Induced polarization effects in inductive source electromagnetic data

Induced polarization (IP) surveys are commonly conducted to map the distribution of electrical chargeability, a diagnostic physical property in mineral exploration and in many environmental problems. Although these surveys have been successful in the past, the galvanic sources required make their application labour intensive and prevents them from being applied in some settings. The ability to detect chargeability with a geophysical technique that employs inductive sources, eliminating the need to inject current into the ground, would provide a valuable tool to applied geophysicists. In this work, two aspects of inductive source induced polarization are examined. First, a new methodology for processing inductive source frequency domain EM data to identify IP effects is presented. The method makes use of the asymptotic behaviour of the secondary magnetic fields at low frequency. A new quantity, referred to as the ISIP datum, is defined so that it equals zero at low frequencies for any frequency-independent (non-chargeable) conductivity distribution. Thus, any non-zero response in the ISIP data indicates the presence of chargeable material. Numerical simulations demonstrate that the method can be applied even in complicated geological situations. A 3D inversion algorithm is developed to recover the chargeability from the ISIP data and the inversion is demonstrated on synthetic examples. Understanding the impacts of IP effects on time-domain electromagnetic data requires the ability to simulate common survey techniques while taking chargeability into account. Most existing techniques preform this modelling in the frequency domain prior to transforming their results to the time domain. Application of those techniques on three dimensional problems can be computationally limiting. In the second part of this thesis, three new techniques for forward modelling the time-domain electromagnetic response of chargeable materials directly in the time domain are developed. The first considers the convolution in the time domain directly, while the others use auxiliary differential equations to analytically transform the governing equations into the time domain. The resulting methods are verified by comparing their results with analytic solutions. The potential application of the method was demonstrated by modelling the occurrence of negative transients in airborne time-domain electromagnetic data.Science, Faculty ofEarth, Ocean and Atmospheric Sciences, Department ofGraduat