Constraining Lyman continuum escape using Machine Learning

The James Webb Space Telescope (JWST) will observe the rest-frame ultraviolet/optical spectra of galaxies from the epoch of reionization (EoR) in unprecedented detail. While escaping into the intergalactic medium, hydrogen-ionizing (Lyman continuum; LyC) photons from the galaxies will contribute to the bluer end of the UV slope and make nebular emission lines less prominent. We present a method to constrain leakage of the LyC photons using the spectra of high redshift (z ≳ 6) galaxies. We simulate JWST/NIRSpec observations of galaxies at z =6–9 by matching the fluxes of galaxies observed in the Frontier Fields observations of galaxy cluster MACS-J0416. Our method predicts the escape fraction fesc with a mean absolute error Δfesc ≈ 0.14. The method also predicts the redshifts of the galaxies with an error $\left\langle \frac{\Delta z}{(1+z)}\right\rangle \approx 0.0003$.