GPGN 605 Geophysical Inverse Theory
Course:  GPGN 605 Inverse Theory
Instructor:  John Scales (jscales@mines.edu, 273-3408)
Wednesday 6:00 - 9:00 PM (i.e., 18:00 - 21:00)
Green Center (room 295)
Fall 1998

What is an Inverse Problem?








News:  8 November 2000

Here is the latest snapshot of the book:  main.pdf

Here is a link to Per Christian Hansen's Matlab regularization package

Link to Kasper Van Wijk's page containing the Matlab code for the VRP problem.
 
 

News:  15 August 2000

This is the latest version of the inversion book available on-line. However, I am busy revising it and will have a new version available by late September or early October. Among the new features will be

Chapters 1 - 12:  1.5 Mb gzipped postscript.  10/20/98.  This include lots of recent bug fixes and has the table of contents and bibliography.  Note that this file uncompresses to about 6 Mb of postscript.  On some Unix systems there is a default limit to the size of files you can print.  This can usually be fixed by giving lpr a command line argument, such as -s.  If you run into this problem read the man page for lpr. Of course, you can always use ghostview or gv to print only selected pages. I also fixed the labeling problem with the lemmas in chapter 10.
 
 


Software Resources (emacs, xlisp-stat, etc)
Some toy software

Some serious software


 
 

Additional background material, lecture notes, software and links

Three recent papers



A new approach to incorporating geologic and geophysical inforamtion into inverse calculations (Geophysical Joural International, 141, 713-723, 2000).   Click here for the postscript.
 

A case study of Bayesian inversion applied to full-waveform seismic data (Journal of Geophysical Research, 103 2759-2779, 1998).   Treats the practical problems of how  compute a priori distributions and data uncertainties for field data.Click here for the postscript.

Comparison of Bayesian methods with Tikhonov regularization.   (Inverse Problems. 13, 323-349, 1997.)  Computes resolution and bias in these two different schemes.  Regularization methods bear a superficial similarity to Bayesian formulae, but the goals and conclusions are completely different.   Click here for the postscript.  Both of these papers were co-written with Wences Gouveia.



 
 

Annotated Reference List (books, papers, etc.)

Residual statics example.  Start with the Readme file.
Here are the data files for the example in Chapter 11.