Reconstruction of the properties of ultra-high energy neutrinos from radio detector data using deep learning

Detection of neutrinos at ultra-high (UHE) energies (>1017eV) would be one of the most important discoveries in astroparticle physics in the 21st century.

Radio detection remains the most promising technique for instrumenting a volume large enough to study UHE neutrinos, and the first detector of sufficient size (RNO-G) is currently being constructed in Greenland. However, the required analysis techniques to make use of the data are not yet developed.

This will be addressed through this project.

The methods to determine the direction, energy and flavor of UHE neutrinos will be developed by employing state-of-the-art deep-learning techniques. The developed algorithms will allow real-time processing of the data to obtain the neutrino direction and energy.

In particular, the fast response time is crucial for multi-messenger campaigns, allowing the radio detector to alert other telescopes. This will provide a unique opportunity to identify and understand the sources of UHE neutrinos.

The starting grant would allow me to build a team at Uppsala University, making Sweden a leading player in this exiting international effort with significant contributions to the success of RNO-G and substantial contributions to the future next-generation neutrino observatory IceCube-Gen2 envisioned to include an order of magnitude larger radio array.