Extended Radio Galaxies in EMU: A Comparative Look at Source-Finding Techniques

Abstract

Extended radio sources present unique challenges for automated detection and classification in wide-field radio surveys. With current surveys such as the Evolutionary Map of the Universe (EMU), robust and scalable methods are essential to identify and catalogue these complex sources. We apply three automatic approaches to detect complex radio emission in EMU observations of the Galaxy And Mass Assembly (GAMA) 09 field (EMU-G09) in order to evaluate their relative strengths and limitations in preparation for large-scale application across future EMU data releases. These include DRAGNHunter, designed to detect likely DRAGNs (Double Radio sources associated with Active Galactic Nuclei) from a component catalogue; coarse-grained complexity, a metric designed to highlight regions of complex emission; and RG-CAT, a machine learning pipeline trained on radio sources identified in the EMU pilot survey. We find that together, the three methods recover nearly all extended sources in EMU-G09 but identify largely distinct, partially-overlapping subsets, with only 375 sources identified by all finders. This demonstrates that a combination of complementary techniques will be required to achieve a complete census of extended radio sources in future large-scale surveys.

Figures (27)

• Figure 1: EMU mosaic constructed from the three EMU tiles that overlap the GAMA09 region with the black box indicating the GAMA09 footprint. The EMU observations are centred on a frequency of $944$ MHz, a native resolution of $11"\times13"$, and an rms sensitivity of $\sim\!25\,\mu$Jy beam$^{-1}$. These data constitute the EMU-G09 region used throughout this paper.
• Figure 2: Distributions of angular separation of identified pairs in EMU-G09 (red bars) in bins of mean misalignment. Black lines mark the local minima, with the shaded region showing the $1\sigma$ uncertainty of their positions.
• Figure 3: Mean misalignment and angular separation for candidate lobe pairs in EMU-G09. Black crosses show the local minima of pair separation in bins of mean misalignment, with vertical bars indicating bin size and horizontal bars the $1\sigma$ uncertainty. The black dashed and grey solid lines show the derived upper limits for likely DRAGNs in EMU and VLASS, respectively, reflecting their different observational parameters. The black dotted line represents the minimum pair separation ($15"$) used to select DRAGNs in this work.
• Figure 4: Distributions of LAS and integrated flux density ($S$) for likely DRAGNs in the EMU G09 field. Density contours of the distributions of LAS and $S$ for EMU DRAGNs (red) for the VLASS DRAGNs (grey). Contours contain $90\%$, $50\%$, and $10\%$ of the data points. The black dotted line represents the minimum pair separation ($15"$) used to select DRAGNs in this work. The difference between these distributions likely arises from the observational differences between EMU and VLASS.
• Figure 5: Illustrative example of the coarse-grained complexity process. An input image (left) is smoothed by a Gaussian filter to produce the smoothed image (right) which is then compressed using gzip. The byte length is then used as a proxy for apparent complexity of the radio source. The images here are $64\times64$ pixels, corresponding to an angular size of $128"\times128"$.