Accelerating lensed quasar discovery and modeling with physics-informed variational autoencoders

¹ Technical University of Munich, TUM School of Natural Sciences, Department of Physics, James-Franck-Str. 1, D-85748 Garching, Germany
² Max-Planck-Institut für Astrophysik, Karl-Schwarzschild-Str. 1, D-85748 Garching, Germany
³ Dipartimento di Fisica, Università degli Studi di Milano, Via Celoria 16, I-20133 Milano, Italy
⁴ INAF – IASF Milano, Via A. Corti 12, I-20133 Milano, Italy
⁵ Academia Sinica Institute of Astronomy and Astrophysics (ASIAA), 11F of ASMAB, No. 1, Section 4, Roosevelt Road, Taipei 10617, Taiwan
⁶ Aix-Marseille Université, CNRS, CNES, LAM, Marseille, France
⁷ Department of Astrophysics, American Museum of Natural History, Central Park West and 79th Street, New York, NY 10024-5192, USA
⁸ Department of Physics and Astronomy, Lehman College of the CUNY, Bronx, NY 10468, USA

^⋆ Corresponding author; irham.andika@tum.de

Received: 17 December 2024
Accepted: 17 January 2025

Abstract

Strongly lensed quasars provide valuable insights into the rate of cosmic expansion, the distribution of dark matter in foreground deflectors, and the characteristics of quasar hosts. However, detecting them in astronomical images is difficult due to the prevalence of non-lensing objects. To address this challenge, we developed a generative deep learning model called VariLens, built upon a physics-informed variational autoencoder. This model seamlessly integrates three essential modules: image reconstruction, object classification, and lens modeling, offering a fast and comprehensive approach to strong lens analysis. VariLens is capable of rapidly determining both (1) the probability that an object is a lens system and (2) key parameters of a singular isothermal ellipsoid (SIE) mass model – including the Einstein radius (θ_E), lens center, and ellipticity – in just milliseconds using a single CPU. A direct comparison of VariLens estimates with traditional lens modeling for 20 known lensed quasars within the Subaru Hyper Suprime-Cam (HSC) footprint shows good agreement, with both results consistent within 2σ for systems with θ_E < 3″. To identify new lensed quasar candidates, we began with an initial sample of approximately 80 million sources, combining HSC data with multiwavelength information from Gaia, UKIRT, VISTA, WISE, eROSITA, and VLA. After applying a photometric preselection aimed at locating z > 1.5 sources, the number of candidates was reduced to 710 966. Subsequently, VariLens highlights 13 831 sources, each showing a high likelihood of being a lens. A visual assessment of these objects results in 42 promising candidates that await spectroscopic confirmation. These results underscore the potential of automated deep learning pipelines to efficiently detect and model strong lenses in large datasets, substantially reducing the need for manual inspection.