Fermi gamma-ray `bubbles from stochastic acceleration of electrons

Gamma-ray data from Fermi-LAT reveal a bi-lobular structure extending up to 50 degrees above and below the galactic centre, which presumably originated in some form of energy release there less than a few million years ago. It has been argued that the gamma-rays arise from hadronic interactions of high energy cosmic rays which are advected out by a strong wind, or from inverse-Compton scattering of relativistic electrons accelerated at plasma shocks present in the bubbles. We explore the alternative possibility that the relativistic electrons are undergoing stochastic 2nd-order Fermi acceleration by plasma wave turbulence through the entire volume of the bubbles. The observed gamma-ray spectral shape is then explained naturally by the resulting hard electron spectrum and inverse Compton losses. Rather than a constant volume emissivity as in other models, we predict a nearly constant surface brightness, and reproduce the observed sharp edges of the bubbles.

💡 Research Summary

The Fermi Large Area Telescope has revealed two giant lobes of gamma‑ray emission extending roughly 50° above and below the Galactic centre, commonly referred to as the “Fermi bubbles”. Their morphology—sharp edges, a nearly uniform surface brightness, and a hard gamma‑ray spectrum that cuts off around a few hundred GeV—implies a relatively recent (≲ few Myr) energetic event in the central region. Two classes of explanations have dominated the literature. The first invokes hadronic interactions: cosmic‑ray protons accelerated near the centre are advected outward by a strong wind, collide with ambient gas, and produce neutral pions that decay into gamma rays. This scenario requires a substantial amount of target material throughout the bubbles and predicts a volume‑filled emissivity that does not naturally yield the observed flat surface brightness or the abrupt boundaries. The second class relies on leptonic processes: shocks within the bubbles accelerate electrons to multi‑TeV energies; those electrons up‑scatter the interstellar radiation field (ISRF) and the cosmic microwave background (CMB) via inverse‑Compton (IC) scattering, generating the observed gamma rays. While this model can reproduce the spectrum, it demands persistent, large‑scale shock fronts whose signatures are not clearly seen in X‑ray or radio data.

In this paper the authors propose an alternative, hybrid mechanism: stochastic (second‑order Fermi) acceleration of electrons by plasma‑wave turbulence that permeates the entire bubble volume. In this picture, magnetohydrodynamic (MHD) fluctuations—presumably Alfvénic or fast‑mode waves—provide a random, resonant scattering environment. Electrons repeatedly interact with these waves, gaining energy on average while diffusing in momentum space. The kinetic description reduces to a Fokker‑Planck equation for the electron distribution f(p,t) with a momentum‑diffusion coefficient D(p) and a systematic acceleration term A(p). Assuming a Kolmogorov‑type turbulence spectrum (energy density ∝ k⁻⁵ᐟ³) and a wave phase speed comparable to the Alfvén speed, the authors derive D(p) ∝ p^{2‑q} with q≈5⁄3, leading to an equilibrium electron spectrum f(p) ∝ p^{‑(3‑q)} ≈ p^{‑1.3}. This is markedly harder than the canonical p⁻2 spectrum expected from diffusive shock acceleration.

Simultaneously, the electrons suffer radiative losses dominated by IC scattering off the ISRF and the CMB. The loss rate scales as p², so at low momenta acceleration dominates, while at high momenta losses dominate. The balance defines a characteristic energy E_c where the acceleration time equals the IC cooling time. Below E_c the electron spectrum retains its hard, acceleration‑limited shape; above E_c it steepens sharply, producing a natural high‑energy cutoff in the gamma‑ray spectrum. By convolving this electron distribution with the full anisotropic ISRF (including optical, infrared, and microwave components) the authors compute the IC gamma‑ray emissivity. The resulting spectrum reproduces the observed flat νF_ν shape from ∼1 GeV to ∼100 GeV and the gentle turnover near a few hundred GeV.

A crucial geometric consequence of the stochastic‑acceleration model is that the volume emissivity is not constant. Because the turbulent energy density and the resulting electron density are assumed to be roughly uniform throughout the bubble, the line‑of‑sight integration yields an almost constant surface brightness across the projected area, exactly as seen in the Fermi‑LAT maps. Moreover, the sharp edges arise naturally: at the bubble boundary the turbulence is expected to decay abruptly (e.g., due to a change in magnetic topology or a drop in plasma density), causing the acceleration term to vanish and the electron population to drop sharply, thereby producing the observed limb‑brightened, well‑defined rims.

The authors discuss several strengths of their proposal. First, it eliminates the need for a sustained, large‑scale wind or for persistent shock fronts, both of which lack clear observational support. Second, the stochastic acceleration can operate continuously over the ∼Myr lifetime of the bubbles, consistent with the relatively young age inferred from the sharp edges. Third, the model naturally explains the combination of a hard spectrum, a high‑energy cutoff, and a uniform surface brightness. However, the paper also acknowledges open questions. The origin of the turbulence—whether it is driven by past jet activity, starburst‑driven outflows, or magnetic reconnection near Sgr A*—is not specified, and the energy budget required to maintain Kolmogorov‑type turbulence over tens of kiloparsecs must be quantified. Additionally, the sensitivity of the results to the assumed turbulence spectrum (e.g., Kraichnan versus Kolmogorov) and to the exact wave mode (Alfvénic versus fast‑mode) warrants further study. Finally, the possibility of a mixed leptonic‑hadronic scenario, where a sub‑dominant proton component contributes to the gamma‑ray flux, is not ruled out.

In summary, the paper presents a compelling case that stochastic second‑order Fermi acceleration by pervasive plasma‑wave turbulence can account for the main observational properties of the Fermi bubbles: a hard, smoothly cutoff gamma‑ray spectrum, a nearly uniform surface brightness, and sharply defined edges. The model offers a parsimonious alternative to wind‑driven hadronic or shock‑driven leptonic explanations, while highlighting the need for future high‑resolution multi‑wavelength observations and detailed MHD simulations to verify the turbulence generation, maintenance, and its coupling to relativistic electrons.