Talk WaveLoc : Locating earthquakes by migrating seismic waveforms

Abstract

Authors: Nadège Langet, Alessia Maggi

WaveLoc is an algorithm for location of earthquakes by brute-force migration of full-waveform seismological data. We pre-process the data using a kurtosis filter to enhance the first arriving seismic phases on the waveforms, and then migrate through a known wave-speed model. Points in time and space where the data from multiple stations stack constructively indicate the occurrence of earthquakes. The entire algorithm has been developed in Python, and relies strongly on numpy, scipy and the obspy suite, and demonstrates that 4-D heavy calculations can be efficiently conducted using pure Python programming. Graphical output is produced using matplotlib and mayavi. WaveLoc is currently being applied to volcano seismicity (e.g. Piton de la Fournaise volcano, Reunion island, France), as well as to tectonic seismicity (e.g. the recent earthquakes in Italy).

Keywords: Seismology, Location, Waveforms, 4D, Volcano

tagged by
no related entity