Dikun Yang

3D inversion of total magnetic intensity data for time-domain EM at the Lalor massive sulphide deposit

The Lalor deposit is a massive sulphide that is characterised as a stack of conductive ore lenses buried more than 600 m deep. We invert helicopter sub-audio magnetics (HeliSAM) data from Lalor collected using a ground large transmitter loop and an airborne total magnetic intensity (TMI) magnetometer measuring the time-domain electromagnetic (EM) responses. The TMI data are modelled as a projection of the anomalous field onto the earth’s magnetic field. Inversion of these data is considered a significant case study because of two challenges. First, the early-time data are contaminated by the infrastructure on the surface. Second, inverting the data with a uniform half-space as the initial model results in a mathematically acceptable, but non-geologic, model. We create workflows to overcome these difficulties. For the contaminated data, we use a locally refined mesh and a constrained inversion to recover highly conductive cells near the surface that effectively represent the infrastructure. This allows us to safely extract geologic information from the early time data. The non-uniqueness in the inversion is reduced by warm-starting the voxel 3D inversion with a more reasonable initial guess, for example, a block model from geometric inversions. Those procedures greatly improve the inversion image from the surface to the bottom of the target at Lalor, and they can easily be incorporated into the industrial production workflows. 3D voxel inversion of HeliSAM data at the Lalor massive sulphide deposit suffers from: (1) contamination of early-time data by the near-surface infrastructure and (2) creation of a mathematically acceptable but non-geologic model from inversion of late-time data. We propose a procedure that incorporates the infrastructure in our inversion, and a warm-start approach to overcome the non-uniqueness.

Electric field data in inductive source electromagnetic surveys

Measurement of the electric field data due to an inductive loop source in a controlled source electromagnetic survey is not common, because electric field data, usually involving grounded electrodes, are expensive to acquire and difficult to interpret. With the recently developed capability of versatile three-dimensional inversion, we revisit the idea of measuring electric field in a large ground loop survey for mineral exploration. The three-dimensional modelling and inversion approach helps us quantitatively understand the detectability and recoverability of the proposed survey configuration. Our detectability study using forward modelling shows that the relative anomaly (percentage difference) in electric field does not decay with a lower induction number, but the conventional magnetic field data (dB/dt) does. Our recoverability study examines how much and what kind of information can be extracted from electric field data for the reconstruction of a three-dimensional model. Synthetic inversions show the following observations. (i) Electric field data are good at locating lateral discontinuity, whereas dB/dt has better depth resolution. (ii) Electric field is less sensitive to the background conductivity and, thus, is prone to misinterpretation because of a bad initial model in inversion. We recommend warm-starting the electric field inversion with an initial model from a separate dB/dt inversion. (iii) Electric field data may be severely contaminated by near-surface heterogeneity, but an inversion can recover the deep target concealed by the geologic noise. (iv) Even one line of single-component electric field data can greatly improve the horizontal resolution in a dB/dt inversion. Finally, we investigate a field dataset of both electric field and dB/dt measurements at a uranium deposit. The field example confirms that the electric field and magnetic field data contain independent information that is crucial in the accurate recovery of subsurface conductivity. Our synthetic and field examples demonstrate the benefit of acquiring electric field data along with magnetic field data in an inductive source survey.

Geophysical survey decomposition and efficient 3D inversion of time-domain electromagnetic data

Rigorous three-dimensional (3D) forward and inverse modeling of geophysical electromagnetic (EM) data can be time-consuming and may require a large amount of memory on expensive computers. In this thesis, a novel framework, called survey decomposition, is proposed to make the 3D EM modeling more efficient. Recognizing the multi-scale nature of the EM modeling problems, the fundamental idea is to break down an EM survey, which consists of many transmitters, receivers and times/frequencies, into a number of subproblems, each of which is only concerned about data modeled by a localized source, receiver and time/frequency. The modeling is then carried out on the subproblems at different scales, instead of the original problem as a whole. Such a decomposition is able to speed up the numerical modeling, because: (1) A subproblem can have highly efficient discretizations in space and time customized to its localized source, receiver, time/frequency and the specific scale of investigation, for example, it uses a local mesh that is much smaller than the one used in the original global problem; (2) A subproblem is a self-contained EM modeling problem that does not depend on other subproblems, so it is suitable for massive parallelization; (3) Upon decomposition, no modeling is carried out on the global mesh and the amount of computation is proportional to the number of subproblems, so the scalability improves significantly. After decomposition, the large number of subproblems is further reduced by adaptive, random and dynamic subsampling of the data. The adaptive scheme matches the number of samples to the scale of investigation so that only the data necessary for the model reconstruction are selected. The framework of survey decomposition is applied to two types of time-domain

Inversion of Multi-source TEM data with Applications to Mt. Milligan

We present a practical formulation for forward modeling and inverting time domain data arising from multiple transmitters. The underpinning of our procedure is the ability to factor the forward modeling matrix and then solve our system using direct methods. We formulate Maxwell’s equations in terms of the magnetic field, H and discretize the equations using a finite volume technique in space and a backward Euler in time. The MUMPS software package is used to carry out a decomposition of the forward operator, with the work distributed over an array of processors. The forward modeling is then quickly carried out using the factored operator. The factorization allows traditional Gauss-Newton inversion mthodologies to be implemented with greater efficiency than could be obtained from iterative techniques. As a demonstration we invert VTEM data at Mt. Milligan which is a Cu-Au porphyry deposit in British Columbia. 1D inversions produce a conductive artifact at depth that is inconsistent with geology. 3D inversions however, even from a limited number of stations, yield a more realistic result. Through the use of a synthetic model that emulates the geology at Mt. Milligan, we are able to show why the geologic artifacts arise from the 1D inversions. Lastly for the field data, we show how the resolution of the inversion result is further enhanced as progressively more transmitters are added.