Seismic Signal Analyis From The Perspective Of Music Information Retrieval
The generation of sound and vibration within the earth is in some respects very similar to the generation of sound in musical instruments (e. g. Johnson and Watson, 2019) or, if one considers the generation of volcanic tremor signals, the human voice. These similarities can be used for example for modeling (e.g. Schlindwein et al, 1995) or for extracting information in general. Research in the field of Music Information Retrieval (MIR) has resulted in paradigms which are largely unknown in the seismology community but which carry a lot of potential to improve current analysis practice in seismology. This refers to techniques for fingerprinting, source separation, onset detection, just to name a few.
Within the framework of the graduate school Natural hazards and risks in a changing world (NatRiskChange) funded by the German Research Foundation (DFG) and supported by the German Academic Exchange Service (DAAD), we are investigating in two research projects to what extent techniques and approaches from the field of Music Information Retrieval (MIR) can be used to investigate seismic and volcanic signals in order to improve hazard assessments or for improved monitoring.
Research topics:
Reza Dokht Dolatabadi Esfahani: Time dependent monitoring of active faults and landslides properties: developing new data processing approaches based on music information retrieval (MIR) strategies
Zahra Zali: Volcanic tremor analysis based on advanced signal processing concepts including music information retrieval (MIR) strategies
Reference:
Johnson, J. B., and L. M. Watson (2019), Monitoring volcanic craters with infrasound “music”, Eos, 100,https://doi.org/10.1029/2019EO123979. Published on 17 June 2019
Schlindwein, V., Wassermann, J., & Scherbaum, F. (1995). Spectral analysis of harmonic tremor signals at Mt. Semeru Volcano, Indonesia. Geophysical Research Letters, 22(13). doi.org/10.1029/95GL01433. (PDF)