Speaker
Description
Previous works on constraining the Epoch of Reionization (EoR) using FRBs have been done, but they all assume that the redshift at which the FRB is emitted is known. This is possible in principle but requires localization of the FRBs. This can’t be provided by the radio telescopes that detect most FRBs, so it would require a second measurement. My work makes it computationally feasible to constrain the EoR with FRBs without assuming redshift information. To do this I created an emulator that outputs the Dispersion Measure probability distribution given a redshift and a set of theoretical parameters that model the EoR. The Dispersion Measure is the only observable provided by an FRB detection. It’s a proxy for the number of electrons along its path. I used simulations from 21cmFAST to obtain the probability distributions on a grid, then interpolated between fits of these distributions, to extend this to a continuum of parameters. This allows a hierarchical Bayesian inference given a set of measured FRBs. The next steps for this project are to write a Hamiltonian Monte Carlo algorithm to perform the Bayesian inference for different sets of simulated FRBs, to determine how many are needed, and from what redshift, to accurately constrain reionization. We also intend to explore how some redshift information (a feasible amount of it) could improve these predictions. This pipeline could then be used with real data, and could also provide, for example, a precise measurement of the Macquart relation, with error bars.
