Optimising Seismic Array Analysis for Forensic Seismology

  • Neil Wilkins

Student thesis: Master's ThesisMaster of Science (MSc)


Seismological techniques have been at the forefront of international efforts to monitor the development and testing of nuclear weapons for more than 70 years. The need for a robust monitoring system has led to the development of seismic instrumentation and data analysis techniques, particularly in arrays of seismometers. Seismic data can be used both to detect clandestine nuclear explosions, and to distinguish them from naturally-occurring earthquakes using discriminants including source depth, regional P/S amplitude ratios and the ratio of body wave magnitude to surface wave magnitude (mb/MS).
Seismic arrays have two main functions useful in forensic studies: stacking the seismic traces from each station to improve the signal-to-noise ratio (SNR) of the data, and improving the quality of the signal used for waveform analysis studies. These stacked signals can be applied as vespagrams, where the direction and arrival time of a particular seismic phase can be identified. Stacked signals can also be used in magnitude calculations towards calculating the explosive yield and mb/MS for a suspected nuclear test.
In this thesis, the history of forensic seismology is discussed and different stacking techniques applied to seismic data from nuclear tests conducted in North Korea and the results compared for different seismic arrays. I have written a software package VesPy, to perform various useful array analysis functions in Python, which is described in this thesis, alongside example applications.
Finally, these techniques are combined to develop slowness-azimuth station corrections (SASCs) for seismic arrays of the International Monitoring System (IMS) for the region surround- ing the North Korean nuclear test site. These corrections are applied to data from five nuclear tests to demonstrate the effect on the signal and magnitude calculations.
I describe a set of recommendations for which stacking methods to use in different situations, and a best practice for developing future SASCs.
Date of Award28 Nov 2019
Original languageEnglish
Awarding Institution
  • The University of Bristol
SupervisorJames M Wookey (Supervisor) & Neil D. Selby (Supervisor)

Cite this