MSU IceCube Neutrino Observatory Group

Associate Professor Tyce DeYoung, from the Department of Physics and Astronomy at Michigan State University, conducts research in particle astrophysics – the observation of high energy particles from space, with the twin goals of understanding how and where they are produced and better understanding the fundamental properties of the particles themselves. He plays a leading role in the IceCube Neutrino Observatory, a billion-ton detector of nearly massless subatomic particles called neutrinos located beneath the South Pole.

IceCube is designed to probe the most violent astrophysical sources in the Universe: events like exploding stars, gamma-ray bursts, and cataclysmic phenomena involving black holes and neutron stars. Neutrinos interact only weakly with other matter, which allows them to escape from these dense, energetic environments, carrying information about their sources with them. But this also poses a challenge for detection, and huge detectors are needed to observe neutrinos from such distant sources. Approximately 300 physicists from 44 institutions in 12 countries work on IceCube, and they also study other important questions in physics like the nature of dark matter and the properties of the neutrino itself. The U.S. National Science Foundation (NSF) provided the primary funding for the IceCube Neutrino Observatory, with assistance from partner funding agencies around the world.

The MSU IceCube research group includes several undergraduate and graduate students and postdoctoral research associates as well as Professor DeYoung and Assistant Professor Kendall Mahn, also from the Physics and Astronomy Department. With nearly half a million of these ghostly particles detected so far, IceCube provides the world's largest neutrino data set, and the MSU group focuses on using this data to measure fundamental properties. Neutrinos come in three types, and quantum mechanical effects cause them to "oscillate" between types as they travel through space. Precision measurements of these oscillations allow for determining important parameters of neutrinos which may shed light on the fundamental structure of matter. 

Professor DeYoung's group relies extensively on high performance computing resources such as iCER's High Performance Computing Center. IceCube observes approximately 100 billion particles per year, but only roughly one in a million is an elusive neutrino. Automated pattern recognition and machine learning techniques are used to reduce the data set to a manageable size, but that is only the first step.

Although billions of Cherenkov photons may be emitted in a high-energy neutrino event, only a few hundred are typically recorded by IceCube sensors. The details of the neutrino event – the type of neutrino that interacted, the direction from which it came, and its energy – must all be reconstructed from this handful of photons. This is done using a maximum-likelihood optimization, and the relative importance of individual detected photons makes them critically dependent on extremely precise modeling of photon production and propagation in the Antarctic ice. The ice is highly transparent, with photons traveling tens of meters between scattering events and hundreds of meters before being absorbed – but also highly complex, with the glaciological history of the Earth imprinted in the medium in the form of spatially variable strata of dust grains which act as scattering centers for the photons. As a result, no analytical description of photon transport is available, and DeYoung relies on computationally expensive numerical simulation of each photon. "General-Purpose Graphics Processing Unit accelerators such as those in the HPCC's newest clusters are perfectly suited to this problem," DeYoung explains. 

The MSU group has pioneered new methods of reconstructing neutrino events at the lowest energies accessible to IceCube, where the effects of neutrino oscillations are most visible. These efforts are paying off with measurements from IceCube, which are competitive with those of the world's leading oscillation experiments. Because IceCube measures these parameters using different techniques and in a higher energy range than other detectors, these data also provide tests of theories of possible new physics beyond particle physics' Standard Model.