Study About 1D Forward And Inversion Algorithm Of Centrl Loop Transient Electromagnetic Method

Transient Electromagnetic Method (TEM), also known as time domain electromagnetic (TDEM), is based on the process of transmitting the primary electromagnetic impulse to underground with ungrounded loop or ground source and analyzing changes of secondary field versus time to get the electrical characters of the medium. Because this method is only receiving the secondary field after turning down the first field power, it has the advantages of resolution, exploration depth and so on. TEM is now widely used in mineral exploration, water exploration, engineering and environmental geophysical exploration.Above all we review the development of transient electromagnetic method and briefly introduce the advantage and device type of TEM. We also introduce the current development and the future trend at the aspects of the theories and devices.On the aspect of forward calculation, we used the central loop transient response calculation, which is derived from the Maxwell's equations. The on-off step wave is usually used in transient electromagnetic prospecting. The real part and imaginary part of transient response is also used to simplify the formulas and acquire forward calculation formulas. There are two key steps --the fast Hankel transform and cosine transform. In this paper we analyze and compare the accuracy and computing speed of the fast Hankel transform between the linear digital filter of Anderson and the linear digital filter of Guptasarma and Singh. It's proved that the filter of Guptasarma and Singh, which we'll adopt finally, is better than that of Anderson on the aspects of accuracy and computing speed. We also analyze and compare the Cosine transform of the broken line approximation method and digital filter method. We analyze the accuracy and computing speed of the broken line approximation cosine transform briefly in different frequency points and bandwidth. On the comparison, numerical filtering cosine transform method has the higher speed than the broken line approximation method, so numerical filtering cosine transform method is advised to use.With the numerical forward calculation algorithm above, we make a computer program for central loop TEM forward calculations. Using this program, we simulate numbers of geo-electric models as the model types of D, G, A, H, K, Q. In addition we also use the one-dimensional forward to simulate the two-dimensional forward modeling which is called pseudo two-dimensional forward. With analyzing the shapes of these forward curves and comparing different experiments, it'll be a help for processing and interpretation of TEM data.On the aspect of apparent resistivity, the conventional central loop transient electromagnetic resistivity definition is classified as early and late apparent resistivity. In this paper, three definitions of all-time apparent resistivity are introduced as the magnetic field strength method, electromotive force method and translation algorithm method. We found that curves of the magnetic field strength method and electromotive force method are complete overlap, and then give the conclusion of the reason. Comparing the TEM showed it sensitivity between low resistivity layer and high resistivity layer and analyzed the sensitivity to low resistivity thin layer and high resistivity thin layer of different models. In addition, we also analyze the pseudo-anomies of electromotive force method and conclude the features of it, which will be benefit for explanation of TEM by using all-time apparent resistivity.On the aspect of inversion, the smoke loop inversion we used is based on smoke theory, which is a simple time-depth conversion method for TEM. In this paper, smoke loop inversion is based on all-time apparent resistivity. It is proved that the result of resistivity reflected from smoke loop inversion is better than all-time apparent resistivity. On the aspect of optimization inversion, we choose the LMF method and take inversions of several typical models with this method and conclude that it's better to use smoking loop inversion at first for qualitative explanation, to use the result of it as the initialized model of LMF inversion, and to use the result of LMF inversion for qualitative explanation. After experiments based the combination of the two methods it is proved that it not only improves the efficiency of the inversion calculation, but also makes the results more closer to real models.