Fermi-LAT observations of the Milky Way Galactic Center (GC) have revealed a spherically symmetric excess of GeV $\gamma$ rays extending to at least 10° from the dynamical center of the Galaxy. A critical uncertainty in extracting the intensity, spectrum, and morphology of this excess concerns the accuracy of astrophysical diffuse $\gamma$-ray emission models near the GC. Recently, it has been noted that many diffuse emission models utilize a cosmic-ray injection rate far below that predicted based on the observed star-formation rate in the Central Molecular Zone. In this study, we add a cosmic-ray injection component which nonlinearly traces the Galactic ${\mathrm{H}}_2$ density determined in three dimensions, and find that the associated $\gamma$-ray emission is degenerate with many properties of the GC $\gamma$-ray excess. Specifically, in models that utilize a large sideband ($40°\times 40°$ surrounding the GC) to normalize the best-fitting diffuse emission models, the intensity of the GC excess decreases by approximately a factor of 2, and the morphology of the excess becomes less peaked and less spherically symmetric. In models which utilize a smaller region of interest ($15°\times 15°$) the addition of an excess template instead suppresses the intensity of the best-fit astrophysical diffuse emission, and the GC excess is rather resilient to changes in the details of the astrophysical diffuse modeling. In both analyses, the addition of a GC excess template still provides a statistically significant improvement to the overall fit to the $\gamma$-ray data. We also implement advective winds at the GC, and find that the Fermi-LAT data strongly prefer outflows of order several hundred km/s, whose role is to efficiently advect low-energy cosmic rays from the inner-few kpc of the Galaxy. Finally, we perform numerous tests of our diffuse emission models, and conclude that they provide a significant improvement in the physical modeling of the multiwavelength nonthermal emission from the GC region.

• Received 21 March 2016

DOI:https://doi.org/10.1103/PhysRevD.94.063504