Search for host galaxies of unlocalized Fast Radio Bursts in the SDSS IV catalog

TL;DR

This paper develops and tests models to predict the host galaxies of unlocalized Fast Radio Bursts using their dispersion measures and galaxy survey data, enabling probabilistic localization of these transients.

Contribution

It introduces four physically motivated models relating DM and redshift, and demonstrates a statistical method to identify likely host galaxies in large surveys.

Findings

Established the probability of ~1000 FRBs being hosted by SDSS galaxies.

Validated models using simulated FRB data and statistical metrics.

Predicted FRBs with ≥0.95 probability to be in SDSS galaxies based on sky position and properties.

Abstract

This theoretical work investigates different models to predict the redshift of Fast Radio Bursts (FRBs) from their observed dispersion measure (DM) and other reported properties. We performed an extensive revision of the FRBs with confirmed galaxy hosts in the literature and compiled an updated catalog. With this sample of FRBs, composed of 117 unique transients, we explore four physically motivated models that relate the DM and redshift ($z$): a linear trend (inspired by the Macquart relation), a log-parabolic function, a power-law, and a combined model from the above. We assess the success of these theoretical proposals by implementing different statistical metrics and ranking them. The DM-$z$ relations are also tested using 100 realizations of 500 simulated FRBs, which follow the observed DM trends. Relying on our theoretical modeling, we establish the probability of $\sim$1000 FRBs with unknown $z$ (from the latest CHIME data release) to be hosted by galaxies in the SDSS archival dataset. Our validation scheme allows us to predict the FRBs with a probability threshold of $\geq$0.95 to originate in these galaxies, using their 2D angular position in the sky, magnitude in the r-band, and redshift. This statistical proposal will be tested with upcoming data releases from DESI and new generations of galaxy surveys, such as Euclid, and it opens brilliant possibilities to localize these transients in an automatic pipeline.