Deriving Photometric Redshifts using Fuzzy Archetypes and Self-Organizing Maps. I. Methodology

TL;DR

This paper introduces a novel method combining fuzzy archetypes and Self-Organizing Maps to improve photometric redshift estimation by capturing galaxy spectral variations and resolving degeneracies, demonstrated on LSST and Euclid mock data.

Contribution

It presents a new approach that enhances template fitting for photometric redshifts using fuzzy archetypes and SOMs, enabling better exploration of galaxy SED variations and degeneracies.

Findings

Accurately recovers full redshift PDFs with MCMC sampling.

Effectively clusters model observables in a topologically smooth manner.

Demonstrates applicability on LSST and Euclid mock photometry.

Abstract

We propose a method to substantially increase the flexibility and power of template fitting-based photometric redshifts by transforming a large numbers of galaxy spectral templates into a corrresponding collection of "fuzzy archetypes" using a suitable set of perturbative priors designed to account for empirical variation in dust attenuation and emission line strengths. To bypass widely seperated degeneracies in parameter space (e.g., the redshift-reddening degeneracy), we train Self-Organizing Maps (SOMs) on a large "model catalogs" generated from appropriate Monte Carlo sampling of our fuzzy archetypes to cluster the predicted observables in a topologically smooth fashion. Subsequent sampling over the SOM then allows full reconstruction of the relevant probability distribution functions (PDFs) using the associated set of inverse mappings from the SOM to the underlying model parameters. This combined approach enables the multi-modal exploration of known variation among galaxy spectral energy distributions (SEDs) using large numbers of archetypes with minimal modeling assumptions. We demonstrate the power of this approach to recover full redshift PDFs using discrete Markov Chain Monte Carlo (MCMC) sampling methods combined with SOMs constructed from model catalogs based on LSST $ugrizY$ and Euclid $YJH$ mock photometry.