AI revolutionizes astrophysics: Glombitza cracks cosmic radiation puzzles!

AI revolutionizes astrophysics: Glombitza cracks cosmic radiation puzzles!

A breakthrough in cosmic ray research could be achieved thanks to the application of artificial intelligence (AI). Dr. Jonas Glombitza from the Erlangen Center for Astroparticle Physics (ECAP) has developed innovative techniques for data analysis. These show that the most energetic particles that reach Earth are predominantly heavy nuclei such as nitrogen or iron atoms, rather than protons. These findings are part of a comprehensive study carried out in collaboration with the world's largest observatory for studying cosmic rays, the Pierre Auger Observatory in Argentina.

Glombitza began programming machine learning tools for analyzing particle showers at RWTH Aachen University in 2017. Since moving to FAU in 2022, he has continued to work on refining these techniques. In 2025, he was awarded the ETI Award, a university talent promotion program, recognizing his significant contributions to astrophysics. However, he says that the term “artificial intelligence” is often defined differently and he is reluctant to use it.

The Pierre Auger Observatory

The Pierre Auger Observatory is the largest facility for research into cosmic rays with an area of ​​3,000 km². The facility includes 27 telescopes and 1,660 surface detectors housed in water tanks. The observations focus on the fluorescent light emission generated by the air showers created when primary particles interact with the Earth's atmosphere.

However, the telescopes only work under optimal conditions, such as clear, moonless weather, which limits data collection. Despite these challenges, AI-assisted analysis can reconstruct the distribution patterns of particles from the surface detectors and estimate the mass of the primary particles. This technology enables the evaluation of around 60,000 particle showers within a relatively short period of time - a task that would have required around 150 years of traditional telescopic observations without AI.

Challenges of cosmic rays

Cosmic rays, described as high-energy particle radiation from the Sun, the Milky Way and distant galaxies, consist primarily of protons, electrons and ionized atoms. Each primary particle creates a particle shower that can release up to 10^11 secondary particles. In the Earth's atmosphere, around 1,000 particles per square meter hit the outer layer every day.

A central mystery in physics concerns the origin of high-energy particles, which have energies above 10^20 eV. These particles occur extremely rarely, about one particle per square kilometer per century, and their study requires large detector facilities such as the Pierre Auger Observatory. Over 250 scientists from 17 countries are working on these challenges as part of the Pierre Auger Collaboration.

In addition to the fundamental research topics, the AugerPrime upgrade of the detector, which combines different types of particle detectors, including optical fluorescence telescopes and radio antennas, is also being investigated. The measured data could not only help solve the mysteries surrounding cosmic rays, but also study their effects on radiation exposure, particularly for people in air travel.

Overall, the current developments in the application of artificial intelligence in astroparticle physics illustrate the enormous possibilities that arise from the symbiosis between modern technology and classical physics. The extraordinary progress in this area could have far-reaching implications for our understanding of cosmic rays and their origins.