Overview

An artificial intelligence algorithm has identified over 5,000 potential gravitational lenses, a phenomenon that can help us understand the origin of the universe and all the matter that makes it up, including the elusive and mysterious dark matter. Experts have set to work to analyze the observations and extract relevant data on the distribution of mass, distance, and movement of distant galaxies, invisible without the help of gravitational lenses.

The science and other stuff to know

Data scientist Dr. Colin Jacobs of the Astro 3D Institute developed an algorithm based on Convolutional Neural Networks (CNN) technique, which allows gravitational lensing to be identified from telescope observations with almost 90% accuracy. The work was published in The Astronomical Journal.

Gravitational lenses are optical phenomena predicted by Einstein in his Theory of Relativity, which postulates that space and time are inseparable dynamic entity that distorts, stretches, and twists in the presence of matter. This ‘sinking’ produced by the mass in the surrounding region causes the light beams that pass through it to divert their trajectory. From here, we can see distant galaxies located just behind other massive galaxies thanks to the distortion caused by our galactic neighbor that allows light from the distant galaxy to pass through, making it perceptible.

Thanks to observations from the Keck Observatory in Hawai’i and the Very Large Telescope in Chile, this algorithm successfully identified 70 unknown gravitational lenses. Then, the research team quickly began studying the properties of these distant galaxies in search of dark matter. Even though it exists more than ordinary matter, dark matter is a theoretical entity that has never been detected because it does not interact with ordinary matter, light, or anything we know. But we know it’s there because its gravitational influence on galaxies is quantifiable.

So what?

The efficient and systematic detection of gravitational lensing would largely increase our database on the behavior of distant galaxies. Moreover, it would give us critical information to understand the role of unknown dark matter in large-scale structures’ formation and the universe’s evolution.

Dark matter is only appreciable on large scales, so having a method that offers us quantitative data on distant galaxies is, without a doubt, a sure step toward understanding this unfamiliar entity.

What’s next?

In addition to continuing to study the images of deep space provided by gravitational lensing, it is necessary to advance in other methods that provide us with data on the distribution of dark matter, as well as improve the current cosmological model. Only in this way do we get closer to unraveling the mysterious nature of this matter which, due to its omnipresence, actually seems to be the ordinary one.