Green Bank Telescope Constraints on Dark Matter Annihilation in Segue I

We use a non-detection in $\nu = 1.4\,$GHz Green Bank Telescope observations of the ultra-faint dwarf spheroidal galaxy Segue I, which could be immersed in a non-negligible halo magnetic field of the Milky Way, to place bounds on particle dark matter properties. We model the galaxy using an Einasto dark matter profile, and compute the expected synchrotron flux from dark matter annihilation as a function of the magnetic field strength $B$, diffusion coefficient $D_0$, and particle mass $m_\chi$ for different annihilation channels. The data strongly disfavor annihilations to $e^+e^-$ for $m_\chi \lesssim 50\,$GeV, but are not sensitive to the $b \bar b$ channel. Adopting a fiducial $B \sim 2\,\mu$G inferred from Segue I's proximity to the Milky Way, our models of annihilation to $\tau^+\tau^-$ with $m_\chi = 30\,$GeV require an intermediate value of $D_0$ for consistency with the data. The most compelling limits are obtained for WIMP annihilation to $\mu^+\mu^-$: we exclude $m_\chi \lesssim 30\,$GeV$\,\rightarrow\mu^+\mu^-$ at 95% confidence, unless $D_0$ exceeds the Milky Way value or $B$ is significantly smaller than we have assumed.