Speaker
Description
Microseisms are continuous ground vibrations excited by ocean-wave activity, typically around 0.1 Hz with amplitudes of the order of micrometers. Although they lie below the observation band of ground-based gravitational-wave detectors, they can still have important impacts on detector operation and site characterization. In LIGO and Virgo, microseisms can couple to the detector through up-conversion associated with scattered-light noise. In KAGRA, winter microseisms driven by rough conditions in the Sea of Japan strongly affected the actuator range of the vibration isolation system and were one of the major causes of interferometer lock loss in the early stage of commissioning.
We are developing microseism forecasting methods for KAGRA, Virgo, and the ET Sardinia candidate site using offshore wave and weather information from Windfinder together with local seismic observations. Our approach includes a PCA-based model function as well as machine-learning methods such as deep neural networks and LightGBM. In this contribution, we present the current status of these studies and discuss the similarities and differences among the three sites. We also compare the different approaches in terms of forecasting capability and physical interpretability, and discuss their relevance for detector characterization, operation planning, and environmental-noise studies for future detectors including ET.