A Cautionary Tale of LyC Escape Fraction Estimates from High Redshift Galaxies

TL;DR

This paper introduces a new methodology for estimating the LyC escape fraction from high-redshift galaxies, accounting for uncertainties and IGM transmission variability, and demonstrates its limitations and biases in current observational approaches.

Contribution

It presents a novel Bayesian method for deriving the posterior distribution of LyC escape fractions considering IGM transmission uncertainties, and critically evaluates existing estimation techniques and galaxy selection biases.

Findings

The new method provides a more accurate posterior distribution of $f_{esc}$.

Average transmission assumptions can significantly overestimate $f_{esc}$.

Photometric redshift overestimations bias LyC candidate selection.

Abstract

Measuring the escape fraction, $f_{\rm esc}$, of ionizing, Lyman Continuum (LyC) radiation is key to our understanding of the process of cosmic reionization. In this paper we provide a methodology for recovering the posterior probability distribution of the LyC escape fraction, $f_{\rm esc}^{\rm PDF}$, considering both the observational uncertainties and ensembles of simulated transmission functions through the intergalactic medium (IGM). We present an example of this method applied to a VUDS galaxy at $z=3.64$ and find $f_{\rm esc}^{\rm PDF}$ = 0.51$^{+0.33}_{-0.34}$ and compare this to the values computed assuming averaged IGM transmission with and without consideration of detection bias along average sightlines yielding $f_{\rm esc}^{\langle T \rangle}$ = 1.40$^{+0.80}_{-0.42}$, and $f_{\rm esc}^{\rm bias}$ = 0.82$^{+0.33}_{-0.16}$. Our results highlight the limitations of methods assuming average, smooth transmission functions. We also present MOSFIRE data for a sample of seven LyC candidates selected based on photometric redshifts at $z > 3.4$, but find that all seven have overestimated photometric redshifts by $\Delta z \sim 0.2$ making them unsuitable for LyC measurements. This results likely due to a bias induced by our selection criteria.