Erlangen 2026 – wissenschaftliches Programm

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

T: Fachverband Teilchenphysik

T 39: Gravitational Waves II

T 39.5: Vortrag

Dienstag, 17. März 2026, 17:15–17:30, KS 00.005

Newtonian Noise mitigation for the Einstein Telescope using Deep Learning — •Jonathan Kuckert, Jan Kelleter, Patrick Schillings, and Johannes Erdmann — III. Physikalisches Institut A, RWTH Aachen University

In the past, gravitational wave interferometers were not able to measure low frequency gravitational waves. The Einstein-Telescope (ET) is a proposed third-generation underground gravitational wave interferometer. For the first time, ET will enable measurements in the 1-10 Hz region. In this region, Newtonian Noise (NN), perturbations in the gravitational field due to density fluctuations in the underground, is the predicted dominant noise source. As a gravitational phenomenon NN, cannot be shielded. Therefore, the most promising mitigation strategy is based on seismometer arrays. The seismometers surround the interferometer mirrors and the gravitational noise on the mirrors is predicted using the seismometer data. For this kind of prediction, Wiener Filters (WFs) were deployed as a standard solution in the past. Based on simulations of simplified seismic events, it has been shown that Deep Learning methods, specifically Graph Neural Networks (GNNs) can match and outperform WFs. In this talk we present further improvements in mitigation and first steps towards optimising seismometer positions using Machine Learning.

Keywords: Deep Learning; Newtonian Noise; Noise Mitigation; Graph Neural Networks; Machine Learning