You are using an old browser with security vulnerabilities and can not use the features of this website.
Timescale: Oct.2018 – Sept.2021
Prof. Dr. Fabrice Cotton, GFZ Potsdam
Prof. Dr. Frank Scherbaum, University of Potsdam
Dr. Matthias Ohrnberger, University of Potsdam
Recent research results suggest that catastrophic events (e.g. earthquakes, landslides) are preceded by a preparation phase of several years. Small earthquakes, tremors, rock velocity changes occurring during this phase may provide key information about the time evolution of state active faults and landslides properties.
In order to detect these time-changes we need to detect, record and analyze these tremors, micro-earthquakes and rock velocity changes. The goal of the PhD topic is thus to exploit the similarity of seismic and acoustic waveforms for information retrieval and to develop innovative seismological data processing methods for seismological monitoring purposes by borrowing from the expertise developed in the field of music information retrieval and audio signal processing.
Objectives and Methods
Nowadays most of seismological stations are continuously recording ground-motions. This continuous recording offers a fantastic opportunity to record signals associated to these “weak” events (small earthquakes, tremors, landslides…). However the amount of data is now too large to use visual quality control or expert classification. There is then an urgent need to develop new strategies to identify, extract and validate the records of earthquakes. There is also a need to develop reduction size strategies while still retaining a comparable level of scientific information. There is finally the need to classify the extracted records according to their origins.
Very similar issues (data reduction, extraction and classification) are also faced in the field of audio signal processing. In the context of music information retrieval (MIR), numerous strategies and methods have already been developed which are widely unknown within the seismological community but potentially very useful for seismological data processing.