Poster abstract details
Identifying reionization-epoch galaxies with extreme levels of Lyman continuum leakage in a JWST survey
Abstract
The NIRSpec instrument on-board the James Webb Space Telescope (JWST) will allow spectroscopic studies of galaxies from the epoch of reionization (EoR). Recent observations of lensed high redshift galaxies suggest that JWST will provide a large sample of galaxy spectra from this era. The fraction of hydrogen-ionizing photons which do not escape into the intergalactic medium will leave an imprint on the spectra in the form of the spectral slope and strength of nebular emission lines.
Here we present a machine learning framework to predict the escape fraction of the hydrogen-ionizing radiation (Lyman continuum) photons using the spectra of high redshift ($z\gtrsim 6$) galaxies. We simulate JWST/NIRSpec observations of galaxies at $z=$6--10 by matching the fluxes of galaxies observed in the Frontier Fields observations of galaxy cluster Abell-2744 and MACS-J0416 for two exposure times with NIRSpec (10 h and 1.5 h). We find that the mean absolute error in predicting escape fractions $\Delta f_{esc}\approx$ 0.12 for the 10 h exposure time and three times worse for a 1.5 h exposure time. The framework also estimates the redshift of those galaxies with a relative error of 0.0128.
Here we present a machine learning framework to predict the escape fraction of the hydrogen-ionizing radiation (Lyman continuum) photons using the spectra of high redshift ($z\gtrsim 6$) galaxies. We simulate JWST/NIRSpec observations of galaxies at $z=$6--10 by matching the fluxes of galaxies observed in the Frontier Fields observations of galaxy cluster Abell-2744 and MACS-J0416 for two exposure times with NIRSpec (10 h and 1.5 h). We find that the mean absolute error in predicting escape fractions $\Delta f_{esc}\approx$ 0.12 for the 10 h exposure time and three times worse for a 1.5 h exposure time. The framework also estimates the redshift of those galaxies with a relative error of 0.0128.