ML-Driven Strong Lens Discoveries: Down to $\theta_E \sim 0.03''$ and $M_\mathrm{halo}< 10^{11} M_\odot$

TL;DR

This paper demonstrates that combining JWST observations with machine learning techniques enables the detection of extremely small Einstein radii and low-mass halos in strong gravitational lensing, surpassing previous capabilities.

Contribution

The study introduces a novel ML approach to identify low-mass halo strong lenses with very small Einstein radii using JWST data, extending the discovery space significantly.

Findings

Achieved near-100% completeness and purity for conventional lenses ($ heta_E extgreater 0.5''$).

Predicted JWST can find about 17 low-mass lenses per square degree.

Validated ML models by discovering new lens candidates in HST images.

Abstract

We present results on extending the strong lens discovery space down to much smaller Einstein radii ($\theta_E\lesssim0.03''$) and much lower halo mass ($M_\mathrm{halo}<10^{11}M_\odot$) through the combination of JWST observations and machine learning (ML) techniques. First, we forecast detectable strong lenses with JWST using CosmoDC2 as the lens catalog, and a source catalog down to 29th magnitude. By further incorporating the VELA hydrodynamical simulations of high-redshift galaxies, we simulate strong lenses. We train a ResNet on these images, achieving near-100\% completeness and purity for ``conventional" strong lenses ($\theta_E\gtrsim 0.5''$), applicable to JWST, HST, the Roman Space Telescope and Euclid VIS. For the first time, we also search for very low halo mass strong lenses ($M_{halo}<10^{11}M_\odot$) in simulations, with $\theta_E\ll 0.5''$, down to the best resolution ($0.03''$) and depth (10,000~sec) limits of JWST using ResNet. A U-Net model is employed to pinpoint these small lenses in images, which are otherwise virtually impossible for human detection. Our results indicate that JWST can find $\sim 17$/deg$^2$ such low-halo-mass lenses, with the locations of $\sim 1.1$/deg$^2$ of these detectable by the U-Net at $\sim100$\% precision (and $\sim 7.0$/deg$^2$ at a 99.0\% precision). To validate our model for finding ``conventional" strong lenses, we apply it to HST images, discovering two new strong lens candidates previously missed by human classifiers in a crowdsourcing project (Garvin et al. 2022). This study demonstrates the (potentially ``superhuman") advantages of ML combined with current and future space telescopes for detecting conventional, and especially, low-halo-mass strong lenses, which are critical for testing CDM models.