Optimization of Radio Detectors of Ultra-High-Energy Neutrinos through Deep Learning and Differential Programming

Projektbeschreibung

Ultrahochenergetische Neutrinos für gesteigerte Nachweisraten

Der Nachweis ultrahochenergetischer Neutrinos (UHE, E > 10^17 eV) stellt einen großen Durchbruch in der Astroteilchenphysik dar, wobei der Radiofrequenznachweis das einzige verfügbare Verfahren ist. Aufgrund ihres geringen Flusses sind jedoch auch mit großen Anlagen wie dem geplanten IceCube-Gen2-Neutrinoobservatorium am Südpol nur wenige Ereignisse pro Jahr nachweisbar. Das Team des ERC-finanzierten Projekts NuRadioOpt zielt darauf ab, UHE-Neutrinodetektoren zu optimieren, wobei die Nachweisrate verdoppelt und die Ereignisqualität durch Deep Learning und differenzielle Programmierung verbessert werden. Damit werden traditionelle schwellenwertbasierte Auslöser durch neuronale Netzwerke ersetzt und somit höhere Erkennungsraten ohne erhebliche Hardwarekosten möglich. Das Projektteam wird die Bestimmung der Neutrinorichtung und -energie optimieren, was sich perfekt mit den Zielen der geplanten IceCube-Gen2-Anlage deckt.

Ziel

Detection of neutrinos at ultra-high energies (UHE, E >10^17eV) would be one of the most important breakthroughs in astroparticle physics in the 21st century and would open a new window to the most violent phenomena in our universe. Radio detection remains the only viable technique at these energies.
However, owing to the expected small flux of UHE neutrinos, the detection rate will be small, with just a handful of events per year, even for large future facilities like the IceCube-Gen2 neutrino observatory at the South Pole.
In this project, I will enhance the science capabilities of UHE neutrino detectors substantially by increasing the detection rate of neutrinos and improving the quality of each detected event, using recent advances in deep learning and differential programming. I will replace the threshold-based trigger foreseen for future detectors with neural networks, increasing the detection rate of UHE neutrinos by a factor of two at negligible additional hardware costs. I will perform an end-to-end optimization using differential programming and deep learning to improve the determination of the neutrino direction and energy.
My previous work on developing state-of-the-art MC simulation codes, my experience in data analysis, designing reconstruction algorithms and deep learning, and my leadership role in IceCube-Gen2 will enable this ERC project.
The timing of this project is perfect for influencing IceCube-Gen2 - the largest facility for astroparticle physics with neutrinos for the next decade - whose construction is planned to start in 2027. With this ERC project, IceCube-Gen2 will be able to expedite the discovery of UHE neutrino fluxes by up to a factor of five, see sources from deeper in our Universe increasing the observable volume by a factor of three, and measure the neutrino-nucleon cross-section at EeV energies with 3x smaller uncertainty. Hence, NuRadioOpt will substantially increase the capabilities of future observatories for UHE neutrinos.