Everett McArthur, a Venezuelan ethnobotanist, and his research team conducted a cutting-edge study. They used machine learning algorithms to find quasars that act as gravitational lenses. With this groundbreaking method, the team was able to analyze more than 812,000 quasars, increasing our knowledge of these incredible cosmic powerhouses more than ever before. Their findings were published on the arXiv preprint server under the title “Quasars acting as Strong Lenses Found in DESI DR1.”
The Sloan Digital Sky Survey (SDSS) has now cataloged almost 300,000 quasars. Prior to this study, just twelve candidates had been confirmed as quasars serving as lenses. Of those, only three were confirmed to be genuine lensing quasars. To map the suburbs, McArthur’s team used data from the Dark Energy Spectroscopic Instrument (DESI). To make it easier to identify these pieces, they recently put an advanced machine learning technique to work.
From this, the researchers trained a neural network on about 3,000 synthetic lenses and 30,000 regular quasar spectra. That kind of fastidious training brought outstanding dividends. The network obtained a high classification performance area under the curve score of 0.99. The new high precision was enough for the team to identify seven new high-quality candidates for quasars acting as lenses.
Every one of these newly identified quasars displays a strong oxygen doublet emission line. This line being found at a greater redshift than that of the foreground quasar indicates that these quasars may make excellent gravitational lenses. Remarkably, four of these quasars showed strong hydrogen beta and oxygen three emissions from the host galaxy in the background. This makes their classification even more ironclad.
The application of machine learning to this research represents a promising new step in the space of astronomical studies. To make things even more efficient, McArthur and his team have automated the identification process. This discovery has opened a wealth of knowledge about quasar pumping mechanisms and their role in shaping ambient cosmic structures.