SMuK 2023 – wissenschaftliches Programm

Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe

T: Fachverband Teilchenphysik

T 119: Neutrino Astronomy IV

T 119.1: Vortrag

Donnerstag, 23. März 2023, 15:50–16:05, POT/0112

Search for neutrinos from AGN using a machine-learning-based source selection — •Sebastian Schindler for the IceCube collaboration — ECAP, University Erlangen-Nürnberg, Germany

The IceCube Neutrino Observatory is currently the world’s largest high-energy neutrino detector. After the detection of a diffuse astrophysical neutrino flux in 2013, one of the main goals has been to associate parts of this flux with specific source classes. A few "hot spots" at or above the three-sigma level have been found and associated with certain classes of Active Galactic Nuclei (AGN). Most recently, the Seyfert II galaxy NGC 1068 was associated with a neutrino flux at a significance of 4.2 σ, and there is growing evidence for a neutrino flux from blazars. However, the underlying physical processes of neutrino production remain poorly understood. One problem for neutrino-source searches comes from the use of historically-driven class definitions of AGN, which are based on specific spectral properties that are not necessarily optimal for the selection of potential neutrino sources.

This talk will motivate a study that aims to address this problem in two stages. The first stage will use multi-wavelength data to define a source selection using modern machine-learning approaches in a way that emphasizes intrinsic physical properties and mostly disregards the general AGN classification. This will allow to identify potential neutrino sources similar in physical properties to those associated with the currently detected "hot spots". The second part will perform a statistical analysis in the form of a correlation analysis, for example a stacking search, using these previously defined source selections.