Study Of The Transient Electromagnetic Response Using Integral Equation Method

Complex electrical structures in earth attribute to the irregular behaviors of3D TEM response. Employing an effective and efficient3D electromagnetic modeling method makes much difference in distinguish the attributions of different factors, and lays a foundation for the3D inversion. The integral equations method (IEM) only needs to divide the anomaly regions which bring to it such advantage of high speed and little storage that more and more attention has been paid to it and much improvement made on it.We have revived and modified the tensor Green’s functions for the multilayer model which have been formulated by the predecessors and employed three new numerical technologies namely the quadrature-with-extrapolation (QWE), three dimensional Newton interpolation and the group theory in the code which is aid to speed up and perfect the performance of program. Then the3D forward code for modeling TEM response of magnetic sources is completed. Finally3D TEM responses for various models are calculated. Numerical results show that:(1) A relatively resistive anomaly within a uniform half space does not produce observable abnormal response while a more conductive anomaly brings to abnormal response which is much larger and decays more slowly and lasts longer. Increase on the depth to top of the anomaly slowly weakens the signal and delays the departure from the background curve. The amplitude of induced voltage is enlarged at all-time due to increase on size of the transmitter loop. Conductive overburden also delays the departure of the anomaly curve from the background curve.(2) The presence of overburden narrows the detective time window of the transient response. The response of overburden can be directly subtracted from the total response and interpreted using the uniform half-space model in case of a much resistive basement. Direct contact of the anomaly with a thin overburden can enhance the abnormal response significantly while with a thick overburden will not.(3) The large fixed loop configuration shows inability to resolute the multiples. Using a small moving loop can eliminate the uncertainty of interpretation after locating the anomalies roughly. The necessary condition to resolute two anomalies in the horizontal plane is their seperation should be no less than half of the loop size. The two minimums in the profile shows precisely the positions of them. If that is not satisfied, the single minimum in the profile just indicates the middle of the two anomalies.(4) Vary in resistivity of the shallow anomaly projects different effects on the TEM response. A relative conductive shallow anomaly affects the response much at early stage but little at late stage while a relative resistive one do much at early stage followed by dramatically drop and even negative attribution shows up at late stage. A shallow anomaly with same resistivity of the target in deep attributes less than a relative resistive one does at early stage while the former surpasses the latter at late stage.(5) The conductive shallow anomaly perplexes the TEM response and imposes a clear mask effect on the deep target at early stage. During late stage, a small size shallow anomaly increases the amplitude of response without damaging the shape of curves thus allowing for determining range of the target while the large one completely destroys TEM response of the target.(6) Induced polarization (IP) effect leads to a negative response (NR) in TEM. In comparison with ID IP effect, the positive response (PR) for3D IP effect is less.The NR for3D IP effect appears earlier and larger. IP effect shows up much later and drops dramatically with depth of the polarization increasing. A double sign reverse phenomenon is modeled using a model with a polarized patch and a conductive basement at certain depth.(7) A TEM decay curve bares a NR peak due to IP effect. The NR peak increases with the loop size increasing and reaches the maximum when the loop size is about half or one third of the body dimension. Cole-Cole parameters shows little effect on the PR while the NR peak increases with the chargeability or frequency parameter increasing and the time constant decreasing. The PR increases with the loop size increasing until the loop size equals the size of the anomaly and it shows a slight decrease if the loop size keeps increasing due to the effect of the relative resistive host. The NR peak curves show a clear maximum for a small size anomaly while the NR peak curves reach’maximum platform’within certain range of loop sizes for a large size anomaly. The PR increases with increase on dimensions of the anomaly especially for large size loop. The thickness of the anomaly has little effect on the PR and NR peak. The NR peak drops fast with increase on the depth to top of the anomaly while the PR drops more slowly. The amplitude of NR peak reduces significantly by almost one or two degrees if the loop reaches edge of the anomaly. The PR decays slowly with the offset increasing and shows a linear dependence on the loop size. Decrease on the host resistivity significantly weakens the NR peak while it has little effect on the PR. Increase on the anomaly resistivity makes both the PR and the NR peak drop dramatically.