Beyond Traditional Diagnostics: Identifying Active Galactic Nuclei with Spectral Energy Distribution Fitting in DESI Data

TL;DR

This paper demonstrates that spectral energy distribution fitting using DESI data effectively identifies active galactic nuclei, complementing traditional methods and enabling more complete AGN samples for large surveys.

Contribution

It introduces a robust SED fitting approach for AGN detection that captures sources missed by conventional diagnostics, improving completeness of AGN samples.

Findings

SED fitting identifies 70% of narrow/broad-line AGN

87% of WISE-selected AGN are detected via SED fitting

Incorporating high SNR WISE data reduces contamination from star-forming galaxies

Abstract

Active galactic nuclei (AGN) are typically identified through their distinctive X-ray or radio emissions, mid-infrared (MIR) colors, or emission lines. However, each method captures different subsets of AGN due to signal-to-noise (SNR) limitations, redshift coverage, and extinction effects, underscoring the necessity for a multi-wavelength approach for comprehensive AGN samples. This study explores the effectiveness of spectral energy distribution (SED) fitting as a robust method for AGN identification. Using {\tt CIGALE} optical-MIR SED fits on DESI Early Data Release galaxies, we compare SED-based AGN selection ({\tt AGNFRAC} $\geq0.1$) with traditional methods including BPT diagrams, WISE colors, X-ray, and radio diagnostics. SED fitting identifies $\sim 70\%$ of narrow/broad-line AGN and 87\% of WISE-selected AGN. Incorporating high SNR WISE photometry reduces star-forming galaxy contamination from 62\% to 15\%. Initially, $\sim50\%$ of SED-AGN candidates are undetected by standard methods, but additional diagnostics classify $\sim85\%$ of these sources, revealing LINERs and retired galaxies potentially representing evolved systems with weak AGN activity. Further spectroscopic and multi-wavelength analysis will be essential to determine the true AGN nature of these sources. SED fitting provides complementary AGN identification, unifying multi-wavelength AGN selections. This approach enables more complete -- albeit with some contamination -- AGN samples essential for upcoming large-scale surveys where spectroscopic diagnostics may be limited.