The Fermi Bubble as a Source of Cosmic Rays in the Energy Range > 10E15 eV

TL;DR

This paper proposes that the Fermi Bubbles, formed by periodic star capture events near the Galactic center, re-accelerate cosmic rays beyond the knee energy, explaining observed cosmic ray spectra and flux.

Contribution

It introduces a model where the Fermi Bubbles re-accelerate cosmic rays via stochastic processes, accounting for spectral features beyond the knee.

Findings

Re-acceleration in the Fermi Bubbles explains cosmic rays beyond the knee.

The model reproduces observed CR spectral indices and flux.

Conversion efficiency from shock energy to CR energy is about 10%.

Abstract

The {\it Fermi} Large Area Telescope has recently discovered two giant gamma-ray bubbles which extend north and south of the Galactic center with diameters and heights of the order of $H\sim 10$ kpc. We suggest that the periodic star capture processes by the Galactic supermassive black hole Sgr A$^*$, with a capture rate of $\tau_{\rm cap}^{-1}\sim 3\times 10^{-5}$ yr$^{-1}$ and an energy release of $W\sim 3\times 10^{52}$ erg per capture, can result in hot plasma injecting into the Galactic halo at a wind velocity of $u\sim 10^8$ cm s$^{-1}$. The periodic injection of hot plasma can produce a series of shocks. Energetic protons in the bubble are re-accelerated when they interact with these shocks. We show that for energy larger than $E> 10^{15}$ eV, the acceleration process can be better described by the stochastic second-order Fermi acceleration. We propose that hadronic cosmic rays (CRs) within the ``knee'' of the observed CR spectrum are produced by Galactic supernova remnants distributed in the Galactic disk. Re-acceleration of these particles in the Fermi Bubble produces CRs beyond the knee. With a mean CR diffusion coefficient in this energy range in the bubble $D_B\sim 3\times 10^{30}$ cm$^2$ s$^{-1}$, we can reproduce the spectral index of the spectrum beyond the knee and within. The conversion efficiency from shock energy of the bubble into CR energy is about 10\%. This model provides a natural explanation of the observed CR flux, spectral indices, and matching of spectra at the knee.