Data Management Center

Institute of Earth Sciences, Academia Sinica

Correcting the background tilt signal of the horizontal seismometer by using a rotation sensor

1. Institute of Earth Sciences, Academia Sinica, Taipei, Taiwan
citation:
Lin, C.-J.*, C.-S. Ku, T.-C. Chi, B.-S. Huang, H.-H. Huang, C.-C. Liu (2022) Seismo. Res. Lett.
DOI: https://doi.org/10.1785/0220210185

Abstract

In observational seismology, the long-period background noise level in horizontal seismometer components is always higher than the vertical component at the same location. It is believed that the horizontal long-period noise is mostly attributed to ground tilt, yet the authentic tilt recordings of background noise in the seismic frequency band (0.01-20 Hz) have not been recorded successfully. We demonstrate the use of a newly developed optical-based rotation sensor, blueSeis-3A, to correct for the tilt effect in a horizontal seismometer. A station with a horizontal long-period (1 < T < 100 s) background noise level higher than the global new high-noise model (NHNM) was chosen, where a seismometer, a tiltmeter, and a rotation sensor were collocated. We first confirmed that the horizontal recordings from a tiltmeter and a seismometer were visually consistent, although it was unclear if tilt contamination existed in both recordings. It became evident until the ground tilt recorded by a rotation sensor was found to be waveform-matched with them, indicating that both sensors were contaminated by tilt. The spectrum of tilt records showed that ground tilts were mostly at period bands > 1 s. By subtracting tilt from seismometer records, the horizontal background noise levels could be improved, approximately to the resolution level of the rotation sensors. Our results demonstrate that collocating rotation sensors with seismometers is a feasible way to improve the quality of the long-period background noise level in seismometer horizontal components when rotation sensors are sufficiently sensitive.

Downloads

CHS2_tilt_correction_data [ZIP] (396MB)