for ≲5 M⊙ yr−1) or typical starburst wind velocities (∼hundreds of km s−1, for ≳5 M⊙ yr−1). Escape of cosmic rays is negligible for ≳30 M⊙ yr−1. A constant ionisation fraction of the interstellar medium fails to reproduce the 150 MHz radio luminosity throughout the whole star formation rate range. Conclusions. Our results reinforce the idea that galaxies with high star formation rates are cosmic-ray calorimeters, and that the main mechanism driving proton escape is diffusion, whereas electron escape also proceeds via wind advection. They also suggest that these winds should be cosmic-ray or thermally driven at low and intermediate star formation rates, respectively. Our results globally support that magneto-hydrodynamic turbulence is responsible for the dependence of the magnetic field strength on the star formation rate and that the ionisation fraction is strongly disfavoured to be constant throughout the whole range of star formation rates.

Athena synergies in the multi-messenger and transient universe

“…It has, therefore, been suggested that starburst galaxies are CR calorimeters, i.e. the total gamma-ray and neutrino output from starburst galaxies may approach the total energy output in cosmic rays [248][249][250][251][252]. In fact, the expected diffuse neutrino emission associated with starburst galaxies is intriguingly close to observations by the IceCube observatory [250].…”

Starburst Nuclei as Light Dark Matter Laboratories

“…This equation shows that scaling the energy-loss timescale by a factor of (Γ − 1) −1 gives a slightly more precise normalization for spectra with Γ = 2 than the simplest approximation that assumes f CR (p) = τ loss • Q CR (p) even though the qualitative results do not change, as already emphasized by Ref. [56]. Finally, we stress that it is important to calculate the CR transport equation into the momentum space, because from a theoretical point of view the CRs are injected with a momentum power-law spectrum [61,94], even though many authors simply assume a power-law spectrum in the energy space.…”

“…For the dimension of the nuclei R SBN , we account for a variation of a factor 2 with respect to the value expected for SBN circumnuclear region [51,54,97,98]. The range for the wind velocity is fixed according to the recent analyses [51,54,56,93,99]. Finally, for the gas density n ISM , the two sources require different priors according to the empirical Kennicut relation [97,98,100].…”

“…We focus on the nuclei of starburst galaxies (hereafter denoted as SBNi), which confine CRs [51][52][53] for ∼ 10 5 years even at energies as large as 100 TeV. While these CRs cannot be directly observed, they produce gamma-rays and neutrinos via hadronic collisions [51][52][53][54][55][56][57]. Therefore, DM-CR interaction can distort the CR spectrum, and in turn the gamma-ray flux observed from SBNi.…”

The cosmic ray ionisation and $γ$-ray budgets of star-forming galaxies

“…Since core collapse SNe rapidly follow star formation, it is therefore natural to expect a linear relationship between star formation rate and CR injection into a galaxy, and thus at least potentially between star formation rate and non-thermal emission that traces CRs. The extent to which such a relationship holds, and to which particular galaxies deviate from it, can then be interpreted as constraining the fraction of CRs that escape ★ E-mail: mark.krumholz@anu.edu.au (MRK) from galaxies; this in turn can be used to illuminate the physics of CR transport through interstellar gas (e.g., Lacki et al 2011;Krumholz et al 2020;Ajello et al 2020;Kornecki et al 2020Kornecki et al , 2022Crocker et al 2021a;Werhahn et al 2021b,c;Ambrosone et al 2022). A crucial input to these interpretive efforts is the total 𝛾-ray production budget associated with star formation -i.e., in a galaxy that is perfectly calorimetric, such that all the CRs accelerated by young stars and their feedback give up their energy within the galaxy, what 𝛾-ray luminosity would we expect per unit mass of stars formed?…”