Researchers from the Department of Astronomy and Space Informatics at Moscow State University, in collaboration with an international team, have developed an innovative method that allows for the assessment of conditions for life on Earth-like planets located around red dwarfs—the most common stars in our Galaxy. This work, supported by the Russian Science Foundation (RSF), was published in the journal Monthly Notices of the Royal Astronomical Society.
The scientists focused on red dwarfs, which, like our Sun, experience flares caused by changes in their magnetic fields. These flares can release powerful bursts of energy lasting several minutes or hours.
As noted by RSF specialists, red dwarfs have significantly stronger magnetic fields compared to yellow dwarfs, making their flares more powerful. The high activity of these stars can negatively impact the potential for life to arise on exoplanets, while a stable magnetic field may indicate a greater potential for habitability.
A group of scientists from the Department of Astronomy and Space Informatics at Moscow State University and their foreign colleagues proposed a new method for analyzing the activity of red dwarfs by recording each flare on these stars. To this end, a catalog of flare images was created, taking into account interference from satellites and other defects that arise during telescope observations.
Capturing such rapidly changing phenomena presents a complex challenge that requires processing vast amounts of data obtained from telescopes. Consequently, machine learning methods were employed to analyze images of the starry sky and search for active red dwarfs.
“When applying machine learning, it is first necessary to train the algorithm. However, the number of real photographs of flares was insufficient, so researchers artificially created such images using data from real flares recorded by space telescopes and overlaying them on images of stars in their 'calm' state. This ensured the creation of a realistic training dataset,” explained the RSF.
According to the new approach, it was found that the most active flares occur on red dwarfs of spectral class M4 with a temperature of about 3000 °C, while hotter stars of class M0, which have a temperature of about 3800 °C, exhibit flares significantly less frequently. It was also discovered that flare activity decreases with the distance of the red dwarf from the plane of the Milky Way: the farther the star, the older it is, and the weaker its magnetic field.
In the future, researchers plan to specifically search for flares in star clusters to improve the understanding of stellar evolution and identify the most stable red dwarfs, as commented by laboratory assistant Anastasia Lavrukhina from the Department of Astronomy and Space Informatics at Moscow State University.
