Skip to main content
SearchLoginLogin or Signup

Estimating Obscured Chandra Source Catalog AGN Redshifts using the XZ Method and Machine Learning

Presentation #404.05 in the session The Chandra Source Catalog version 2.1: New Avenues for Discovery in X-ray Datasets.

Published onJul 01, 2023
Estimating Obscured Chandra Source Catalog AGN Redshifts using the XZ Method and Machine Learning

Computing conventional spectroscopic and photometric redshifts for obscured AGN is a notoriously challenging and costly process. In this context, X-rays are a valuable resource because they can both select obscured AGN and estimate their redshifts using low-count spectra. The XZ Method is a Bayesian X-ray Analysis-based pipeline for performing this task on large data sets. In this work, I present a 121-source redshift catalog of obscured AGN with no literature redshift values, computed using the XZ Method. To produce the catalog, a ~400 source data set of obscured AGN was obtained from the Chandra Source Catalog 2.0 (CSC2) and selected primarily using the X-ray hardness ratio. ~100 such AGN had literature-documented spectroscopic or photometric redshifts, which were combined with a set of 1000 simulations to train a neural network model to identify satisfactory XZ redshift estimates. The fitted model was then applied to the remaining CSC2 AGN to yield a robust, minimally contaminated redshift catalog.

Comments
0
comment
No comments here