Chapter 2

Magnetovariational Method in Deep Geoelectrics

Mark N. Berdichevsky1 Vladimir I. Dmitriev Nina S. Golubtsova Natalia A. Mershchikova Pavel Yu. Pushkarev    Faculty of Geology, Lomonosov Moscow State University, Moscow, Russia

Abstract

Deep resistivity structure of the Earth's crust and upper mantle can be studied by two natural-source methods: magnetotelluric (MT) sounding, which uses electric and magnetic field variations, and magnetovariational (MV) sounding, which uses magnetic field variations only. Integrated MT and MV data interpretation is a multicriterion problem. MT and MV data have different sensitivity to resistivity structures, as well as different robustness to their 2-D approximation. We consider two approaches to MT and MV data integrated interpretation: parallel inversion and successive inversions of data components. In the first case, the selection of weights of data components is critical. We suggest successive inversions approach and test it on model data, calculated for the schematic 2-D resistivity model of the Kyrgyz Tien Shan, and experimental data, collected along the profile in the Cascadia subduction zone.

Keywords

magnetovariational sounding integrated data interpretation Tien Shan Cascadia subduction zone

Chapter Outline

2.1 Introduction 23

2.2 On Integrated Interpretation of MV and MT Data 26

2.3 Model Experiments 29

2.4 MV-MT Study of the Cascadian Subduction Zone (EMSLAB Experiment) 34

Acknowledgments 45

References 45

2.1. Introduction

Deep geoelectric studies of the Earth's crust and upper mantle include two methods: (1) the magnetotelluric (MT) method using the electric and magnetic fields and (2) the magnetovariational (MV) method using only the magnetic field. Following a common practice, a leading part belongs to the MT method with impedance tensor ^ and apparent resistivity ?a (vertical stratification of the medium, geoelectric zoning, mapping of underground topography, detection of conductive zones in the Earth crust and upper mantle, recognition of deep faults), whereas the MV method with tipper vector W and horizontal magnetic tensor ^ helps in tracing of horizontal conductivity contrasts, localization of geoelectric structures, determination of their strike. Such a partition of MT and MV methods is reflected even in the MT nomenclature: if the MT studies are referred to as MT soundings, the MV studies are considered as MV profiling (Rokityansky, 1982). The MT-MV geoelectric complex is widely and rather successfully used throughout the world. It provides unique information on the Earth's interior (porosity, permeability, graphitization, sulfidizing, dehydration, melting, fluid regime, ground-water mineralization, rheological characteristics, thermodynamic, and geodynamic processes). The weak point of deep geoelectrics with MT priority is that inhomogeneities in the uppermost layers may severely distort the electric field and consequently the impedance tensor along with the apparent resistivity. The distortions are of galvanic nature - they extend over the whole range of low frequencies causing static ("conformal") shifts of the low-frequency branches of apparent resistivity curves. The near-surface inhomogeneities affect the apparent resistivities, no matter how low the frequency is. They spoil the information on the deep conductivity. There is a plethora of techniques for correcting these distortions. But all these techniques are fraught with information losses or even with subjective (sometimes erroneous) decisions resulting in false structures. We can considerably improve the MT-MV complex by realizing to the full extent the potentialities of the MV method. The generally recognized advantage of MV method is that with lowering frequencies the induced currents penetrate deeper and deeper into the Earth, so that their magnetic field and consequently the tipper and magnetic tensor are less and less distorted by subsurface inhomogeneities and convey more and more information about buried inhomogeneities. This remarkable property of the magnetic field gives us the chance to protect the deep geoelectric studies from the static-shift problem (no electric field is measured). But excluding the electric field, we face the problem of informativeness of the MV method. It is commonly supposed that "MV studies determine only horizontal conductivity gradients, while the vertical conductivity distribution is not resolved" (Simpson and Bahr, 2005). Is it true? The fallacy of this statement is clearly seen from Figure 2.1, which shows a two-dimensional (2-D) model with an inclusion of higher conductivity in the upper layer resting on the resistive strata and conductive basement. The half width of the inclusion is 8 km. A depth to the conductive basement ranges from 25 km to 150 km. Let us compare the longitudinal apparent resistivity curves ?xy measured outside the inclusion (site O1, ? = -9 km), with the real-tipper curves Wzy=ReHza/Hy, measured at the same site O1, and with the magnetic-tensor curves yy-1=ReHya/Hyn, measured inside the inclusion (site O2, ? = 0). In the model under consideration, the bell-shaped MV curves Wzy and yy-1=ReHya/Hyn, derived from the ratio between the vertical component of the anomalous magnetic field to the horizontal component of the magnetic field and from the ratio between the horizontal component of the anomalous magnetic field to the horizontal component of the normal magnetic field, resolve the vertical conductivity distribution no worse than the customary MT curves ?xy. Generalizing these indications, we can say that the MV method reveals not only horizontal variations in the Earth's conductivity but the vertical variations as well. Moreover, we can appeal to the uniqueness theorem proved by Dmitriev for 2-D tipper and 2-D horizontal magnetic tensor and state that the 2-D piecewise analytical distribution of conductivity is uniquely defined by exact values of the tipper or the horizontal magnetic tensor given over all points of infinitely long transverse profile in the entire range of frequencies from 0 to 8 (Berdichevsky et al., 2003; Dmitriev and Berdichevsky, Chapter 7, this volume). The physical meaning of this unexpected result is rather simple. Naturally, the MV studies of horizontally homogeneous media with zero MV anomalies make no sense. But in the case of the horizontally inhomogeneous medium, the MV studies can be considered as ordinary frequency soundings using the magnetic field of excess currents distributed within a local horizontal inhomogeneity, which plays a role of the buried source. Figure 2.1 Illustrating the resolution of MT and MV soundings. Model parameters: ´1=100?Om, ?1=10?Om, w = 8 km, h1 = 1 km, ?2 = 10,000 Om, h2 = 24, 49, 99, 149 km, ?3 = 1 Om. Curve parameter: h = h1 + h2. So, we have every reason to revise the traditional MT-MV complex and consider a new MV-MT complex, within which the MV method, as being tolerant to subsurface distortions, plays a leading part and gives a sound geoelectric basis for MT-detailed specification. This approach goes back to the MT experiments that were performed in 1988-1990 in the Kirghiz Tien Shan mountains by geophysical teams of the Institute of High Temperatures, Russian Academy of Sciences (Trapeznikov et al., 1997; Berdichevsky and Dmitriev, 2002). These measurements were carried out at a profile characterized by strong local and regional distortions of apparent resistivities that dramatically complicated the interpretation of resulting data. The situation has normalized only with MV soundings. Figure 2.2 shows the real tippers, ReWzy, and the geoelectric model fitting these observation data. The model contains an inhomogeneous crustal conductive layer (a depth interval of 25-55 km) and vertical conductive zones confined to the known faults, the Nikolaev line (NL) and the Atbashi-Inylchik fault (AIF). The figure also presents the model reconstructed from seismic tomography data. The geoelectric model agrees remarkably well with the seismic model: low resistivities correlate with lower velocities. This correlation confirms the validity of geoelectric reconstructions based on MV data. We see that MV soundings not only outline crustal conductive zones but also stratify the lithosphere. Figure 2.2 Magnetovariational sounding in the Kyrgyz Tien Shan Mountains. (a) Plots of the real tipper along a profile crossing the Kyrgyz Tien Shan. (b) The resistivity section...