Decoupling induced polarisation data based on Cole-Cole parameters

Abstract

All resistivity/induced polarisation (IP) data are contaminated to varying degrees by inductive coupling. Spectral IP (SIP) data sets are usually more affected than conventional IP data sets as data are collected at higher frequencies, where coupling is stronger. For SIP surveys interpretation is based on examination of the phase response over the entire frequency range so it is necessary to use relatively sophisticated methods to remove coupling. Unfortunately, most of the methods described in the literature are complicated to implement, or are not described well enough to be practical. Pelton et al. suggest a method for decoupling based on the commonly used Cole-Cole equation that is relatively simple to implement using a Monte Carlo examination of the solution space. In this approach the Cole-Cole response is separated into two terms, the ground response and the inductive response. For each data spectrum, a random number generator is used to determine a set of test values for each the seven parameters in the extended Cole-Cole equation. The values for each of the seven variables are limited to â??reasonableâ?? levels based on initial testing of the solution space. The randomly generated Cole-Cole parameters are then used to generate magnitude and phase responses at the frequencies used for the survey. These are compared with the field data and the fit is calculated. This is run 1 million times for each data point and the best fit chosen from the complete run. Results from this study are encouraging, but the method requires refinement.