Signals in the Noise – Resolving Small Seismic Signals at Very Long Periods
Session: Regional Earthquake Centers: Highlights and Challenges [Poster]
Type: Poster
Date: 4/28/2020
Time: 08:00 AM
Room: Ballroom
Description:
Increasingly data mining methods are being applied to seismic data to search for “phenomena” which may or may not be associated with typical seismic signals visible on seismograms. We explore whether seismic signals, particularly at periods longer than 50s, can be discovered in the noise recorded by a seismometer/data logger system using standard methods, like spectral analysis and stacking. First, we investigate how the intrinsic instrumental noise at very long periods affects the ability to resolve signals. Modern wide- and broadband seismometers with data loggers of 24 bits resolution or more typically resolve the earth noise as represented by the low noise model at periods between 0.1s and 20s, acceptable for detecting and locating local, regional and teleseismic events. For longer periods we looked for normal modes in the range of 182-667s in the data from the vertical channels of the BDSN recorded in the 120 hours following the M7.1 Indonesia earthquake on 14Nov2019. They are visible in the spectra of many of the stations. Data from the same quake are available from a huddle test in the Byerly Vault from the following data logger-sensor combinations: Q330HR&STS-1, Q330&STS-2, Q330S+&MBB2, Minimus+&CMG-3T, Centaur&Horizon, Centaur&TrilliumCompact120, Centaur&TrilliumPH120. In spectral plots of the vertical channels before the earthquake’s waves arrive, the station’s standard equipment, the Q330HR&STS-1, shows the lowest noise in the band from 20s–400s. The combinations with CMG-3T and TrilliumPH120 also exhibit low noise, while the noise of the other sensors has a minimum at 20s-30s and increases towards longer periods. To further investigate the ability to resolve tiny signals in the band from 20s-100s, we select data from earthquakes in selected magnitude ranges from the very productive Puerto Rico sequence, a “point source” from the California perspective, and explore the resolution of surface wave trains in the stacks of data from representative stations of the BDSN and NCSN. The magnitude ranges are M6.4 (mainshock), M5.6-5.9, M4.9-5.2, M4.4-4.6 and M3.9-4.2
Presenting Author: Margaret Hellweg
Authors
Margaret Hellweg peggy@seismo.berkeley.edu University of California, Berkeley, Berkeley, California, United States Presenting Author
Corresponding Author
|
Claire Doody claired@berkeley.edu University of California, Berkeley, Berkeley, California, United States |
Horst Rademacher horst@berkeley.edu University of California, Berkeley, Berkeley, California, United States |
Taka'aki Taira taira@berkeley.edu University of California, Berkeley, Berkeley, California, United States |
Robert Uhrhammer bob@seismo.berkeley.edu University of California, Berkeley, Berkeley, California, United States |
Signals in the Noise – Resolving Small Seismic Signals at Very Long Periods
Category
Regional Earthquake Centers: Highlights and Challenges