Simplified dark matter models confront the gamma ray excess

Inspired by the excess of gamma rays from the Galactic Center, we confront a number of simplified dark matter models with experimental data. Assuming a single dark matter particle coupled to standard matter via a spin-0 mediator, we compare model evidences for Majorana fermion, real scalar, and real vector dark matter candidates. We consider dark matter annihilation into various fermionic final states contributing to the observed differential gamma ray flux. Our likelihood function also includes the dark matter relic density, its elastic scattering cross section with nuclei, and collider limits. Using Bayesian inference we confine the mass and couplings strengths of the dark matter and mediator particle. Our results show that, if the gamma ray excess is due to dark matter the above parameters are well constrained by the observations. We find that the Majorana fermion dark matter model is supported the most by the data.

Figure 1

The spectrum of gamma rays produced from dark matter annihilation into individual $b\overline{b},{\tau }^{+}{\tau }^{-}$, and $t\overline{t}$ final states with 20 and 200 GeV dark matter mass, respectively.

Figure 2

Posterior probability distribution as a function of (a) $m_{\chi }$, (b) $m_S$, and (c) ${\lambda }_f$ ($f=b,t,\tau$) for Majorana fermion DM.

Figure 3

Posterior probability distribution as a function of (a) $m_{\varphi }$, (b) $m_S$, and (c) ${\lambda }_f$ ($f=b,t,\tau$) for real scalar DM.

Figure 4

Posterior probability distribution as a function of (a) $m_X$, (b) $m_S$, and (c) ${\lambda }_f$ ($f=b,t,\tau$) for real vector DM.

Figure 5

Posterior probability distribution marginalized to (a) ${\lambda }_{\tau }$ vs ${\lambda }_b$, (b) ${\lambda }_t$ vs ${\lambda }_b$, (c) $m_S$ vs $m_{\chi }$, and (d) ${\sigma }_{\mathrm{SI}}$ vs $m_{\chi }$ for Majorana fermion DM. The light and dark regions correspond to 1 and $2\sigma$ credible regions, respectively.

Figure 6

Posterior probability distribution marginalized to (a) ${\lambda }_{\tau }$ vs ${\lambda }_b$, (b) ${\lambda }_t$ vs ${\lambda }_b$, (c) $m_S$ vs $m_{\varphi }$, and (d) ${\sigma }_{\mathrm{SI}}$ vs $m_{\varphi }$ for real scalar DM.

Figure 7

Posterior probability distribution marginalized to (a) ${\lambda }_{\tau }$ vs ${\lambda }_b$, (b) ${\lambda }_t$ vs ${\lambda }_b$, (c) $m_S$ vs $m_X$, and (d) ${\sigma }_{\mathrm{SI}}$ vs $m_X$ for real vector DM.