Flexible Analysis of Earthquake Datasets Using a Modular, High-Throughput Seismic Processing System

Session: Applications and Technologies in Large-Scale Seismic Analysis
Type: Oral
Date: 4/28/2020
Time: 11:45 AM
Room: 120 + 130
Description: 

Analysis of large earthquake and controlled-source seismic datasets is often complex and requires combining different types of data from multiple sources, standard and customized types of data filtering and imaging with interpretation and elaborate analysis. It is also important to facilitate stream, batch and interactive processing by using multiple software packages, file formats nd programming languages in a cross-platform manner. With all generality of these objectives, seismic records are the primary data type in seismology, and therefore it is convenient and effective to build data analysis on the backbone of a high-throughput, modular seismic processing system, as it is common in the reflection seismic industry. We implement this design by using a Unix-based, open-source, reflection/refraction-seismology style processing package IGeoS (http://seisweb.usask.ca/igeos) combined with Octave (or Matlab), Generic Mapping Tools (GMT) and other software tools. The IGeoS system natively supports multicomponent, variable-length records with unlimited headers (metadata) and its tools provide essentially unlimited options for input/output, flexible data sorting, access to databases (including SQL and custom), several types of displays and numerous tools for single- and multichannel signal filtering and imaging. IGeoS tools calculate the IASP91 travel times, manage georeferenced metadata and correct for instrument responses. By integrating Matlab/Octave scripts, arbitrary interactive and non-interactive data analysis is easily implemented without knowledge of the details of the core system. This data-handling method is demonstrated on a large dataset provided by the Iranian Seismological Center and containing over 66,000 miniSEED files from 83 seismic stations in Iran. We illustrate several key data analysis tasks including quality control, editing, manual and automatic travel-time picking, filtering, plotting common-station and event record sections, event relocations, measurement of arrival and coda amplitudes and spectra and inversions for Lg and coda Q.

Presenting Author: Igor Morozov

Authors