met-jobs@lists.reading.ac.uk
July 2013
Message 101

[Periods|Index by:DateThreadSubjectAuthor|Date:PreviousNext|Thread:(Previous)(Next)|List Information]

[Met-jobs] PhD position: Joint Inversion of Multiple Data Types for Rupture Models of Large Earthquakes

From Sebastien Allgeyer <sebastien.allgeyer@gmail.com>
To met-jobs@lists.reading.ac.uk
Date Wed, 31 Jul 2013 09:38:42 +1000

Summary
The candidate will develop techniques for the inversion of multi sensor data for earthquake fault rupture. The different types of data will include near-field and far-field seismic waveforms and continuous GPS measurements, as well as aftershock and sea level data. The methodology developed should account for non-linearity in the dependence of data on fault geometry, as well as determining the level of ambiguity resolution in the two candidate fault planes.



Research Project
The dynamics of fault rupture during large earthquake is still poorly understood, and many fundamental questions remain unanswered; Why does rupture initiate in one part of a fault rather than another, why does rupture sometimes stop after generation of a small earthquake, and why does it sometimes grow to the size of a mega-earthquake? Kinematic rupture models, which describe how fault rupture evolves in space and time, are the fundamental data used to answer questions such as these. Such models are typically obtained through inversion of seismic, geodetic or tsunami data. Each type of data has its strengths and weaknesses. Ideally, we would perform joint inversions of multiple data types in a way that would maximize their strengths, thereby obtaining models with the optimal spatial and temporal resolution.

Furthermore, kinematic rupture inversion methods generally assume knowledge of the geometry of the rupture surface, but point-source estimates of fault mechanism cannot distinguish between the actual rupture plane and its conjugate, auxiliary plane, nor can they provide information on extent of the rupture area. This problem is particularly acute when rapid estimates of rupture are required, as is the case when a tsunami warning or rapid impact estimate is required. Information on fault plane geometry and rupture extent generally requires near-field and/or aftershock data system, but even when theese are availbale some level of ambiguity often remains.

This project seeks overcome some of these challenges by developing robust inversion methods that can be applied to multiple data types, and account for variable spatial and temporal resolution as well as ambiguity in fault geometry. We propose to investigate these questions using a trans-dimensional Bayesian approach, recently developed within our group, which allows the model parametrisation to adapt to the data as part of the inversion process.   

This position is supported by an Australian Research Council Discovery Project to undertake research in new methods for studying fault rupture using multi-sensor data. It is open to all candidates with a strong background in physics, geophysics, or applied mathematics. Of particular interest are candidates with expertise in earthquake source characterization.

The Australian National University's Research school of Earth Sciences
The Earth Physics Area of the Research School of Earth Sciences (RSES) combines world-leading expertise in earthquake, tsunami and data inference science. RSES carries out a diverse range of research into the physics, chemistry, material properties and environmental conditions of the Earth, and hosts the Centre for Advanced Data Inference (CADI). CADI has its own Terrawulf III computing cluster, and in addition the ANU also hosts the national facility for computing infrastructure on campus. RSES also has close interaction with Geoscience Australia, where the seismological component of the Joint Australian Tsunami Warning System is located.


For additional information
Please send a cover letter and CV to Dr. Sébastien Allgeyer (sebastien.allgeyer@anu.edu.au)


Go to: Periods · List Information · Index by: Date (or Reverse Date), Thread, Subject or Author.