Detectability of accretion-induced bosenovae in the Milky Way

Abstract: We estimate collapse rates of axion stars in our galaxy based on the axion minicluster mass function of the Milky Way dark matter halo. We consider axion-like particles with different temperature evolution of the axion mass, including the QCD axion with $m_a=50\,\mu$eV. Combining estimates for the present-day axion star mass function from our previous work with the axion star accretion model predicted by self-similar growth, we can infer the expected number of bosenovae occurring within the Milky Way. Our estimates suggest that for an observation time of $t_\mathrm{obs}=1\,$yr, the majority of the up to $\sim 10<sup>{13}$ bosenovae per galaxy occur in the densest miniclusters with initial overdensity parameter $\Phi\lesssim 10<sup>4$. We discuss the detectability of such recurring axion bursts within our galactic vicinity and find that, for models with derivative couplings including axion-fermion interactions, potential broadband axion DM experiments can probe a large range of ALP masses $m_a\lesssim 10<sup>{-6}\,$eV and with moderate improvements even the QCD axion case. For axions with non-derivative-type interactions like the axion-photon coupling, our analysis suggests that optimistic predictions with order-one dark matter abundance of axion stars $f_\star \sim 1$ can be probed by dedicated burst searches.