Physicists have developed a new method for measuring the expansion of the Universe that could be four times more accurate than current approaches. The method relies on data from stellar explosions in galaxies, which is analyzed by artificial intelligence.
Standard candles
Type Ia supernovae occur when white dwarfs explode. All such explosions have nearly the same absolute brightness, so astronomers know how bright such a supernova should appear from any distance. By comparing this with the actual observed brightness, researchers determine the distances to distant galaxies and thus measure the expansion of the Universe.
This method helped establish in 1998 the fact that the universe is expanding at an accelerating rate. This phenomenon is linked to dark energy, one of the major unsolved mysteries of physics. However, it turns out that Type Ia supernovae are not all identical. Over the past twenty years, researchers have discovered that the galaxy where the explosion occurred influences the recorded brightness of the flash.
An all-encompassing model
An international team led by researchers from the Institute of Space Sciences at the University of Barcelona (ICCUB) has developed a tool called CIGaRS. It models several interrelated factors at once: the parameters of the explosions, the characteristics of the host galaxies, interstellar dust that absorbs light and shifts it toward the red end of the spectrum, the frequency of supernovae throughout cosmic history, and the expansion of the universe.
Instead of considering each factor separately, the team developed a single, integrated model that links them through physical and statistical relationships.
Role of artificial intelligence
To make this large-scale approach work in practice, the researchers used a method of statistical inference based on simulations. First, the scientists generate thousands of simulated universes based on physical equations. Then, a neural network is trained on these simulations to identify how the observed data relates to fundamental parameters.
Once trained, the system can analyze actual observations and estimate these parameters simultaneously for tens of thousands of supernovae. Traditional methods cannot handle such a large scale.
Preparing for new data
The Vera Rubin Observatory, currently under construction in Chile, will begin a decade-long survey of the sky in a few years. It is estimated that about 99% of the supernovae it detects will be recorded only as images in different colors, without spectroscopic observations.
The Vera Rubin Observatory, recently completed in Chile, has already taken its first images and will soon begin a ten-year survey of the sky. It is predicted about 99% of supernovae which will be received in the future will be observed photometrically only, using direct imaging in different colors, without spectroscopy. The CIGaRS instrument was designed specifically to process such data.
In addition to refining measurements of dark energy, the method could help us better understand how and when Type Ia supernovae form. By reconstructing the relationship between the frequency of explosions and the age of stars in galaxies, the model opens the door to studying long-standing questions about the stellar systems that give rise to these explosions. The study was published in the journal Nature Astronomy.
According to scitechdaily.com
