*******************************************************************************************
PhD opportunity- Annecy France- radar interferometry- volcanoes
From: Virginie Pinel <Virginie.Pinel@xxxxxxxxxxxxxx>
*******************************************************************************************
The LISTIC (Laboratoire d'Informatique, Système, Traitement de
l'Information et de la Connaissance, http://www.listic.univ-savoie.fr)
at the University of Savoie (France) is looking for candidates for a
PhD thesis, starting in October 2010.
Title: Information Fusion in the context of data assimilation in Geosciences
Keywords: image processing, radar interferometry (InSAR), information
fusion, data assimilation
Subject: The ANR EFIDIR project (Information mining and data fusion
for measuring displacement by radar imagery, http://www.efidir.fr)
brings together researchers in the fields of Earth science and
information processing (signal, image, computer ...). The observed
geophysical phenomena (earthquakes, volcanoes, glaciers) are the
source of surface displacements measured by GPS and radar imagery.
These displacement measurements are used to model the studied
phenomena and find the unknown parameters by a step called inversion
or, more generally, data assimilation into dynamic models. The combination of
different data sets or types of measures aims to validate the physical
models used, and where appropriate, to improve their accuracy.
Strategies for information fusion are often conditioned by the
upstream stages, ie the alignment of available information sources and
the extraction of attributes that are then combined according to their
properties (statistics, confidence ...). Data assimilation in
geosciences offers an alternative framework where the model of the
studied phenomenon connects the various inputs (attributes,
measurements) and the sought-after outputs: values, interpolation or
prediction of physical parameters. The assimilation of image data
requires an important reflection on the space where the comparison is
made between data and model outputs, the metric and minimization
methods used and on the representation and propagation of
uncertainties. This thesis will make first a synthesis of methods
currently implemented in
assimilation of images. It will develop a methodology for information
fusion adapted to the problem of displacement field assimilation in
dynamic models. The experimental part will be carried out on
geophysical models and remote sensing data to test the proposed
solutions in a complete information processing chain. In particular,
time series of SAR data acquired on volcanoes in Alaska (Okmok and
Westdahl) will be treated using a dynamic model of shallow magma
chamber replenishment.
Work will be performed at LISTIC (Laboratoire d'Informatique, Système,
Traitement de l'Information et de la Connaissance,
http://www.listic.univ-savoie.fr ) in collaboration with the LGIT
(Laboratoire de Géophysique Interne et de Tectonophysique,
http://www-lgit.obs.ujf-grenoble.fr).
Candidate Profile: Engineering / Master degree in one of the following
specialties: Signal /image processing, Applied Mathematics,
Geophysics. The candidate must have a training experience in image
processing or remote sensing as well as good programming skills.
Applicants must send to the contacts listed below their detailed CV, a
covering letter and their grades of Master or Engineering school, as soon as possible.
Contacts:
- Emmanuel Trouvé, Professor at University of Savoie, LISTIC, Polytech
Annecy-Chambery; emmanuel.trouve@xxxxxxxxxxxxxx
- Virginie Pinel, Researcher at IRD, LGIT Chambery;
virginie.pinel@xxxxxxxxxxxxxx
============================================================== To unsubscribe from the volcano list, send the message: signoff volcano to: listserv@xxxxxxx, or write to: volcano-request@xxxxxxxx
To contribute to the volcano list, send your message to: volcano@xxxxxxxx Please do not send attachments. ==============================================================
