Improving Optical Photo-z Constraints for Dusty Star-forming Galaxies Using Submillimeter-based Priors

TL;DR

This paper introduces a method to improve photometric redshift estimates for dusty star-forming galaxies by incorporating far-infrared and millimeter data as priors, significantly reducing errors in large surveys.

Contribution

The authors develop a novel approach that integrates FIR-to-millimeter photometry into standard optical/near-infrared photo-z frameworks, leveraging empirical dust temperature relations.

Findings

Achieved threefold reduction in catastrophic outliers.

Demonstrated effectiveness on Herschel H-ATLAS data.

Enhanced redshift accuracy for FIR-bright galaxies.

Abstract

Photometric redshifts (photo-z's) provide an efficient alternative to spectroscopic redshifts, enabling redshift estimation for large galaxy samples. However, traditional photo-z methods primarily rely on optical and near-infrared (OIR) photometry, which can struggle with dusty star-forming galaxies that are often faint in the OIR but bright at far-infrared (FIR) and millimeter wavelengths. We present a method for incorporating FIR-to-millimeter photometry as a prior within standard OIR-based photo-z frameworks, explicitly folding in the observed empirical relationship between total infrared luminosity and dust temperature. This approach is particularly suitable for wide-area surveys, such as those anticipated with the Euclid satellite or Rubin Observatory, where OIR photo-z's can be complemented with longer-wavelength data to help with the dustiest and highest star-forming galaxies. Applying this method to the Herschel Astrophysical Terahertz Large Area Survey (or H-ATLAS) catalog, which combines FIR photometry from Herschel-SPIRE with OIR observations, we achieve a threefold reduction in catastrophic outliers compared to traditional OIR-based photo-z techniques, demonstrating its utility for improving redshift estimates in FIR-bright galaxies.