Artificial intelligence helps scientists understand cosmic explosions

Scientists at the University of Warwick are using artificial intelligence (AI) to analyze cosmic explosions known as supernovae. Their work is published in Monthly Notices of the Royal Astronomical Society.

Many stars in the universe will end their lives as white dwarfs – compact stars containing about the mass of the Earth the size of the Sun. Some of these white dwarfs eventually explode as supernovae. This process is highly energetic and results in the creation of heavy elements that are the building blocks of life, such as calcium and iron, which are released back into space.

Despite their importance, astronomers still do not know exactly how and why these supernovae occur.

To gain a better understanding, the new research will use a type of artificial intelligence known as machine learning to speed up supernova experiments – processes that are currently very computationally expensive and time-consuming. This will help reveal how these cosmic explosions took place by comparing explosion models to real-life observations.

Lead author Dr. Mark Magee, from the Department of Physics, University of Warwick, said: “When we study supernovae, we analyze their spectra. Spectra show the intensity of light at different wavelengths, which is affected by the elements created in the supernova.” Each element interacts with light at unique wavelengths and therefore leaves a unique signature on spectra.

“Analyzing these signatures can help identify what elements are created in a supernova and provide more details about how supernovae explode.

“Based on this data, we build models that are compared to real supernovae to find out what type of supernova it is and exactly how it exploded. It can typically take 10-90 minutes to generate one model, and we want to compare hundreds or thousands of models to get a complete understanding supernovae This is not possible in many cases.

“Our new research will step away from this tedious process. We will train machine learning algorithms on what different types of explosions look like and use them to generate models much faster. In a similar way to how we can use artificial intelligence to create new works of art.” or text, we will now be able to generate supernova simulations, which means we will be able to generate thousands of models in less than a second, which will be a huge boost to supernova research.”

In addition to speeding up the supernova analysis process, the use of AI will also enable better accuracy in research. This will help determine which models most closely resemble real explosions and the range of their physical properties.

Dr. Magee added: “Examining the elements released by supernovae is a crucial step in determining the type of explosion that occurred, as certain types of explosions produce more of certain elements than others. We can then relate the properties of the explosion back to the properties of the supernova’s host galaxies, creating a direct link between how the explosion occurred, and the type of white dwarf that exploded.”

The job accepted now is just the first step. Future research will expand to include an even larger range of explosions and supernovae and directly link the properties of the explosion and the host galaxy. Such research is now only possible thanks to advances in machine learning.

Dr. Thomas Killestein from the University of Turku, who was also involved in the research, added: “With modern surveys, we finally have data sets of the size and quality to tackle some of the key remaining questions in supernova science: exactly how they explode.” Machine learning approaches like this make it possible to study a larger number of supernovae, in greater detail and more consistently than previous approaches.”