Ray-tracing Fast Radio Bursts Through IllustrisTNG: Cosmological Dispersion Measures from Redshift 0 to 5.5

TL;DR

This paper develops a new method to accurately measure the dispersion measures of Fast Radio Bursts in cosmological simulations, improving modeling of cosmic baryons and matching recent observational data.

Contribution

It introduces a continuous ray-tracing technique through IllustrisTNG, producing extensive DM catalogs and correcting previous sampling issues in DM distribution estimates.

Findings

The new method reduces estimation errors of DM distribution moments by over 50%.

The functional form for p(DM|z) fits observational data well across all redshifts.

Simulation box size and resolution significantly impact the accuracy of DM signals.

Abstract

The dispersion measures (DMs) of Fast Radio Bursts (FRBs) arise predominantly from free electrons in the large-scale structure of the Universe. The increasing number of FRB observations have started to empirically constrain the distribution of cosmic baryons, making it crucial to accurately forward model their propagation within cosmological simulations. In this work, we present a method for measuring FRB DMs in IllustrisTNG that continuously traces rays through the simulation while reconstructing all traversed line segments within the underlying Voronoi mesh. Leveraging this technique, we create over $20$ publicly available DM catalogs, including a full-sky DM map observed from a Milky Way-like environment. Our method addresses a problem in previous TNG-based studies, in which a sparse snapshot sampling in the path integral leads to a misestimation of the standard deviation and higher moments of the DM distribution $p(\rm{DM}|z)$ by over $50\%$. We show that our results are consistent with the most recent observational data from the DSA-110, ASKAP, and CHIME. We offer a functional form for $p(\rm{DM}| z)$ that provides very good fits across all redshifts from $0$ to $5.5$, showing that the previously proposed log-normal distribution is not well matched to the data. We find that using simulation box sizes smaller than $35\,h^{-1}\,$Mpc or resolutions with baryonic masses of less than $5\times10^{8}\,h^{-1}\,M_{\odot}$ can distort the derived DM signal by more than $8\%$. Our findings provide new insights into how the cosmic web shapes FRB signals, and highlight the importance of accurate methodology when comparing cosmological simulations against observations.