Activities of the Area

The core activity of the Imaging and Numerical Geophysics area is the study and development of numerical solution methods and parallel algorithms to reconstruct the map of the subsurface properties. We thus continue the work initiated within the former Geophysics Area and Parallel Computing Group on developing industrial high performance codes for geophysics exploration.

Forward and inverse problems for prospecting for oil, gas, mineral and groundwater resources provide eloquent demands for parallel computing. A good example is the echo-reconstruction of subsurface structure based on non-intrusive experiments. The inversion process requires a propagation model to help determine inaccessible field values from measured data.

Seismic data migration, presently our most important activity in collaboration with ENI/AGIP, is an image reconstruction technique based on the inversion of the scalar wave equation. In the migration process, the recorded pressure waves are used as initial conditions for a wave field governed by the scalar wave equation, which propagates downward and in reverse time. 3D seismic imaging, a computationally- and I/O-intensive task, is reportedly the key technological advance in oil exploration today. 

Another application of the inverse problem is the reconstruction of electromagnetic properties of the medium through the comparison between acquired data and numerical simulations. We developed an application to invert mono- and multi-frequency data in the lower part of the frequency spectrum.

The resulti ng inversion scheme apply, for instance, to geophysical electromagnetic surveys reconstructing the conductivity profile of the near subsurface from the measured apparent conductivity. 

In collaboration with the ESAOTE Group, we imported wave-field back propagation algorithms into ultrasound reconstruction, potentially significantly increasing frame rate in 2D data imaging and opening the way to actual 3D reconstruction

More recently, the area diversified its domain of application extending inverse problem techniques to environmental studies, in particular for the characterization and monitoring of contaminated sites, including evaluation of cost/benefit of different remediation techniques. 

Most of our codes were developed in High Performance Fortran (HPF), a straightforward extension of Fortran 90 for data parallel programming. Since MPI has become the first widely accepted portable parallel programming language, we began the porting of our SPMD codes to this programming model. Our new large scale seismic applications, working on irregular data structures, were originally designed in F90/MPI.

Area presentation (in italian)

The Area old home page

For further information go:

Staff

Enrico Pieroni (Area leader)

Giovanni Cardone

Antonio Maria Cristini

Cristina Manzi

  [Home ] 

Aggiornato il: 16 maggio 2002