Special Session: Interaction Between Observations and Models in Seismo-volcanic Studies
Type: Oral
Day: 5/15/2018
Time: 10:45 AM
Room: Flagler
Abstract
We developed a network-based method for detecting and locating seismo-volcanic signals such as volcano-tectonic earthquakes, long-period events, and volcanic tremors. The method is based on the analysis of eigenvalues and eigenvectors of the seismic network covariance matrix, the equivalent in the frequency domain of the cross-correlation matrix. First, the width of the network covariance matrix eigenvalues distribution is used to detect different types of seismo-volcanic events. Then, the first eigenvector of the covariance matrix corresponding to each event is extracted. Our main hypothesis is that, by representing the principal component of the recorded wavefield, this first eigenvector characterizes the dominant event filtering the information related to the noise. Sources of the different events are finally located using the cross-correlation moveout information of the corresponding first eigenvector.
The method is applied to two data sets to automatically detect and locate volcano-tectonic earthquakes, long-period events and volcanic tremors. We analyze seismic data continuously recorded by the Red Sísmica Canaria, a permanent monitoring network composed of 15 broadband stations operated by the Instituto Volcanológico de Canarias (INVOLCAN) on the island of Tenerife (Canary Islands). Results are also obtained from four and a half years of continuous data recorded by 19 permanent seismic stations operated by the Kamchatka Branch of the Geophysical Survey (KBGS) in the vicinity of the Klyuchevskoy volcanic group in Kamchatka (Russia). The developed method does not require a priori knowledge, is fully automatic, and the database of network-based events fingerprints can be continuously enriched with newly available data.
Author(s):
Soubestre J. Instituto Volcanológico de Canarias (INVOLCAN)
Barrancos J. Instituto Volcanológico de Canarias (INVOLCAN)
D'Auria L. Instituto Volcanológico de Canarias (INVOLCAN)
Padilla G. D. Instituto Volcanológico de Canarias (INVOLCAN)
Shapiro N. M. Institut de Physique du Globe de Paris (IPGP)
Seydoux L. Université Grenoble Alpes, Institut des Sciences de la Terre (ISTerre)
Peréz N. Instituto Volcanológico de Canarias (INVOLCAN)
Automatic Detection and Location of Seismo-Volcanic Signals Based on the Seismic Network Covariance Matrix: Examples from Tenerife (Canary Islands) and the Klyuchevskoy Volcanic Group (Kamchatka)
Category
Interaction Between Observations and Models in Seismo-volcanic Studies