An extinction free AGN selection by 18-band SED fitting in mid-infrared in the AKARI NEP deep field

TL;DR

This paper introduces an 18-band mid-infrared SED fitting method for AGN selection that effectively identifies obscured AGNs and outperforms previous techniques, using data from the AKARI NEP deep field.

Contribution

The study develops a novel 18-band mid-IR SED fitting approach combining AKARI, WISE, and Spitzer data to improve AGN detection, especially obscured ones, surpassing prior methods.

Findings

Recovered 20% more X-ray detected AGN than previous methods.

Identified an AGN fraction of approximately 30% in the sample.

Detected Seyfert 1.8 and 2 types often missed by earlier surveys.

Abstract

We have developed an efficient Active Galactic Nucleus (AGN) selection method using 18-band Spectral Energy Distribution (SED) fitting in mid-infrared (mid-IR). AGNs are often obscured by gas and dust, and those obscured AGNs tend to be missed in optical, UV and soft X-ray observations. Mid-IR light can help us to recover them in an obscuration free way using their thermal emission. On the other hand, Star-Forming Galaxies (SFG) also have strong PAH emission features in mid-IR. Hence, establishing an accurate method to separate populations of AGN and SFG is important. However, in previous mid-IR surveys, only 3 or 4 filters were available, and thus the selection was limited. We combined AKARI's continuous 9 mid-IR bands with WISE and Spitzer data to create 18 mid-IR bands for AGN selection. Among 4682 galaxies in the AKARI NEP deep field, 1388 are selected to be AGN hosts, which implies an AGN fraction of 29.6$\pm$0.8$\%$ (among them 47$\%$ are Seyfert 1.8 and 2). Comparing the result from SED fitting into WISE and Spitzer colour-colour diagram reveals that Seyferts are often missed by previous studies. Our result has been tested by stacking median magnitude for each sample. Using X-ray data from Chandra, we compared the result of our SED fitting with WISE's colour box selection. We recovered more X-ray detected AGN than previous methods by 20$\%$.