BCGS Technical Talk – November 16, 2017
Speaker: Seogi Kang, PhD Candidate, UBC
Title: On recovering distributed induced polarization information from time domain electromagnetic data
Date/Time: Thursday November 16, 2017 @ 4:30pm
Location: 4th Floor Conference Room, Room 451, 409 Granville St. (UK Building at Granville and Hastings), Vancouver
Abstract:
The electrical conductivity of earth materials is frequency-dependent. This is due to a phenomenon known as induced polarization (IP), wherein electrical charges build-up under the application of an electric field. Macroscopically, earth rocks may be considered chargeable, as they act like electric capacitors. The goal of this thesis is to show how IP data can be extracted from geophysical data, then inverted to recover information about chargeable targets. Although both frequency and time-domain electromagnetic (EM) surveys measure IP signals, this presentation will focus solely on time-domain EM (TEM). To recover chargeability information, the following TEM-IP inversion workflow is developed. (1) Extracts a background conductivity model that is free of IP signals. (2) Decouple the TEM and IP signals by subtracting the fundamental responses estimated using the background conductivity. (3) Inverts the resultant IP data to recover pseudo-chargeabilities at multiple times for a set of 3D volumes. This is used to infer the location and dimensions of chargeable targets. (4) Carry out further analyses of pseudo-chargeabilities at multiple times to estimate intrinsic parameters such as a Cole-Cole chargeability and its associated time constant. For grounded sources, the workflow is implemented for a synthetic DC-IP example.
Results show that the early time signals, which are often discarded, can be used to estimate the background conductivity. Applying the workflow to inductive sources such as airborne EM (AEM) is more challenging, as steady-state electric fields are not produced. This was overcome by developing an IP function which (1) accurately characterizes how electric fields from inductive sources buildup in the earth and (2) allows the recovery of a 3D chargeability by solving a linear inverse problem. The efficacy of the aforementioned approach is validated using field AEM surveys over the Mt. Milligan porphyry deposit in British Columbia and Tli Kwi Cho kimberlite deposit in Northwestern Territories. For the kimberlite deposit, the recovered chargeability information is able to distinguish two distinct kimberlite units. To validate the approach, a 3D rock model is constructed using the recovered chargeability and background conductivity. This model is compared against geological models obtained through drilling and shows good agreement.