/ga-aem

Modelling and Inversion of Airborne Electromagnetic (AEM) Data in 1D

Primary LanguageC++OtherNOASSERTION

GA-AEM Source Code Repository

Geoscience Australia Airborne Electromagnetics Programs

  • Author: Ross C Brodie, Geoscience Australia (ross.c.brodie at ga.gov.au)
  • Language: mostly C++, some matlab, some python

Releases

Release-20160606

- Added Python 3.x interface for simple forward modelling and derivatives only.
- Added Matlab interface for simple forward modelling and derivatives only.
- Changed how the PPM normalisation is carried out. Now PPM normalisation is by directional-component-wise with respect to the maximum primary dB/dt or B-field at the receiver for a reference system geometry (which is usually estimated on a per flight or per survey basis). Previously PPM normalisation was with respect to the system geometry for the forward model being run.
- Added GEOTEM (1996 ppm system) and SPECTREM (ppm system) examples.
- Fixed a bug in the thickness derivative of the second bottom layer. This may have effected few-layer inversions, but not multi-layer fixed-thickness inversion.

Release-20160428

- Initial public release.

Currently included programs

  1. GAFORWARDMODELTDEM - 1D forward modelling program for AEM data
  2. GALEISBSTDEM - deterministic 1D sample by sample inversion of AEM data
  3. GARJMCMCTDEM - stochastic 1D sample by sample inversion of AEM data

Documentation

  • User manual (see ga-aem/docs/GA AEM Programs User Manual.pdf)
  • Theoretical details for GALEISBSTDEM (see ga-aem/docs/GALEISBSTDEM Inversion Algorithm Theoretical Details .pdf)

Building on Linux

  • cd makefiles
  • edit the file run_make.sh to setup for your compiler
    • set the C++ compiler (e.g., cxx=g++)
    • set the MPI C++ compiler (e.g., mpicxx=mpiCC)
    • set the C++ compiler flags (e.g. cxxflags='-std=c++11 -O3 -Wall')
    • set the executable directory (e.g., exedir='../bin/raijin/gnu')
  • run_make.sh
  • Matlab shared library should go into ga-aem/matlab/bin/linux (.dll on Windows or .so on linux)
  • Python shared library should go into ga-aem/python/gatdaem1d (.dll on Windows or .so on linux)

Building on Windows

  • You can build the programs with the free Microsoft Visual Studion 2013 Express.
  • Visual Studio solution and project files are supplied.
  • Open ga-aem\vs2013\ga-aem-all\ga-aem-all.sln to compile all programs plus the matlab and python shared libraries.
  • Alternatively open individual program solutions files in their respective directories.

Third party software dependencies

  1. Template Numerical Toolkit (TNT)

  2. The Fastest Fourier Transform in the West (FFTW)

    • see http://www.fftw.org
    • Windows binaries (dlls) are include in the repository.
    • Linux users will need to install a suitable FFTW for your system.

  3. Message Passing Interface (MPI)

    • see https://www.mpi-forum.org
    • Both Windows and Linux users will need to install a suitable MPI for your system.
    • The pre-compiled Windows inversion programs require Microsoft HPC Pack 2012 because thats the flavour of MPI they have been compiled and linked with.
      • They can be recompiled using MPICH or other flouvours of MPI if required.
      • If you do not want to install MPI on your Windows system, you can use the galeisbstdem-nompi.exe.
      • This program has not been linked with MPI but uses OpenMP shared memory parallelism instead (see manual for details).