The Discovery of 63 Giant Radio Galaxies in the FIRST Survey

Giant Radio Galaxies (GRGs) are Active Galactic Nuclei (AGN) with radio emission that extends over projected sizes $>0.7\,$Mpc. The large angular sizes associated with GRGs complicate their identification in radio survey images using traditional source finders. In this Note, we use DRAGNhunter, an algorithm designed to find double-lobed radio galaxies, to search for GRGs in the Faint Images of the Radio Sky at Twenty cm survey (FIRST). Radio and optical images of identified candidates are visually inspected to confirm their authenticity, resulting in the discovery of $63$ previously unreported GRGs.


INTRODUCTION
Radio galaxies with a projected "Largest Linear Size" (LLS) over 700 kpc are considered Giant Radio Galaxies (GRGs, e.g., Andernach et al. 2021).Their jets should be intrinsically powerful (Wiita et al. 1989) and relatively uninhibited (Dabhade et al. 2020a) to reach such sizes, and can provide insights into the AGN central engine and host galaxy environment.GRGs can be difficult to identify in radio surveys: for a classic Double Radio Source associated with an Active Galactic Nucleus (DRAGN, Leahy 1993), the lobes of the radio galaxy may be detected as multiple sources despite belonging to the same physical object.Consequently, many detectable GRGs may remain unidentified.
Recently, Gordon et al. (2023, hereafter G23) introduced DRAGNhunter, an algorithm that groups 'extended components' into physical sources, and applied it to the Very Large Array Sky Survey (VLASS, Lacy et al. 2020), discovering 31 new GRGs in the process.Given the lower frequency (1.4 GHz) and better sensitivity to extended sources of FIRST (Becker et al. 1995), we used DRAGNhunter on the FIRST catalog (Helfand et al. 2015) to identify further unidentified GRGs.

IDENTIFYING CANDIDATE GRGS
DRAGNhunter identifies DRAGNs by pairing cataloged extended radio sources based on their separation and relative alignment, then uses the likelihood ratio approach (McAlpine et al. 2012) to search for the probable host galaxy in the AllWISE catalog (Cutri et al. 2014).Hosts are then cross-matched with a number of legacy catalogs to obtain redshifts (for details see Sections 2-4 of G23).Selecting extended FIRST sources as having deconvolved sizes > 5.5 ′′ , DRAGNhunter finds > 9, 000 DRAGNs with a probable host and redshift.We determine their LLS from their measured Largest Angular Size (LAS) assuming a flat ΛCDM cosmology with H 0 = 70 km/s/Mpc, Ω M = 0.3, and Ω Λ = 0.7.We obtain LLS > 0.7 Mpc for 213 DRAGNs, but because DRAGNhunter is an automated algorithm, we inspect their radio and optical images to exclude any mistaken groupings of independent sources, and ensure that the identified host and redshift are correct.Furthermore, we manually remeasure the LAS that DRAGNhunter automatically estimates, ensuring these truly are giants.In total this leaves us with 80 GRGs identified by DRAGNhunter in FIRST.
Only 6 of our GRGs have spectroscopic redshifts, all from the Sloan Digital Sky Survey (SDSS-DR16, Ahumada et al. 2020).The remaining GRGs have photometric redshifts (photo-zs) obtained from the Dark Energy Spectroscopic Instrument imaging Legacy Surveys (LS-DR8, Duncan 2022).Alternative photo-zs are available for many of our GRGs, and the choice of photo-z may impact the GRG classification for sources with LLS ≈ 0.7 Mpc.However, Duncan (2022) test the robustness of their photo-zs for radio galaxies, finding them to be reliable over the redshift range of our sample, thus these are an appropriate choice here.
Although none of our 63 sources have been reported as GRGs, they may have been identified as extended radio galaxies in previous catalogs (e.g., Amirkhanyan 2009;Proctor 2011).For example, we measure LAS = 96 ′′ for SDSS J025551.47-025939.2, whereas Kuźmicz & Jamrozy (2021) only measure 80 ′′ leading to LLS = 0.64 Mpc in their work.Here, diffuse structure in the northern lobe extends beyond the hotspot used to measure the LAS by Kuźmicz & Jamrozy (2021).Finally, we note that DRAGNhunter is biased toward finding relatively unbent and symmetric FR II galaxies (G23).Consequently, GRGs with atypical morphologies will likely be missed.Nonetheless, this Note demonstrates the usefulness of automated routines in finding GRGs.2003), Topcat (Taylor 2005)