Identifying multiplets of IceCube alert events

¹ Technische Universität München, TUM School of Natural Sciences, Physics Department, James-Frank-Str. 1, D-85748 Garching bei München, Germany
² European Southern Observatory, Karl-Schwarzschild-Straße 2, 85748 Garching bei München, Germany
³ Institute for Advanced Study, Technische Universität München, Lichtenbergstrasse 2a, D-85748 Garching bei München, Germany
⁴ Center for Astrophysics and Space Science (CASS), New York University Abu Dhabi, PO Box 129188 Abu Dhabi, United Arab Emirates
⁵ Associated to INAF, Osservatorio Astronomico di Brera, Via Brera, 28, I-20121 Milano, Italy

^⋆ Corresponding author: martina.karl@tum.de

Received: 6 December 2024
Accepted: 5 March 2025

Abstract

Context. The IceCube Neutrino Observatory publishes “alert events”, which are detections of high-energy neutrinos with a moderate-to-high probability of being of astrophysical origin. While some events are produced in the atmosphere, a fraction of alert events should point back to their astrophysical sources.

Aims. We aim to identify multiple alert events possibly related to a single astrophysical counterpart by searching for spatial and temporal clusterings in 13 years of alert data.

Methods. We identify spatial clusters (“multiplets”) by checking for events overlapping within their uncertainty regions. In order to reduce chance coincidences and to improve the signal purity of our sample, we apply different thresholds. We investigate the weighted mean position of these multiplets for an over-fluctuation of γ-ray counterparts. As a final step, we apply expectation maximization to search for temporal clusters around the identified weighted mean positions.

Results. We find no statistically significant clustering of alert events around a specific origin direction or in time.

Conclusions. This could be because the selections are still dominated by atmospheric background. Another possibility is that we are not yet sensitive enough and only detect single events from sources. In this case, we need more data in order to observe a clustering of events around their origin.