High Energy Astrophysics 5 JAN, 2015

The Star-Forming Galaxy Contribution to the Cosmic MeV and GeV Gamma-Ray Background

The Star-Forming Galaxy Contribution to the Cosmic MeV and GeV Gamma-Ray Background

While star-forming galaxies could be major contributors to the cosmic GeV $\gamma$-ray background, they are expected to be MeV-dim because of the “pion bump” falling off below ~100 MeV. However, there are very few observations of galaxies in the MeV range, and other emission processes could be present. We investigate the MeV background from star-forming galaxies by running one-zone models of cosmic ray populations, including Inverse Compton and bremsstrahlung, as well as nuclear lines (including $^{26}$Al), emission from core-collapse supernovae, and positron annihilation emission, in addition to the pionic emission. We use the Milky Way and M82 as templates of normal and starburst galaxies, and compare our models to radio and GeV–TeV $\gamma$-ray data. We find that (1) higher gas densities in high-z normal galaxies lead to a strong pion bump, (2) starbursts may have significant MeV emission if their magnetic field strengths are low, and (3) cascades can contribute to the MeV emission of starbursts if they emit mainly hadronic $\gamma$-rays. Our fiducial model predicts that most of the unresolved GeV background is from star-forming galaxies, but this prediction is uncertain by an order of magnitude. About ~2% of the claimed 1 MeV background is diffuse emission from star-forming galaxies; we place a firm upper limit of <~10% based on the spectral shape of the background. The star-formation contribution is constrained to be small, because its spectrum is peaked, while the observed background is steeply falling with energy through the MeV-GeV range.

💡 Research Summary

The paper investigates how much star‑forming galaxies contribute to the diffuse cosmic gamma‑ray background in the MeV and GeV energy ranges. While it is well established that normal and starburst galaxies can dominate the unresolved GeV background through hadronic (pionic) emission, the same galaxies are expected to be faint in the MeV band because the characteristic “pion bump” falls off sharply below ~100 MeV. However, the MeV sky is poorly explored, and additional processes—inverse‑Compton (IC) scattering, bremsstrahlung, nuclear line emission (e.g., ⁶²Al), core‑collapse supernova (CCSN) gamma‑rays, and positron‑annihilation (511 keV line)—could provide a non‑negligible MeV flux.

Methodology
The authors construct a one‑zone cosmic‑ray (CR) model that treats a galaxy as a homogeneous volume filled with gas, radiation fields, and magnetic fields. Primary CR protons and electrons are injected with power‑law spectra; their energy losses (hadronic collisions, ionization, synchrotron, IC, bremsstrahlung) are computed self‑consistently. The model includes:

  1. Pionic gamma‑rays from π⁰ decay.
  2. IC scattering of CR electrons on the interstellar radiation field (IR, optical, CMB).
  3. Bremsstrahlung from CR electrons interacting with gas.
  4. Nuclear de‑excitation lines, especially the 1.809 MeV line from ⁶²Al decay.
  5. Gamma‑rays from CCSN shock acceleration and associated neutrino production.
  6. Positron annihilation radiation (511 keV line and continuum).

Two well‑studied galaxies serve as templates: the Milky Way (normal star‑forming galaxy) and M82 (starburst). Their radio spectra, Fermi‑LAT GeV data, and TeV measurements from VERITAS/H.E.S.S. are used to calibrate model parameters such as gas density (n≈1 cm⁻³ for the Milky Way, n≈250 cm⁻³ for M82), magnetic field strength (B≈5 µG vs. B≈200 µG), and CR injection efficiency. For high‑redshift (z≈1–2) normal galaxies, the authors assume an average gas density 2–3 times higher than the present Milky Way, reflecting the observed increase in cosmic star‑formation rate density.

A cascade component is also modeled: high‑energy (> GeV) gamma‑rays can be absorbed via γγ → e⁺e⁻ pair production on the intense infrared radiation fields of starbursts. The resulting secondary electrons and positrons then radiate via IC and bremsstrahlung, feeding power back into the MeV band.

Key Results

  1. High‑z Normal Galaxies – The enhanced gas density strengthens hadronic collisions, producing a pronounced pion bump that sharply suppresses emission below ~100 MeV. Consequently, these galaxies contribute negligibly to the MeV background.

  2. Starburst Galaxies – If the magnetic field is relatively weak (B ≲ 50 µG), synchrotron losses are reduced and CR electrons lose energy predominantly through IC and bremsstrahlung, boosting MeV emission by up to a factor of two compared with a strong‑field case.

  3. Cascade Contribution – In starbursts where the primary gamma‑ray output is hadronic, internal γγ absorption can be significant. The cascade re‑processes a fraction (≈10–20 %) of the GeV–TeV power into the MeV band, providing an additional, albeit modest, MeV component.

  4. Overall Background Contribution – The fiducial model predicts that star‑forming galaxies can account for the majority (≈70–90 %) of the unresolved GeV background, though this estimate carries an order‑of‑magnitude uncertainty due to unknown CR injection efficiencies and magnetic field configurations. In the MeV band, the same population contributes only ~2 % of the measured background. By comparing the spectral shape of the observed MeV–GeV background (which declines roughly as E⁻²·⁵) with the peaked, bump‑like spectrum of star‑forming galaxies, the authors place a firm upper limit of ≲10 % on the star‑forming galaxy contribution.

  5. Uncertainties – Varying the CR injection efficiency from 10 % to 30 % changes the GeV background prediction by ~0.5 dex. Adjustments to gas density, magnetic field strength, and CCSN rates similarly affect the MeV output, but the overall conclusion that star‑forming galaxies are a sub‑dominant MeV source remains robust.

Discussion and Implications

The steeply falling MeV background observed by COMPTEL and INTEGRAL cannot be reproduced by the relatively narrow, peaked emission from star‑forming galaxies. Therefore, other astrophysical contributors—such as low‑luminosity active galactic nuclei, blazars, or exotic processes (e.g., dark‑matter annihilation/decay)—must dominate the MeV sky. The paper highlights the importance of future MeV missions (e.g., AMEGO, e‑ASTROGAM) which will have the sensitivity to detect individual normal and starburst galaxies in the MeV band, directly testing the model predictions. Improved measurements of high‑z gas fractions and magnetic fields will also refine the high‑z normal galaxy contribution.

Conclusions

  • Star‑forming galaxies are likely the principal source of the unresolved GeV gamma‑ray background, but the prediction is uncertain by roughly an order of magnitude.
  • In the MeV regime, their contribution is modest (~2 %) and constrained to be < 10 % by spectral shape arguments.
  • The MeV background’s steep decline indicates that other source classes dominate, and that the “pion bump” characteristic of star‑forming galaxies cannot explain the observed spectrum.
  • Future MeV observatories and more sophisticated multi‑wavelength modeling are essential to disentangle the various contributions and to confirm the limited role of star‑forming galaxies in the MeV background.