Constraints on the cosmic ray diffusion coefficient in the W28 region from gamma-ray observations

GeV and TeV gamma rays have been detected from the supernova remnant W28 and its surroundings. Such emission correlates quite well with the position of dense and massive molecular clouds and thus it is often interpreted as the result of hadronic cosmic ray interactions in the dense gas. Constraints on the cosmic ray diffusion coefficient in the region can be obtained, under the assumption that the cosmic rays responsible for the gamma ray emission have been accelerated in the past at the supernova remnant shock, and subsequently escaped in the surrounding medium. In this scenario, gamma ray observations can be explained only if the diffusion coefficient in the region surrounding the supernova remnant is significantly suppressed with respect to the average galactic one.

💡 Research Summary

The paper investigates the diffusion of cosmic‑ray (CR) protons in the vicinity of the supernova remnant (SNR) W28 by exploiting the spatial and spectral characteristics of the GeV–TeV γ‑ray emission detected by Fermi‑LAT and H.E.S.S. The γ‑rays are tightly correlated with several massive molecular clouds (MC A, MC B, MC C, etc.) that surround W28, suggesting a hadronic origin: CR protons accelerated at the SNR shock interact with the dense gas, producing neutral pions that decay into γ‑rays. The authors adopt a standard picture in which protons are injected with a power‑law spectrum (∝ E⁻²) at the shock, escape the remnant after a characteristic escape time tₑₛ꜀, and subsequently diffuse through the ambient medium. Diffusion is parameterised as D(E)=χ D_Gal(E)·(E/10 GeV)^δ, where D_Gal(E)=10²⁸ cm² s⁻¹ (E/10 GeV)^δ is the canonical Galactic diffusion coefficient, χ is a suppression factor, and δ (≈0.3–0.6) encodes the energy dependence. By varying χ, δ, and tₑₛ꜀, the authors compute the expected γ‑ray spectra and surface‑brightness profiles for each cloud and compare them with the observations.

The key result is that only models with a strongly suppressed diffusion coefficient (χ≈0.01–0.1, i.e., D≈10²⁶–10²⁷ cm² s⁻¹ at 10 GeV) can simultaneously reproduce the observed flux levels, spectral shapes, and the spatial coincidence of the γ‑ray emission with the clouds. This suppression is one to two orders of magnitude below the average Galactic value. The best‑fit parameters also imply an escape time of order 10⁴ yr, consistent with the estimated age of W28 (≈3–4 × 10⁴ yr). The authors discuss plausible physical mechanisms for such suppression: enhanced magnetic turbulence generated by the SNR shock, damping of Alfvén waves in the high‑density environment, and compression of the interstellar magnetic field by the expanding remnant. All these effects reduce the mean free path of CRs, leading to slower diffusion.

The paper further examines the uniformity of the suppression across different clouds. Despite variations in cloud mass, distance from the SNR, and density, a single χ value provides a satisfactory fit for all, indicating that the suppressed diffusion region extends over several tens of parsecs around W28. This finding challenges the common assumption that Galactic‑average diffusion applies universally and underscores the importance of local environmental conditions in shaping CR propagation.

In the discussion, the authors highlight the broader implications for Galactic CR transport models. If diffusion can be locally reduced by up to two orders of magnitude near SNRs interacting with dense gas, the contribution of such sources to the overall CR spectrum observed at Earth may need to be reassessed. Moreover, the suppressed diffusion zone could act as a reservoir, retaining freshly accelerated particles for longer periods and enhancing the γ‑ray luminosity of nearby clouds.

Finally, the paper outlines future prospects. Upcoming facilities such as the Cherenkov Telescope Array (CTA) will deliver higher sensitivity and finer angular resolution, enabling detailed mapping of the diffusion coefficient’s spatial variation and its temporal evolution. Complementary observations in radio (to probe magnetic turbulence) and in molecular lines (to refine cloud density estimates) will further constrain the physical conditions responsible for diffusion suppression.

In summary, by coupling γ‑ray observations with a physically motivated diffusion model, the authors demonstrate that the CR diffusion coefficient in the W28 region must be significantly lower than the Galactic average. This result provides a robust, observationally anchored constraint on CR propagation in complex SNR‑molecular cloud environments and opens new avenues for refining Galactic CR transport theories.