Streamlined Lensed Quasar Identification in Multiband Images via Ensemble Networks

Abstract

Quasars experiencing strong lensing offer unique viewpoints on subjects related to the cosmic expansion rate, the dark matter profile within the foreground deflectors, and the quasar host galaxies. Unfortunately, identifying them in astronomical images is challenging since they are overwhelmed by the abundance of non-lenses. To address this, we have developed a novel approach by ensembling cutting-edge convolutional networks (CNNs) -- for instance, ResNet, Inception, NASNet, MobileNet, EfficientNet, and RegNet -- along with vision transformers (ViTs) trained on realistic galaxy-quasar lens simulations based on the Hyper Suprime-Cam (HSC) multiband images. While the individual model exhibits remarkable performance when evaluated against the test dataset, achieving an area under the receiver operating characteristic curve of $>$97.3% and a median false positive rate of 3.6%, it struggles to generalize in real data, indicated by numerous spurious sources picked by each classifier. A significant improvement is achieved by averaging these CNNs and ViTs, resulting in the impurities being downsized by factors up to 50. Subsequently, combining the HSC images with the UKIRT, VISTA, and unWISE data, we retrieve approximately 60 million sources as parent samples and reduce this to 892,609 after employing a photometry preselection to discover $z>1.5$ lensed quasars with Einstein radii of ${\theta }_{\mathrm{E}}<5$ arcsec. Afterward, the ensemble classifier indicates 3080 sources with a high probability of being lenses, for which we visually inspect, yielding 210 prevailing candidates awaiting spectroscopic confirmation. These outcomes suggest that automated deep learning pipelines hold great potential in effectively detecting strong lenses in vast datasets with minimal manual visual inspection involved.

Accepted for publication in the Astronomy & Astrophysics journal. 28 pages, 11 figures, and 3 tables. We welcome comments from the reader

References in corpus (32)

• Array Programming with NumPy
• The Gaia mission
• The UKIRT Infrared Deep Sky Survey (UKIDSS)
• The propagation of uncertainties in stellar population synthesis modeling I: The relevance of uncertain aspects of stellar evolution and the IMF to the derived physical properties of galaxies
• EAZY: A Fast, Public Photometric Redshift Code
• The unWISE Catalog: Two Billion Infrared Sources from Five Years of WISE Imaging
• Photometric redshift and classification for the XMM-COSMOS sources
• Fast Automated Analysis of Strong Gravitational Lenses with Convolutional Neural Networks
• The Sloan Digital Sky Survey Quasar Lens Search. I. Candidate Selection Algorithm
• The UKIRT Hemisphere Survey: Definition and J-band Data Release
• LinKS: Discovering galaxy-scale strong lenses in the Kilo-Degree Survey using Convolutional Neural Networks
• The Cosmic Horseshoe: Discovery of an Einstein Ring around a Giant Luminous Red Galaxy
• The Sloan Digital Sky Survey Quasar Lens Search. II. Statistical Lens Sample from the Third Data Release
• Correcting the z~8 Galaxy Luminosity Function for Gravitational Lensing Magnification Bias
• Strong lensing time-delay cosmography in the 2020s
• PyAutoLens: Open-Source Strong Gravitational Lensing
• A survey of luminous high-redshift quasars with SDSS and WISE II. the bright end of the quasar luminosity function at z ~ 5
• AGN Populations in Large Volume X-ray Surveys: Photometric Redshifts and Population Types found in the Stripe 82X Survey
• Data Mining for Gravitationally Lensed Quasars
• HOLISMOKES. VI. New galaxy-scale strong lens candidates from the HSC-SSP imaging survey
• HOLISMOKES. VIII. High-redshift, strong-lens search in the Hyper Suprime-Cam Subaru Strategic Program
• The Sloan Digital Sky Survey Quasar Lens Search. IV. Statistical Lens Sample from the Fifth Data Release
• Discovery of Four Gravitationally Lensed Quasars from the Sloan Digital Sky Survey
• Discovery of a Gravitationally Lensed Quasar from the Sloan Digital Sky Survey: SDSS J133222.62+034739.9
• Detecting gravitational lenses using machine learning: exploring interpretability and sensitivity to rare lensing configurations
• Red quasars blow out molecular gas from galaxies during the peak of cosmic star formation
• Revisiting the Lensed Fraction of High-Redshift Quasars
• A Survey for High-redshift Gravitationally Lensed Quasars and Close Quasars Pairs. I. the Discoveries of an Intermediately-lensed Quasar and a Kpc-scale Quasar Pair at $z\sim 5$
• A machine learning based approach to gravitational lens identification with the International LOFAR Telescope
• A Mock Catalog of Gravitationally Lensed Quasars for the LSST Survey
• ULAS J234311.93-005034.0: A gravitational lens system selected from UKIDSS and SDSS
• Discovering strongly lensed quasar candidates with catalogue-based methods from DESI Legacy Surveys