Release of Double Radio Galaxy (DRAGNs) Catalog and Machine Learning-Based Morphological Annotations for VLASS
Susan Xu (CIRADA) and Steven Myers (NRAO)
CIRADA is excited to announce the release of a double radio galaxy catalog and new machine learning classifications for the VLASS Quick Look epoch 2 data. The Double Radio Galaxy (DRAGNs) Catalog contains over 17,000 double-lobed radio galaxies and supplementary information such as host IDs and, where possible, redshift measurements. We also include a value-added list of more than 500,000 single-component radio sources. DRAGN catalogs are now available for each of the existing component catalogs, namely Quick Look 1, 2 and Single Epoch 2. Researchers can access the catalog and the accompanying software on our website at cirada.ca<http://cirada.ca/>. Detailed descriptions are provided in Gordon et al., 2023, ApJS, 267, 37.
CIRADA also now introduces morphological annotations of radio sources created using a Self-Organizing Map (SOM) for the epoch 2 data. This unsupervised machine learning technique classifies VLASS sources based on their distinct morphologies into 100 different classes or ‘neurons.’ Such classifications not only provide quality flags to identify bad data but are also beneficial for researchers interested in creating samples based on radio morphologies. Details of potential issues and guidance on such selection are provided in the Catalog User Guide at cirada.ca<http://cirada.ca/>.
With this release, we have also corrected any astrometry errors that corrupted the previous versions of the Quick Look epoch 1 catalog.
See NRAO E-News article.
