A method is described for removing the ionospheric-induced electromagnetic signals from electromagnetic time series in order to improve the detection of signals caused by oceanic, industrial, environmental, or seismic activity. The method is based on transforming the ionospheric-induced signals at a remote site to the local site using smoothly varying transfer functions. This assumes the remote site is relatively free of the non-ionospheric-induced signals. The transformed ionospheric-induced signal is then subtracted from the local time series. The method is tested on data recorded at Hollister and Parkfield that have been provided by F. Morrison. The sites are both near the San Andreas fault in California. Parkfield is used here as the remote site because it will be less contaminated by man-made sources since it is in a remote part of central California. The electric dipoles are approximately 100 m long with electrodes buried about 3 m deep and the magnetic sensors are three buried coils.
Between March 1989 and March 1994, annual self-potential (SP) surveys were carried out on Izu-Oshima, a small volcanic island. A terrain-related SP distribution of about -1 mV per meter of elevation was observed outside the caldera in all five surveys. Inside the caldera, SP increases from about -350 mV to near 0 mV (relative to the coastline) as the summit crater is approached, although negative anomalies of small spatial extent are manifest. Self-potential inside the caldera decreased by about 100 mV between the March 1989 and the March 1990 surveys, which appears to be correlated with a significant decline in the degassing rate from the summit crater. After 1990, the SP distribution is quite steady along the entire survey line which extends from the west coast through the southern part of the caldera, and ends east of Ura-sabaku. Recently a postprocessor has been developed to calculate space/time distributions of electrokinetic potentials resulting from histories of underground conditions (pressure, temperature, salt concentration, flowrate etc.) computed by multiphase multi-component unsteady geothermal reservoir simulations (Ishido and Pritchett, 1996). We applied this postprocessor to a simple two-dimensional model of hydrothermal activity in a volcanic island. The low potentials in areas of high elevation are reproduced in the model, and are caused by downflow of meteoric waters. The high potential centered at the summit crater is found to be produced by upflows of volcanic gas and vapor which diminish meteoric water downflow near the volcanic conduit.
Wide band and dense MT soundings on the hypocentral area of the Ebino earthquake swarm, which occurred in 1968, in southern Kyushu were performed to examine the existence of fluid in the hypocentral area, which possibly caused the earthquake swarm. The regional strike and the induction arrow were calculated first. Then a two-dimensional resistivity structure beneath the surveyed area is calculated by using an inversion technique. The best fitted model shows that the hypocentral area is relatively more resistive than the surrounding area. We conclude that a large mass of fluid is not likely to exist in the hypocentral area.
Miyake-jima Island, about 150 km south of Tokyo in Izu-Bonin Arc, is one of the most active basalt volcanoes in Japan. Big eruptions took place in 1940, 1962 and 1983. In this volcano, magma ascends towards a depth of a few km below the summit without any significant earthquakes or deformation, then gives rise to flank fissure eruptions because of the blockaded vent just beneath the summit crater. Hence eruption forecasts are very difficult to make with mechanical methods (i.e., seismic and deformation measurements) alone. We have developed an electromagnetic monitoring system of the volcano that combines magnetic, resistivity and electric field (SP) measurements. We expect that magma injection and the hydrothermal materials dispatched from it will result in thermal demagnetization, resistivity change and SP variations together with the electrokinetic-magnetic effect. Since October, 1995, we have continuously operated eight well distributed proton magnetometers over the island as well as two SP measurement systems on the NE and SW fissure zones. SP surveys brought to light distinct anomalies, which strongly suggest a close relation to the eruption mechanism. They are a positive anomaly up to 700 mV centered around the summit, and two negative ones amounting to -250 mV on the north and -100 mV on the southwestern mountainside. These anomalies can originate from a common mechanism: Rainwater penetrates from fissure zones along fractures toward the center of the volcano, a few km deep, where it is warmed by the heat supplied from deep-seated magma to rise through the summit vent. The down flow makes the negative, while the upwelling the positive SP anomalies, respectively. Miyake-jima island is located near the path of the Kuroshio, the most dominant ocean current in the western Pacific. A large magnetic variation amounting to several nT was observed to result from the meander of the flow path. This phenomenon produces serious EM noise and complicates monitoring for volcanic activity. It is crucial to investigate the characteristics of motionally-induced EM fields and to properly eliminate their effects.
The magnetic field, generated by the electrokinetic effect is calculated for a spherical, time-varying pressure source in a layered half-space chosen due to its exact solvability. In addition to a widely discussed steady-state phase, electromagnetic signals of electrokinetic origin have a transient phase which corresponds to their propagation through a conductive crust. The duration of the transient phase depends on a distance from there pressure gradient source and on the crustal conductivity. For characteristic conductivities of 0.1-10-3 S/m and distances of several kilometers the transient phase lasts for 10-100 sec. Although in the layered model both transient and steady state external magnetic fields are zero, this is not so in the general case. Therefore, if the electrokinetic effect occurs in a water filled fault, the transient magnetic field can appear at the surface as an ULF pulse.
Magnetotelluric (MT) data were collected at three sites around the eastern rim of the caldera of Axial Seamount, on the Juan de Fuca Ridge. The seamount has been observed to be volcanically and hydrothermally active over the last ten years, and is therefore an excellent target for electromagnetic induction studies on the seafloor. This paper follows an initial interpretation by Heinson et al. (1996) with a more complete analysis of the MT data, to investigate both oceanographic induction effects and the resistivity structure beneath the seamount. From time series analysis of electric field data using multitaper methods, coherences between electric field data from different sites are significant at the 95% confidence level at periods less than 1 day, and generally greater than 0.8 at solar and ocean tidal periods. Spectral peaks at 16.7 hours and 4 days are observed; the former is due to inertial currents in the area, and the latter is probably a ridge-trapped Rossby wave. Robust MT impedance tensors are derived using a remote-reference, and tensor decomposition shows that there is no galvanic distortion and almost isotropic responses at each site. The MT data are inverted for 1D structure, and more complex 3D forward models used to assess the lateral extent of the resistivity structure. 1D inversions show that the data are consistent with a crust with a very high electrical conductance of 1200 ± 200 S and an asthenosphere of 5-50 Ω·m at a depth of 40 km, connected by a low resistivity lithosphere of 50-100 Ω·m. The low resistivity lithosphere acts as a leakage path to the mantle for induced currents in the ocean. 3D forward modelling suggests that this region may be present only beneath Axial Seamount, surrounded by a resistive lithosphere of 500-50, 000 Ω·m. The tectonic implications from these models are that a small fraction of melt is presently migrating from the melt source in the mantle to a crustal magma chamber beneath Axial Seamount. Bulk estimates of melt fractions are 1-10% for the asthenosphere, and 1% between the asthenosphere and the crustal magma chamber, although melt may be concentrated in fractures or pipes.
Test measurements of the electric voltage in Kilpisjärvi, northern Finland during September 9-October 12, 1995 were made with the aim of studying the possibility of detecting a motionally induced electric field by means of a horizontal electric dipole 3800 meters in length located at a distance of about 50 km from the nearest fjord of the Norwegian Sea. In addition to the electric voltage we measured at Kilpisjärvi, the water level, magnetic field and meteorological data from Tromsø, Norway, the meteorological data from Kilpisjärvi and the magnetic field data from Sodankylä, Finland were used in analysis. Raw data were preprocessed to eliminate non-stationarities (remove outliers) and to remove solar daily variations, and were then smoothed. No significant influence of the local temperature, humidity and rainfall on the smoothed voltages was found. Probably due to a source effect the voltage was not coherent with the magnetic field from Tromsø and Sodankylä. The long-period anomaly with the period of about 16 days was observed in the smoothed voltage, water level and air pressure. More data are necessary to derive a definite conclusion about the nature of the observed phenomenon, which is probably associated with direct atmospheric forcing on the coastal zone of the Norwegian Sea.
For some time, oceanographers have used measurements of the small electromagnetic fields generated by the flow of the electrically conductive oceans through the Earth's main magnetic field to infer values for the ocean flow velocities. An overview of the process of motional induction is given including a description of the global electromagnetic fields generated by a global model ocean. We describe how electromagnetic methods are currently used in oceanography and outline the most important challenges presently faced.
We have examined magnetotelluric (MT) data from the Juan de Fuca plate and the Tasman Sea to understand the influence of galvanic distortion by local and regional topography. Galvanic (non-inductive) distortion, caused by the build up of charge along conductivity gradients, is present in both regions. Its effects can be removed by decomposition of the impedance tensor assuming that both electric and magnetic field distortion are present. Unlike for most land-based data, magnetic field distortion is necessary to explain the seafloor data. Electric field distortion parameters resulting from the full decomposition can be interpreted equivalently in terms of either 2-D or 3-D distorting bodies, but these interpretations cannot be differentiated using MT observations alone. The data are compared to the overall geology to determine whether 2-D or 3-D distortion is more probable. The electric field distortion parameters in both locations behave appoximately frequency-independently, as expected for galvanic distortion. When interpreted in terms of 2-D distorting bodies, the superficial strikes were found to parallel the trends of coastlines and large-scale submarine features (e.g. mid-ocean ridges). This suggests that distortion both in the Tasman Sea and on the Juan de Fuca plate is dominantly galvanic and appears to be caused by regional, large-scale, 2-D features rather than small-scale, 3-D distorting bodies.
Wide-band (20, 000 Hz to 0.002 Hz) magnetotelluric (MT) observations have been conducted since 1990 in the central part of the Tohoku district in the northeastern part of Japan to investigate the deep electrical resistivity structure. MT data were obtained on three east-west-trending MT transects about 140 km long and separated by about 20 km. In this paper we investigate three-dimensional galvanic distortions on the three transects using the Groom-Bailey tensor decomposition technique. We have focused on the regional strike direction and are not considering static shifts in the data. The decomposition method worked well at most sites but failed at some sites suggesting considerable regional three-dimensionality there. We found that the north-northwest to south-southeast direction is the predominant regional strike. This direction coincides with the regional geological strike being nearly perpendicular to the subduction direction of the Pacific plate. The induction vectors are generally in agreement with the regional strike direction deduced by the decomposition method on the southernmost line, but they deviate systematically on northernmost and middle lines suggesting conductivity contrasts in these areas.
Magnetic induction responses of elongated pairs and single conductors with strikes perpendicular to a local coastline of a deep ocean are studied with the aid of laboratory model measurements. The dependence on period, the width of the conductor, and the distance from the end of the conductors are examined for traverses over the conductors. It is shown that for the model geometry studied, the ocean induction effects can be subtracted to yield the Parkinson induction arrows for the conductor alone. At sites between pairs of parallel conductors the quadrature arrows compared with the in-phase arrows show a very differing period dependence. These results have particular application to the interpretation of geomagnetic measurements at sites between major elongated conductive structures such as sediment filled grabens.
One of the most challenging problems in electromagnetic (EM) geophysical methods is developing fast and stable methods of imaging inhomogeneous underground structures using EM data. In our previous publications we developed a novel approach to this problem, using EM migration. In this paper we demonstrate that there is a very close connection between the method of EM migration and the solution of the conventional EM inverse problem. Actually, we show that migration is an approximate inversion. It realizes the first iteration in the inversion algorithm, based on the minimization of the residual field energy flow through the profile of observations. This new theoretical result opens a way for formulating a new imaging condition. We compare this new imaging condition with the traditional one, obtained for simplified geoelectrical models of the subsurface structures. This result also leads to the construction of a solution of the inverse EM problem, based on iterative EM migration in the frequency domain, and gradient (or conjugate gradient) search for the optimal geoelectrical model. However, the authors have found that in the framework of this method, even the first iteration, based on the migration of the residual field, generates a reasonable geoelectrical image of the subsurface structure.
Electrical properties in most geologic materials have been known to be frequency dependent, and resulting dispersion relationship can be a useful diagnostic tool for investigating the shallow subsurface. In this paper we investigate the determination of dispersive electrical properties of the shallow subsurface with inversion of high-frequency electromagnetic (EM) fields. We have limited the dispersive characteristics to the electrical permittivity and used the Cole-Cole model to describe the frequency dependence of the permittivity. For horizontally layered earth models high-frequency EM fields are successfully inverted via Marquardt-Levenberg least-squares method and simulated annealing method. Inversion experiments show that the simulated annealing yields slightly better parameter resolution than the least-squares inversion.
Magnetotelluric (MT) data acquired during September-October, 1994, in northern Canada, were strongly influenced by non-uniform source field contributions from the auroral electrojet, and especially by intense auroral episodes. The largest effect on the estimate of the magnetotelluric impedance tensor elements was during intervals of highest magnetic activity, which primarily correlated with high auroral activity and was observed during local nighttime. In comparison, during the day the effect on the normal magnetotelluric impedance tensor response was usually, but not always, small. A robust controlled-leverage processing algorithm was applied to these data in an attempt to extract the stable uniform field estimates of the impedance. The differences between nonrobust and robust processing of the entire data set is compared to that obtained after dividing the time series into daytime and nighttime segments. The nonrobust estimate using all data is controlled by the nocturnal data, which are, in turn, dominated by non-uniform source effects. However, nonrobust processing of only the daytime data fails to recover a useful result. There is little difference between the robust response for the entire and daytime data provided that the fraction of auroral activity is not large, i.e., in excess of half of the available data series. In addition, examination of the time-dependence of the response functions shows that the strongest bias is observed during the initial quarter of an auroral event.
The two-dimensional (2-D) magnetotelluric (MT) inverse problem still poses difficult challenges in spite of efforts to develop fast and efficient methods for its solution. In this paper, we present a new approach for the solution of overparameterized cases based on regularization theory and full 2-D, quasi-analytic, calculation of the Frechet derivatives. For the forward solution we use a fast and efficient finite difference formulation to the solution of the MT equations in both transverse electric (TE) and transverse magnetic (TM) modes based on the balance method. The Frechet derivative matrix is obtained as a solution to simple forward and back substitution of the LU decomposed matrix of coefficients from the forward problem utilizing the principle of reciprocity. Magnetotelluric data is usually contaminated by noise, so that its inverse problem is ill-posed. In order to constrain the solution to a set of acceptable models, Tikhonov regularization is applied and yields a regularized parametric functional. The regularized conjugate gradient method is then utilized to minimize the parametric functional. Results of inversion for a set of synthetic data and for a set of CSAMT data from Kennecott Exploration show that the method yields models which are physically and geologically reasonable for both synthetic and real data sets.
We present an algorithm for modeling the magnetotelluric response of three-dimensional multilayered structures with irregular interfaces. In this formulation, based on a Rayleigh-Fourier technique, the effect of vertical anisotropy in the electrical conductivity has also been included. This method has an applicability range complementary to other solutions based on finite differences or on integral equations, which are especially adequate to model localized bodies intruded in a host medium. To test the method, the MT response of a simple conductive structure was modeled and compared with the solutions obtained using integral equations. A good agreement between the results with similar processing times have been observed. Finally, the effect of anisotropy was estimated for the particular case of a conductive basin, showing a non-negligible contribution, depending on the relation between the vertical and horizontal values of the conductivity.
We present a new, accurate, high-performance, wide-band three-dimensional (3-D) solver for the electromagnetic (EM) field scattering problem in an isotropic earth. The solver relates to those based on the volume integral equation (IE) approach and exploits a modified Neumann series (MNS) technique to solve Maxwell's equations. The solver allows for the conduction, polarization and displacement currents to be taken into account and admits for 3-D earth excitation by arbitrary electric or/and magnetic sources. We estimate the solver efficiency for scatterers discretized into Nx × Ny × Nz prisms, where it requires only about 6NxNyNz (log2(2Nx) log2(2Ny) + 6Nz) multiplications to get one term of the MNS expansion and about 200 NxNyNz2 bytes of memory. Our experience show that the number of terms N which are to be summed up to get the solution to 1% accuracy doesn't exceed fifty for the models with the conductivity contrast of up to 100. We demonstrate the solver versatility for magnetotellurics (MT) and controlled-source simulations. EM fields arising from a 3-D model with two high-contrast thin layers residing in layered earth were simulated due to a 10 Hz electric dipole located at the surface. When the layers were discretized into 16, 384 prisms our code on a Pentium-100 MHz took T ∼ 58 minutes, M ∼ 7 Mbytes and N ∼ 280. We also modeled the 0.1 Hz and 0.01 Hz MT responses within 3-D model with 1 Ω·m and 100 Ω·m blocks. When the blocks were discretized into 8, 000 prisms the code took T ∼ 5 minutes, M ∼ 8 Mbytes, and N ∼ 25. Finally fields for a crosswell model including a 3-D conducting target were simulated for 0.1 kHz and 10 kHz electric and magnetic dipoles in the wellbores. While the target was discretized into 6, 250 prisms the code took T ∼ 16 minutes, M ∼ 13 Mbytes, and N ∼ 24. All simulations showed from very good to excellent agreement with those of the other 3-D solvers.
We have developed a modified Neumann series (MNS) technique to solve Maxwell's equations for three-dimensional anisotropic earth. We assume that both conductivity and dielectric permittivity are 3 × 3 matrices, elements of which are complex-valued functions of space coordinates and frequency, whereas the magnetic permeability is symmetric 3 × 3 matrix with elements being real-valued functions of depth. Both conduction and displacement currents are taken into account. In order to derive the MNS solution, we impose the positiveness condition of Joule losses inside anisotropic earth. The MNS solution so obtained is valid for any frequency range and for arbitrary three-dimensional anisotropic earth.
Neogene extensional forces generated narrow rift zones (deep subbasins) in the Pannonian Basin, among others one of the deepest, the 7 km deep Békés graben. Above this structure strong magnetotelluric (MT) phase anisotropy (phase-deviation in two orthogonal directions) has been observed indicating the upwelling of the partially molten asthenosphere. This upwelling is certainly the first to prove the validity of the deep mantle structure of Buck's extensional narrow rift model by an electromagnetic method. A geotectonic relation will also be pointed out between the causes of MT phase anisotropy in cratonic and orogenic belts.
Research into the complex geological structure of the Middle European Variscides in the western part of the Bohemian Massif has largely intensified in connection with the KTB deep borehole project in the Oberpfalz, Germany. First results of long period magnetotelluric studies along an about 100 km long profile in SW Bohemia, close to the KTB drilling site, are presented here. They represent a partial step within a broader project of constructing a 3-D geoelectrical model of the western margin of the Bohemian Massif and resolving the relations of anomalous geoelectrical features to the geological structures of the region. The data roughly fit a composite model of a 2-D regional structure, striking E-W, overlaid by a layer of highly distorting near-surface inhomogeneities. The distorting layer displays large anisotropy with preferred conductivity in NW-SE to NNW-SSE direction. Attempt is made to estimate the absolute static shift of the MT curves. Several variants of 2-D models for the deep regional structure are discussed, including an anisotropic model which can qualitatively explain the discrepancy of the principal directions indicated by the MT and GDS data. The relation of the results obtained along the particular profile to the regional geoelectrical background is studied.
This contribution is aimed at suggesting and analyzing a thin sheet model of the electrical conductivity structure beneath Central Europe that would fit the induction response data, specifically the in-phase induction vector distribution. In particular, attention is paid to fitting the reversals of the induction vectors that suggest the existence of elongated zones of high electrical conductivity distinguishing main structural units.
Wideband magnetotelluric measurements were conducted along three traverses in the central Tohoku district of the northeastern Japan arc. These data provide independent constraints for understanding seismo-tectonics and geology where the Pacific plate subducts beneath the Eurasian plate. Smooth two-dimensional models of the data acquired along the three transects were obtained using the Generalized Rapid Relaxation Inversion algorithm. Modeling suggest that the area can be reasonably treated as two-dimensional with the structural direction being north-south in agreement with the strikes of geological units. The three 2-D models correlate well with the regional geological morphology in the central part of the Tohoku district which is divided by three high altitude regions. The resistivity profiles indicate two clear conductive anomalies in the Central Basin Range and in the Kitakami, Abukuma river regions that show considerable variability in each of the three 2-D models. Boundaries of conductive bodies correlate well with mapped faults, including those of pre-Tertiary age. The northernmost cross section is more complex possibly because of the presence of Quaternary volcanoes and an active geothermal source area in the vicinity.
After almost three decades of study, from its initial discovery in the 1960s to laboratory analyses of rock samples last year, we can now identify the most probable cause of the North American Central Plains (NACP) conductivity anomaly for much of its 1, 500-km strike extent. Tectonic processes operating during Paleoproterozoic Trans-Hudson orogenesis, with closure of the 5, 000-km-wide Manikewan ocean, included subduction and compression of sediments deposited during a hiatus in volcanism as the first of the advancing arcs approached the Archean continental margin to the west (Wyoming and Rae/Hearne cratons). These sediments were folded, and syngenetic sulphides within them migrated to concentrate along fold hinges, preferentially along strike, leading to high anisotropy in electrical conductivity (over 2-3 orders of magnitude). Mapping of the anomaly in similar tectonic environments, from the southern Dakotas to northern Manitoba, suggests that these processes were active along the whole western and northern margin of the orogen. However, other processes, possibly invoking graphitic emplacement in a foredeep, more likely account for the southern terminus of the anomaly from the Black Hills to southeastern Wyoming.
Nighttime transient geomagnetic variations recorded by an array of 29 magnetometers in the equatorial region of north-northeast Brazil have been subjected to robust regression analysis in order to derive transfer functions as a diagnostic indication of lateral conductivity variations. A thin sheet conductance model was developed to explain and interpret the variety of conductive anomalies evidenced by the induction arrow maps. The two main structures revealed by the present study are: a large NE-SW trending conductive anomaly with an embedded resistive zone in the central part of the Parnaíba Basin (Parnaíba Basin Conductive Anomaly-PBCA) and a relatively weaker anomaly branching off from the northwestern corner of the PBCA and extending towards the Marajo basin (LINK anomaly). The major PBCA anomaly is provisionally interpreted as a graben in the Precambrian basement filled with carbonaceous carbonates. The embedded resistive body, also characterized by a high density, is shown to be consistent with the presence of a diabase intrusive related to a magmatic event. The LINK anomaly is tentatively considered to be the relics of a sedimentary channel connecting the Parnaíba and Marajó basins.
A 3-D thin sheet model is developed synthesizing the data from different small array studies in the frontal Himalaya region. Most of the regional anomaly is simulated by a single highly conducting zone approximating an integral sign in shape (∫). The integral sign conductor (ISC) is basically an extension of Trans-Himalayan Conductor at the northern and southern ends towards east and west respectively. The conductance map indicates that the crust beneath the Indo-Gangetic Plains (IGP) lying on the western side of the ISC is electrically one order more conducting than on the eastern part. The synthesis of many geophysical signatures has permitted to infer that the Indian shield contiguous to the frontal Himalaya is mosaic of two crustal blocks of contrasting geophysical properties. In such tectonic scenario, the ISC is seen as a northeast continuation of the Aravalli range beneath the IGP and is interpreted to represent accretion zone resulting from the collision or underthrusting of the eastern crustal block beneath the western block. The clustering of epicenters along the ISC marks the continued activation of this ancient structure.
We discuss in a simplified surface-to-borehole example the resolution of the coefficient of anisotropy A for an anisotropic layer sandwiched in an isotropic half-space as a function of transmitter-receiver configuration and magnetic field components. Our analysis supposes that the horizontal resistivity of the anisotropic bed has been established by previous logging, and that only horizontal magnetic fields will be used to determine the vertical resistivity. The resolution analysis shows that for the model studied the coefficient of anisotropy λ can be resolved within 10% for a broad range of transmitter-receiver configurations if ∂Bz/∂t and ∂By/∂t are used jointly in the interpretation. Using only ∂Bx/∂t or ∂By/∂t individually, A can only be resolved for a limited range of transmitter-receiver configurations. A joint interpretation of both horizontal magnetic field components facilitates a better resolution of the coefficient of anisotropy compared to a single interpretation.
Various types of electrodes designed for the measurement of the electric field in the soil or in sea water at periods larger than one minute have been compared in a one-year experiment in Garchy, France. The experiment included more than fifty electrode pairs with liquid or absorbed electrolytes and Pb/PbCl2, Ag/AgCl, Cu/CuSO4 and Cd/CdCl2 metal-ion couples. The electrode parameters were systematically measured in the laboratory and the electrodes were installed in the field to constitute 50-meter long parallel dipoles separated by 2 meters. Pairs of electrodes used for sea measurements were monitored in a salted water vessel. Fourty-two potential differences were recorded with a sampling interval of 1 minute between May 1995 and April 1996. When electrodes are compared, large differences are observed in the long term stability as well as in the sensitivity to diurnal variations, rainfall and soil saturation. For measurements in soil, the installation method of the electrodes plays an important role. In salted water, the best performing electrode pair has a drift of the order of 0.1 mV per year. In soil, typical drifts for the best sensors are of the order of 0.2 mV per month in dry soil and 0.5 mV per month in soaked soil. Preferred electrode designs and installation methods, depending on the external conditions or the type of geophysical measurement, emerge from this experiment. In addition to the magneto-telluric field, potential variations which are not electrode or installation effects are observed and attributed to electrical sources in the soil.
MT and AMT surveys were conducted in the Minami-Kayabe area, southern Hokkaido, over a period of about two months in the fall of 1988. In this survey, 161 sites were arranged in a square lattice shape. FFT techniques were used in the initial data processing in 1998. Here we describe improved data processing using the cascade decimation technique. Data were processed efficiently by constructing a system using coherency and remote-reference. More than 10 sets of threshold values for the coherencies of (Ex, Hy) and (Ey, Hx) were then determined. Power spectra were sorted by these threshold values and then stacked. Satisfactory results were obtained for the frequency range from 3.4 Hz to 250 Hz. At frequencies below 3.4 Hz, however, the result of cascade decimation was not satisfactory because a sufficient number of stacks could not be obtained due to high local noise and limited measurement time.
Within the MT-DIW3 project of the comparison of various MT interpretation techniques, the controlled random search procedure was applied to the 2-D interpretation of the MT data from the Minamikayabe MT polygon in Japan. Three NE-SW profiles together with one perpendicular profile were 2-D modelled, simultaneously for both polarizations of the MT field. Although the 2-D approximation seems to be rather far from the real resistivity conditions, we could obtain relatively consistent interpretation results particularly for data within the short period range, up to 1 s. Better fit was achieved for apparent resistivities than for phases. For all NE-SW profiles, a near-surface conductive body is indicated below the central part of the polygon. At both sides of this body two deeper conductive zones are obvious. As for the perpendicular profile kb*s06, a large conductive zone is required below the SE margin of the profile.