|
Spectral Induced Polarization |
IntroductionThis page is meant to provide an awareness of spectral IP. It should become evident that chargeability is not a simple phenomenon. Using it for enhancing the ability to discriminate between mineral types is an attractice proposition, but this is far from easy to do. Definitions
There have been several studies suggesting that rock types have distinctive spectral characteristics. They are most often described in terms of how phase response depends on frequency. The figure to the right shows the phase response as a function of frequency (phase spectra) for typical target rocks (porphyry) and for inductive coupling. The so called Cole-Cole model for complex impedance is the most often used relation for modeling the ground's impedance. The Cole-Cole model is written as: This relation describes a complex impedance as a function of frequency, ω with three parameters. M is chargeability, τ is a time constant (of the decay curve), and c is a parameter controlling the frequency dependance.
If EM coupling is included, then the impedance relation will look like a combination of two Cole-Cole models: as discussed further in Appendix II: Noise. Attempts to use spectral IPBelow are some examples of phase spectra for two different rock types (from Telfordd, Geldart and Sherriff, 1990). Data such as these clearly provide some optimism that differentiation between minerals should be possible.
However, there has been rather mixed success at actually differentiating rock types in real field situations. This is perhaps because the causes for variations in phase spectra are not yet well understood. For example, grain size probably has at least as much influence on phase spectra as mineral type. Electrolyte type may also be very influential. Small amounts of clay are likely to have significant effects. None of these aspects is necessarily consistent amongst ore deposits. The use of phase spectra to remove EM coupling effects has been more successful because there is a significant difference between the spectra of all rocks, and inductive coupling. However, the most successful method of removing EM effects so far appears to be the method of Routh and Oldenburg, which does not depend upon quantifying Cole-Cole parameters for the various contributions to signals. Yuval and Oldenburg (1997) describe an inversion approach for extracting maps of τ and c from time domain IP data. Their method appears to work but they caution that the problem has not yet been solved. Also, relating these parameters to geology is still a poorly constrained problem. There have been some studies on use of spectral IP to help detect contamination in various soils, but it is not yet a main-stream technique. The example below is a good lab study from Soininen, H. and Vanhalan, H. (1996). It shows field data over a contaminated site on the left, and results of testing contaminated soil samples in the lab on the right. These data do provide some reason to believe that the technique may be useful, but there have not been many examples of routine use in field situations.
References
|
Spectral IP has been a topic of much research because there is a need both to discriminate between different types of rocks, and to remove the EM coupling effect.
