Strong Gravitational Lens Candidates Research Articles

Astronomaly at scale: searching for anomalies amongst 4 million galaxies

ABSTRACT Modern astronomical surveys are producing data sets of unprecedented size and richness, increasing the potential for high-impact scientific discovery. This possibility, coupled with the challenge of exploring a large number of sources, has led to the development of novel machine-learning-based anomaly detection approaches, such as astronomaly. For the first time, we test the scalability of astronomaly by applying it to almost 4 million images of galaxies from the Dark Energy Camera Legacy Survey. We use a trained deep learning algorithm to learn useful representations of the images and pass these to the anomaly detection algorithm isolation forest, coupled with astronomaly’s active learning method, to discover interesting sources. We find that data selection criteria have a significant impact on the trade-off between finding rare sources such as strong lenses and introducing artefacts into the data set. We demonstrate that active learning is required to identify the most interesting sources and reduce artefacts, while anomaly detection methods alone are insufficient. Using astronomaly, we find 1635 anomalies among the top 2000 sources in the data set after applying active learning, including eight strong gravitational lens candidates, 1609 galaxy merger candidates, and 18 previously unidentified sources exhibiting highly unusual morphology. Our results show that by leveraging the human–machine interface, astronomaly is able to rapidly identify sources of scientific interest even in large data sets.

The impact of human expert visual inspection on the discovery of strong gravitational lenses

ABSTRACT We investigate the ability of human ‘expert’ classifiers to identify strong gravitational lens candidates in Dark Energy Survey like imaging. We recruited a total of 55 people that completed more than 25 per cent of the project. During the classification task, we present to the participants 1489 images. The sample contains a variety of data including lens simulations, real lenses, non-lens examples, and unlabelled data. We find that experts are extremely good at finding bright, well-resolved Einstein rings, while arcs with g-band signal to noise less than ∼25 or Einstein radii less than ∼1.2 times the seeing are rarely recovered. Very few non-lenses are scored highly. There is substantial variation in the performance of individual classifiers, but they do not appear to depend on the classifier’s experience, confidence or academic position. These variations can be mitigated with a team of 6 or more independent classifiers. Our results give confidence that humans are a reliable pruning step for lens candidates, providing pure and quantifiably complete samples for follow-up studies.

DenseLens – Using DenseNet ensembles and information criteria for finding and rank-ordering strong gravitational lenses

ABSTRACT Convolutional neural networks (CNNs) are the state-of-the-art technique for identifying strong gravitational lenses. Although they are highly successful in recovering genuine lens systems with a high true-positive rate, the unbalanced nature of the data set (lens systems are rare), still leads to a high false positive rate. For these techniques to be successful in upcoming surveys (e.g. with Euclid) most emphasis should be set on reducing false positives, rather than on reducing false negatives. In this paper, we introduce densely connected neural networks (DenseNets) as the CNN architecture in a new pipeline-ensemble model containing an ensemble of classification CNNs and regression CNNs to classify and rank-order lenses, respectively. We show that DenseNets achieve comparable true positive rates but considerably lower false positive rates (when compared to residual networks; ResNets). Thus, we recommend DenseNets for future missions involving large data sets, such as Euclid, where low false positive rates play a key role in the automated follow-up and analysis of large numbers of strong gravitational lens candidates when human vetting is no longer feasible.

Hubble Asteroid Hunter

Context. The Hubble Space Telescope (HST) archives constitute a rich dataset of high-resolution images to mine for strong gravitational lenses. While many HST programmes specifically target strong lenses, they can also be present by coincidence in other HST observations. Aims. Our aim is to identify non-targeted strong gravitational lenses, without any prior selection on the lens properties, in almost two decades of images from the ESA HST archive (eHST). Methods. We used crowdsourcing on the Hubble Asteroid Hunter (HAH) citizen science project to identify strong lenses, along with asteroid trails, in publicly available large field-of-view HST images. We visually inspected 2354 objects tagged by citizen scientists as strong lenses to clean the sample and identify the genuine lenses. Results. We report the detection of 252 strong gravitational lens candidates, which were not the primary targets of the HST observations. A total of 198 of them are new, not previously reported by other studies, consisting of 45 A grades, 74 B grades and 79 C grades. The majority are galaxy-galaxy configurations. The newly detected lenses are, on average, 1.3 magnitudes fainter than previous HST searches. This sample of strong lenses with high-resolution HST imaging is ideal to follow up with spectroscopy for lens modelling and scientific analyses. Conclusions. This paper presents the unbiased search of lenses that enabled us to find a wide variety of lens configurations, including exotic lenses. We demonstrate the power of crowdsourcing in visually identifying strong lenses and the benefits of exploring large archival datasets. This study shows the potential of using crowdsourcing in combination with artificial intelligence for the detection and validation of strong lenses in future large-scale surveys such as ESA’s Euclid mission or in James Webb Space Telescope (JWST) archival images.

Mining for Strong Gravitational Lenses with Self-supervised Learning

We employ self-supervised representation learning to distill information from 76 million galaxy images from the Dark Energy Spectroscopic Instrument Legacy Imaging Surveys’ Data Release 9. Targeting the identification of new strong gravitational lens candidates, we first create a rapid similarity search tool to discover new strong lenses given only a single labeled example. We then show how training a simple linear classifier on the self-supervised representations, requiring only a few minutes on a CPU, can automatically classify strong lenses with great efficiency. We present 1192 new strong lens candidates that we identified through a brief visual identification campaign and release an interactive web-based similarity search tool and the top network predictions to facilitate crowd-sourcing rapid discovery of additional strong gravitational lenses and other rare objects: github.com/georgestein/ssl-legacysurvey.

Gaia GraL: Gaia DR2 Gravitational Lens Systems

Aims. In this work, we aim to provide a reliable list of gravitational lens candidates based on a search performed over the entire Gaia Data Release 2 (Gaia DR2). We also aim to show that the astrometric and photometric information coming from the Gaia satellite yield sufficient insights for supervised learning methods to automatically identify strong gravitational lens candidates with an efficiency that is comparable to methods based on image processing. Methods. We simulated 106 623 188 lens systems composed of more than two images, based on a regular grid of parameters characterizing a non-singular isothermal ellipsoid lens model in the presence of an external shear. These simulations are used as an input for training and testing our supervised learning models consisting of extremely randomized trees (ERTs). These trees are finally used to assign to each of the 2 129 659 clusters of celestial objects extracted from the Gaia DR2 a discriminant value that reflects the ability of our simulations to match the observed relative positions and fluxes from each cluster. Once complemented with additional constraints, these discriminant values allow us to identify strong gravitational lens candidates out of the list of clusters. Results. We report the discovery of 15 new quadruply-imaged lens candidates with angular separations of less than 6″ and assess the performance of our approach by recovering 12 of the 13 known quadruply-imaged systems with all their components detected in Gaia DR2 with a misclassification rate of fortuitous clusters of stars as lens systems that is below 1%. Similarly, the identification capability of our method regarding quadruply-imaged systems where three images are detected in Gaia DR2 is assessed by recovering 10 of the 13 known quadruply-imaged systems having one of their constituting images discarded. The associated misclassification rate varies between 5.83% and 20%, depending on the image we decided to remove.

Testing Convolutional Neural Networks for finding strong gravitational lenses in KiDS

Convolutional Neural Networks (ConvNets) are one of the most promising methods for identifying strong gravitational lens candidates in survey data. We present two ConvNet lens-finders which we have trained with a dataset composed of real galaxies from the Kilo Degree Survey (KiDS) and simulated lensed sources. One ConvNet is trained with single \textit{r}-band galaxy images, hence basing the classification mostly on the morphology. While the other ConvNet is trained on \textit{g-r-i} composite images, relying mostly on colours and morphology. We have tested the ConvNet lens-finders on a sample of 21789 Luminous Red Galaxies (LRGs) selected from KiDS and we have analyzed and compared the results with our previous ConvNet lens-finder on the same sample. The new lens-finders achieve a higher accuracy and completeness in identifying gravitational lens candidates, especially the single-band ConvNet. Our analysis indicates that this is mainly due to improved simulations of the lensed sources. In particular, the single-band ConvNet can select a sample of lens candidates with $\sim40\%$ purity, retrieving 3 out of 4 of the confirmed gravitational lenses in the LRG sample. With this particular setup and limited human intervention, it will be possible to retrieve, in future surveys such as Euclid, a sample of lenses exceeding in size the total number of currently known gravitational lenses.

The DES Bright Arcs Survey: Hundreds of Candidate Strongly Lensed Galaxy Systems from the Dark Energy Survey Science Verification and Year 1 Observations

Abstract We report the results of searches for strong gravitational lens systems in the Dark Energy Survey (DES) Science Verification and Year 1 observations. The Science Verification data span approximately 250 sq. deg. with a median i-band limiting magnitude for extended objects (10σ) of 23.0. The Year 1 data span approximately 2000 sq. deg. and have an i-band limiting magnitude for extended objects (10σ) of 22.9. As these data sets are both wide and deep, they are particularly useful for identifying strong gravitational lens candidates. Potential strong gravitational lens candidate systems were initially identified based on a color and magnitude selection in the DES object catalogs or because the system is at the location of a previously identified galaxy cluster. Cutout images of potential candidates were then visually scanned using an object viewer and numerically ranked according to whether or not we judged them to be likely strong gravitational lens systems. Having scanned nearly 400,000 cutouts, we present 374 candidate strong lens systems, of which 348 are identified for the first time. We provide the R.A. and decl., the magnitudes and photometric properties of the lens and source objects, and the distance (radius) of the source(s) from the lens center for each system.

The Sloan Lens ACS Survey. V. The Full ACS Strong‐Lens Sample

We present the definitive data for the full sample of 131 strong gravitational lens candidates observed with the Advanced Camera for Surveys (ACS) aboard the Hubble Space Telescope by the Sloan Lens ACS (SLACS) Survey. All targets were selected for higher redshift emission lines and lower redshift continuum in a single Sloan Digital Sky Survey (SDSS) spectrum. The foreground galaxies are primarily of early-type morphology, with redshifts from z similar or equal to 0.05 to 0.5 and velocity dispersions from sigma similar or equal to 160 to 400 km s(-1); the faint background emission-line galaxies have redshifts ranging from z similar or equal to 0.2 to 1.2. We confirm 70 systems showing clear evidence of multiple imaging of the background galaxy by the foreground galaxy, as well as an additional 19 systems with probable multiple imaging. For 63 clear lensing systems, we present singular isothermal ellipsoid and light-traces-mass gravitational lens models fitted to the ACS imaging data. These strong-lensing mass measurements are supplemented by magnitudes and effective radii measured from ACS surface brightness photometry and redshifts and velocity dispersions measured from SDSS spectroscopy. These data constitute a unique resource for the quantitative study of the interrelations between mass, light, and kinematics in massive early-type galaxies. We show that the SLACS lens sample is statistically consistent with being drawn at random from a parent sample of SDSS galaxies with comparable spectroscopic parameters and effective radii, suggesting that the results of SLACS analyses can be generalized to the massive early-type population.

Three Gravitational Lenses for the Price of One: Enhanced Strong Lensing through Galaxy Clustering

We report the serendipitous discovery of two strong gravitational lens candidates (ACS J160919+6532 and ACS J160910+6532) in deep images obtained with the Advanced Camera for Surveys on the Hubble Space Telescope, each less than 40 arcsec from the previously known gravitational lens system CLASS B1608+656. The redshifts of both lens galaxies have been measured with Keck and Gemini: one is a member of a small galaxy group at z~0.63, which also includes the lensing galaxy in the B1608+656 system, and the second is a member of a foreground group at z~0.43. By measuring the effective radii and surface brightnesses of the two lens galaxies, we infer their velocity dispersions based on the passively evolving Fundamental Plane (FP) relation. Elliptical isothermal lens mass models are able to explain their image configurations within the lens hypothesis, with a velocity dispersion compatible with that estimated from the FP for a reasonable source-redshift range. Based on the large number of massive early-type galaxies in the field and the number-density of faint blue galaxies, the presence of two additional lens systems around CLASS B1608+656 is not unlikely in hindsight. Gravitational lens galaxies are predominantly early-type galaxies, which are clustered, and the lensed quasar host galaxies are also clustered. Therefore, obtaining deep high-resolution images of the fields around known strong lens systems is an excellent method of enhancing the probability of finding additional strong gravitational lens systems.

Sloan Digital Sky Survey Spectroscopic Lens Search. I. Discovery of Intermediate-Redshift Star-forming Galaxies behind Foreground Luminous Red Galaxies

We present a catalog of 49 spectroscopic strong gravitational lens candidates selected from a Sloan Digital Sky Survey sample of 50996 luminous red galaxies. Potentially lensed star-forming galaxies are detected through the presence of background oxygen and hydrogen nebular emission lines in the spectra of these massive foreground galaxies. This multiline selection eliminates the ambiguity of single-line identification and provides a very promising sample of candidate galaxy-galaxy lens systems at low to intermediate redshift, with foreground redshifts ranging from 0.16 to 0.49 and background redshifts from 0.25 to 0.81. Any lenses confirmed within our sample would be important new probes of early-type galaxy mass distributions, providing complementary constraints to those obtained from currently known lensed high-redshift quasars.