Abstract
Computer programs to process and interpret resistivity sounding data are abundant these days, but only few of them are flexible, userfriendly, and commercially available. The program RESIST has been designed to process sounding data according to these criteria of flexibility and userfriendliness.
Designed to process data obtained with Wenner, Schlumberger, and dipole-dipole arrays, the program is built around three main procedures: smoothing of noisy field data, accurate computation of apparent resistivity models and inversion of resistivity data in an iterative procedure which includes a priori information of the model parameters without generating convergence problems.
Data smoothing includes single point correction, curvebranch shifting, and eccentricity correction (for Schlumberger arrays). These procedures are all carried out on the screen.
For the computation of apparent resistivity model curves the linear filter theory was used. Preference was given to accuracy over speed of calculation and therefore tests were performed on the various published filters to select the most adequate ones.
The inversion of resistivity data is based on the “Marquardt-Levenberg” technique. Experience has demonstrated that inclusion of a priori information of model parameters in this technique can result in convergence problems during the iterative procedure. Therefore, a modification of the “Marquardt-Levenberg” inversion algorithm was developed which considers a probabilistic treatment of the field observations and the model parameters. The approach allows the program to reach quickly a solution which is consistent with additional information.
Designed to process data obtained with Wenner, Schlumberger, and dipole-dipole arrays, the program is built around three main procedures: smoothing of noisy field data, accurate computation of apparent resistivity models and inversion of resistivity data in an iterative procedure which includes a priori information of the model parameters without generating convergence problems.
Data smoothing includes single point correction, curvebranch shifting, and eccentricity correction (for Schlumberger arrays). These procedures are all carried out on the screen.
For the computation of apparent resistivity model curves the linear filter theory was used. Preference was given to accuracy over speed of calculation and therefore tests were performed on the various published filters to select the most adequate ones.
The inversion of resistivity data is based on the “Marquardt-Levenberg” technique. Experience has demonstrated that inclusion of a priori information of model parameters in this technique can result in convergence problems during the iterative procedure. Therefore, a modification of the “Marquardt-Levenberg” inversion algorithm was developed which considers a probabilistic treatment of the field observations and the model parameters. The approach allows the program to reach quickly a solution which is consistent with additional information.
| Original language | English |
|---|---|
| Pages (from-to) | 691-703 |
| Journal | Computers & geosciences |
| Volume | 19 |
| Issue number | 5 |
| DOIs | |
| Publication status | Published - 1993 |
Keywords
- ADLIB-ART-1859
- ESA
Fingerprint
Dive into the research topics of 'RESIST: A computer program to process resistivity sounding data on PC compatibles'. Together they form a unique fingerprint.Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver