2015MNRAS.451.4277D

We present constraints on the origins of fast radio bursts (FRBs) using large cosmological simulations. We calculate contributions to FRB dispersion measures (DMs) from the Milky Way, from the Local Universe, from cosmological large-scale structure, and from potential FRB host galaxies, and then compare these simulations to the DMs of observed FRBs. We find that the Milky Way contribution has previously been underestimated by a factor of ∼ 2, and that the foreground-subtracted DMs are consistent with a cosmological origin, corresponding to a source population observable to a maximum redshift z ∼ 0.6-0.9. We consider models for the spatial distribution of FRBs in which they are randomly distributed in the Universe, track the star formation rate of their host galaxies, track total stellar mass, or require a central supermassive black hole. Current data do not discriminate between these possibilities, but the predicted DM distributions for different models will differ considerably once we begin detecting FRBs at higher DMs and higher redshifts. We additionally consider the distribution of FRB fluences, and show that the observations are consistent with FRBs being standard candles, each burst producing the same radiated isotropic energy. The data imply a constant isotropic burst energy of ∼ 7x10⁴⁰ erg if FRBs are embedded in host galaxies, or ∼ 9x10⁴⁰ erg if FRBs are randomly distributed. These energies are 10^–100 times larger than had previously been inferred. Within the constraints of the available small sample of data, our analysis favours FRB mechanisms for which the isotropic radiated energy has a narrow distribution in excess of 10⁴⁰ erg.