Submission #309

Submission information
Submitted by Anonyme (not verified)
Sat, 04/28/2018 - 13:56
86.201.23.152
No
Industry
Applied Mathematician or Research Engineer in numerical modelling and imaging processing
Computer Science
PhD Thesis
CNRS
http://www.get.obs-mip.fr
roland.martin@get.omp.eu
Toulouse
France
Title: DIMSCALE3D: basin and subsurface water mass detection and imaging via high performance computing and joint inversion of seismic and potential data

Funding: Doctoral grant of Region Occitanie and Federal University of Toulouse Midi-Pyrénées; gross pay: 1768 euros; net pay: 1414 euros.
Thesis supervisors and contacts:
Dr. Ing. Didier El Baz HDR, LAAS-CNRS, 05 61 33 63 03 elbaz@laas.fr
Dr. Roland Martin HDR, GET, 05 61 33 25 64 roland.martin@get.omp.eu

Project summary:
Due to current environmental and climatic issues, discovering, producing and managing new resources like underground water are major challenges. Our project aims at developing new tools for prospecting, monitoring and management of water and fluid contents at regional scale. The design and application of powerful imaging tools will be done at different spatial and temporal scales (basins and near-surface) with multiple data, e.g., seismic, electrical and gravimetric data. This project involves together geophysicists of the Geosciences Environment Toulouse Laboratory (GET) and computer scientists and applied mathematicians of the Laboratory of Analysis and Systems Architecture (LAAS-CNRS). In order to estimate water contents at sedimentary and near-surface scales, we propose to improve, optimize and use imaging tools to invert simultaneously various data at different scales. This is a major challenge. The project falls within two priority research topics of the GET laboratory : 1)Geo-resources, Geophysics and Geodesy, and 2) the priority axis of the University Paul Sabatier in Earth Sciences and High Performance Computing as well as Axis 1 of the Regional Competitiveness Pole AQUAVALLEY: Identification and assessment of water resources (identification and characterization of surface and deep water resources - quality and quantity of the resource - anticipation of the evolution of the resource).

Given the large volume of data to be processed, the optimization of parallel code on the regional platform CALMIP or some local platforms like NUWA at OMP will make it possible to accurately estimate water contents. The calculations will be carried out with data collected at the controlled laboratory scale and in two pilot sites, i.e., the Saint-Lary Soulan basin (kilometric scale data) and a medium scale site on the Mediterranean coast using seismic data and the BGI gravimetric database. The applications to the natural environment will be conducted under the supervision of experts in data acquisition and processing. In the medium term, we will be able to carry out systematic studies in the Pyrenees and monitor the transfer of water content masses at local and regional scales.

Goals of the Ph.D. work:
The Ph.D. student will focus his efforts on:
1) The improvement, adaptation and optimization of numerical simulation (forward problem using high order numerical schemes) and High Performance Computing (HPC) data inversion codes for the modeling and coupling of different physics during inversions .
2) The application and validation of digital acoustic data inversion codes in a controlled laboratory environment at a frequency range of a few kHz.
3) Taking into account of petrophysical and geological data a priori during the inversion process at the scales of the subsurface and basins.

Required profile: Master's degree in Earth and Planetary Science, applied mathematics or civil engineering, preferably with an engineering degree( INSA, INP, ENSICA, ISAE, .. .engineering schools)
Scientific curiosity, a taste for geophysical exploration, numerical simulation and parallel computing; faculties of teamwork. Strong background in finite elements, fiinite differences anf finite volumes.

Skills: good to excellent level in Fortran90/2008 and/or C / C ++ code development on Linux; knowledge of MPI, OpenMP, CUDA, OpenACC or OpenCL; numerical simulation, parallel algorithms; experience appreciated in computing accelerators: GPU, Intel Xeon Phi or cluster computing.

Professional perspectives of the PhD student: know-how in the fields of numerical computation and high-performance computing (HPC); expertise in geophysics and civil engineering for private companies as well as the academic career at regional or international level.
1414 Eurros net pay; 1768 Euros gross pay
Euro
Tue, 06/26/2018