Geophysical Prospecting, 2000, 48, 871885

An approximate analytical approach to compute geoelectric dipoledipole responses due to a small buried cube
Sa ndor Szalai1 and La szlo Szarka1

Abstract A simple analytical solution is presented for computing direct current (DC) electric field distortion due to a small cube in a homogeneous half-space, measured with a dipoledipole array on the surface. Both the transmitter and the receiver may have any orientation; furthermore their position on the horizontal surface and the depth of the cube can be freely selected. It is shown that a simple approximate analytical method may replace more complicated 3D numerical modelling algorithms. The approximation lies in the linearization of the problem: the secondary source (i.e. the cube) is considered as a system of three perpendicular electric dipoles. In spite of this first-order approximation, in the case of realistic depths z z=R < 0:10:5; where R is the transmitterreceiver distance), this approximate solution fits very well with true 3D numerical modelling results, and with analogue modelling results if a=R # 0:1; where a is the length of the side of the cube. Due to its simplicity, this method could be used for computing DC field distortion effects, estimating parameter-sensitivities, or even determining some initial models for further inversions. Introduction The DC dipoledipole technique has a Russian origin (Alpin 1941, 1950; Alpin et al. 1966). At that time the main reason for using a dipoledipole array instead of in-line AMNB configurations was to reduce the required cable length and to simplify the layout problem. Nowadays, DC dipoledipole systems are being used in new applications, e.g. in monitoring near-surface resistivity changes. Such problems require small and local resistivity changes to be detected. The 3D DC direct problems have been solved by using different numerical dara and Kaikkonen 1998), but modelling codes (e.g. Zhdanov and Keller 1994; Hvoz there is a requirement for faster and more efficient solutions. Analytical solutions for 3D DC anomalies have usually been applied using uniform primary fields (Honkura 1976; Ward and Hohmann 1987). In our solution a dipole
Received September 1998, revision accepted March 2000. Geodetic and Geophysical Research Institute of the Hungarian Academy of Sciences, H-9401 Sopron, POB 5, Hungary. E-mail: szalai@ggki.hu and szarka@ggki.hu

q 2000 European Association of Geoscientists & Engineers

871

872

S. Szalai and L. Szarka

field is used instead of a homogeneous and uniform primary field. In induced polarization phenomena, a somewhat similar technique was applied by Seigel (1959). Due to redundancy in the Maxwell equations and the related relationships, there are two approaches to the solution of DC problems. The first is based on the continuity equation and on the differential Ohm's law. In this approach the secondary source is given in terms of 7(s a7U), where s a is the anomalous conductivity and U is the scalar potential. In the second approach the secondary sources are assumed to be charges, accumulated at conductivity interfaces (Alpin 1941; Alfano 1959; Roy and Apparao 1971; Price 1973; Kaufmann and Keller 1985; Li and Oldenburg 1991; Szarka 1992). We use an extremely simple assumption about the charge distribution: the system of the resulting positive and negative charges on opposite surfaces of the cube is considered as an electric dipole. This is a simplification similar to the Born-like approximation of Zhdanov and Keller (1994). Such an assumption makes it possible to compute the secondary electric field as if it originated from a system of three perpendicular dipoles associated with the cube. Although the mechanism we describe is not new, we demonstrate for the first time that such a very simple approximate solution may serve as a substitute for much more complicated 3D algorithms. First some basic equations are summarized (definitions for the dipole moment, the electric field components due to a dipole, the rotational relationships and the surface charge accumulation), then the steps of the derivations (computation of the primary field, computation of the secondary source, computation of the secondary field) are presented. Parameter-sensitivity maps obtained by our approximate analytical method are compared with those obtained by an analogue modelling technique. A fairly good agreement illustrates that the proposed approximate solution gives realistic results. Finally, the verification of the method is demonstrated by comparing the approximate analytical results and 3D numerical results for different cube sizes, depths and positions, and 16 different dipoledipole configurations. The initial formulae The dipole moment If static charges +Q and 2Q are separated from each other by a distance AB, the classical definition of the moment p of an electrostatic dipole is pstatic QAB: 1a

In geoelectrics, instead of a static charge Q, a continuous current supply I is applied and the following dipole moment definition is used, pgeoel I AB: 1b

Among the other geophysical definitions, we mention that given by Alpin et al. (1966), who defined the moment of an electric dipole situated on the surface of the
q 2000 European Association of Geoscientists & Engineers, Geophysical Prospecting, 48, 871885

Analytical dipoledipole response

873

half-space with a near-surface resistivity @ 1 as pAlpin

@1 I AB: 2p

1c

As shown in the Appendix, I 101 @1 Q is just the continuous current supply needed to keep the charge constant in a conducting medium (1 0 is the dielectric permittivity, @ 1 is the resistivity of the medium around the dipole). In this way, the geoelectric dipole moment in the electrostatic approach is pstatic 10 @1 I AB: 1d

In order to resolve the paradox among various approaches, a common definition for the dipole moment is suggested as follows: m @ I AB; 1e 2p 1 where, in the case of the classical electrostatic definition, m 2p10 ; in the case of the electric definition, m 2p=@1 ; and, in the case of Alpin et al.'s (1966) definition, m 1: The dipole moment is given by either (1d), (1b) or (1c), depending on m. p Electric field inside a homogeneous half-space due to a horizontal electric dipole situated on the surface of the earth The potential U of a dipole of unit intensity, orientated in the direction of the unit vector e, can be obtained directly from the potential F of a point source of unit intensity, situated at the origin of the coordinate system (Korn and Korn 1968) as U r 2e7Fr: 2a According to (2a), the electric potential of a horizontal, x-orientated electric dipole m with a moment of p 2 p @1 I AB; situated on the surface of a homogeneous half-space, at the origin of an (x,y,z)-coordinate system is 2 @1 I x p x AB 3 2 ; 2b 2p r m r3 where r 2 x2 y2 z2 : The electric field components inside the homogeneous conducting half-space due to this dipole are U AB x; y; z E x x; y ; z E y x; y; z E z x; y ; z 1 p 3x2 2 r 2 ; m r5 1 p 3xy; m r5 1 p 3xz: m r5 3a 3b 3c

q 2000 European Association of Geoscientists & Engineers, Geophysical Prospecting, 48, 871885

874

S. Szalai and L. Szarka

According to the reciprocity theorem (Dachnov 1953; Harrington 1961), if both the transmitter/receiver positions and the transmitter/receiver components are interchanged, the measurement will be the same. In this way, the electric field at the surface due to a dipole situated inside a homogeneous half-space can also be expressed directly using (3a), (3b) and (3c). Charge accumulation on a conductivity interface According to Kaufman and Keller (1985), an electric field Eb, normal to a conductivity interface, will induce a surface charge density t , given by

t 210 kE b ;

where k s1 2 s2 =s1 s2 is the conductivity contrast and Eb is the normal component of the field vector (i.e. the average normal component of the electric field across the interface), orientated from medium 1 to medium 2 (Li and Oldenburg 1991). The total charge at a surface A is given by
A

t dA 210 k E b dA < 210 kE bmean A:

A

This approximation is the same as the Born-like approximation of Zhdanov and Keller (1994). Solution of the problem The faces of the small buried cube are orientated parallel or perpendicular to the transmitterreceiver line. The geometry of the problem is shown in Fig. 1. Three right-hand Cartesian coordinate systems are used: X xC, yC, zC: the coordinate system in which the faces of the cube are parallel to the axes. (The coordinates of the centre of the cube are xC, yC, zC. The coordinates of the centre of the transmitter in this coordinate system are (0,0,0) and those of the receiver centre are (0, R, 0), where R is the transmitterreceiver distance.) X xT, yT, zT: the coordinate system of the transmitter, rotated by an angle b from the xC,yC,zC system: xT xC cos b yC sin b; yT 2xC sin b xC cos b; zT zC : The transmitter dipole is orientated in the direction of positive xT. The 2 2 2 2 2 2 transmittercube distance is RT, where R2 T x T y T zT xC y C z C : X xR, yR, zR: the coordinate system of the receiver, rotated by an angle g from the xC,yC,zC system: xR xC cos g yC 2 R sin g; yR 2xC sin g yC 2 R cos g; 2 2 2 2 zR zC : The receivercube distance is RR, where R2 R x R y R zR xC 2 2 yC 2 R zC : The receiver dipole is orientated in the direction of positive xR.
q 2000 European Association of Geoscientists & Engineers, Geophysical Prospecting, 48, 871885

Analytical dipoledipole response

875

Figure 1. The geometry of the problem.

The steps of the derivation are as follows: 1 computation of the electric field E hom in the homogeneous half-space; R 2 computation of the anomalous (secondary) source (i.e. the determination of the three dipoles corresponding to the effect of the cube), pC {pxC ; pyC ; pzC }; 3 computation of the anomalous electric field due to the three subsurface dipole cube cube components, E cube E cube R pxC E R pyC E R pzC ; R hom total E cube 4 summing the primary and secondary fields, E R E R R : The mathematical derivation is summarized in Table 1 in a very condensed form. Sections (1), (2), (3) and (4) in Table 1 correspond to the steps of derivation. The last three columns of Table 1 give the three components x,y and z of the electric field. The derivation is equally valid for both the classical electrostatic and the different geophysical definitions of the electric dipole. It is easy to see that the effect of small targets is ultimately due to one single dipole. We decomposed it into three dipoles corresponding to the cube faces. In this way it is possible to study the individual effect of the side-pairs.

q 2000 European Association of Geoscientists & Engineers, Geophysical Prospecting, 48, 871885

Table 1. Geoelectric field distortion due to a small cube, buried in a homogeneous half-space in horizontal dipoledipole configuration. (For the rotation, the different coordinate systems and more information about m see the text.)
x (1a) The x,y,z components of Ex p m the electric field due to a horizontal electric dipole with a dipole moment m px 2 p @1 I AB at a distance of r, where r 2 x2 y2 z2 (1b) The xR component of the electric field due to a horizontal electric dipole with a dipole moment pxT 2 3x2 T 2r r5 x 3xy Ey p m r5 x 3xz Ez p m r5

876 S. Szalai and L. Szarka

q 2000 European Association of Geoscientists & Engineers, Geophysical Prospecting, 48, 871885

xT E hom pm G hom h 2  i R 3xT hom 1 1 T yT cosg 2 b 3xR where G R 2R sing 2 b 5 3 5 R5

pxT T (2a) The xC,yC,zC components E iC x; y and z i C ; where 2 m GT i 2 2 1 1 3 x 2 R cos b 2 3 x y sin b 3x2 of the electric field at the GT G yC R T T 5 T T 2 RT sin b 3xT yT cos b T xC R5 T T centre of the cube, obtained from ExT and EyT (2b) Dipole moments of the ti C 210 kE i C ; where i x; y and z electrostatic dipole pi C 210 kE i C a3 Qa ta2 a at the xC, yC, zC surfaces of the cube, where 1 2s2 t 210 kE ; k s s1 s2 pi C R E cube (2c) The secondary fields from R p i C p10 G i C ;where i x; y and z 2 2R R 2 2 1 1 2 dipoles pxC, pyC, pzC, G xC R5 3xR 2 RR cos g 2 3xR yR sin g G yC R 5 3xR 2 RR sin g 3xR yR cos g R R obtained using (1b), the rotation relation and the reciprocity relationship

1 GT zC R5 3xT zT
T

1 GR zC R5 3xR zR
R

(3) Sum of effects of dipoles pxC, pyC and pzC (4) The resulting total electric field

cube cube E cube E cube R R pxC E R pyC E R pzC pxT ka3 T R E cube G G C C R m p T T T R R R R where GT C {G xC ; G yC ; G zC } and GC {G xC ; G yC ; G zC }   3 T R xT E total E hom E cube pm G hom ka R R R p GC GC

Analytical dipoledipole response

877

Comparison with analogue and numerical modelling results The results of the analytical solution given in Table 1 were first compared to those of analogue modelling experiments. In Fig. 2, computed and measured parameter n 1996) are shown for sensitivity maps (Roy and Apparao 1971; Gyulai 1989; Hursa an axial dipoledipole configuration in the case of a burial depth of z=R 0:2: The isolines in this special case are almost identical, which means that the approximate analytical solution gives realistic results. In the next step, a systematic numerical comparison was carried out using a true 3D numerical code, based on an integral-equation approach (Zhdanov and Keller 1994). The transmitterreceiver configuration, the size and the relative position of the cube (its horizontal coordinates and its burial depth) were systematically varied. In this way, the validity of the proposed approximate analytical solution could be tested under various circumstances. The results as ratios of the approximate and numerical values (Eanalytical/ Enumerical) are illustrated in three selected examples, in Figs 3, 4 and 5. Figure 3 shows the results for a cube which is situated at the midpoint of the transmitter receiver line (cube position C1: xC =R 0; yC =R 0:5; in Fig. 4 the cube is off the transmitterreceiver line at a position xC =R 0:2 and yC =R 0:2 (cube position C2), while in Fig. 5 the cube is again placed along the transmitterreceiver line, but it is behind the transmitter (cube position C3: xC =R 0; yC =R 20:2: Each of Figs 3, 4 and 5 shows 16 different transmitterreceiver configurations (each subfigure corresponds to one configuration), six different cube sizes and five different burial depths. The rotation angles of the transmitter T and the receiver R both vary between 0 and 908 in steps of 308, indicated in Figs 3, 4 and 5 as TXX or RXX, where XX 0; 30, 60 or 908; the side of the cube varies between a=R 0:005 and 0.2 and the burial depths vary between z=R 0:1 and 0.5. If the cube is on the transmitterreceiver line (i.e. C1 (Fig. 3) and C3 (Fig. 5)), the 16 subfigures can be divided into four groups depending on the transmitter and receiver angles. (1) One of the angles is zero and the other angle is not 908 (Figs 3a,b,c, Figs 5a,b,c and Figs 3a,e,i, Figs 5a,e,i; altogether five subfigures in each figure). (2) One of the angles is 908 and the other angle is not zero (Figs 3h,l,p, Figs 5h,l,p and Figs 3n,o,p, Figs 5n,o,p; altogether five subfigures). (3) Neither the transmitter angle nor the receiver angle is zero or 908 (Figs 3f,g,j,k and Figs 5f,g,j,k). (4) The transmitter and the receiver are perpendicular to each other (Figs 3d,m and Figs 5d,m; altogether two subfigures). The following should be noted about the Eanalytical/Enumerical curves in Figs 3, 4 and 5. X In group (4) there is no signal and the ratio Eanalytical/Enumerical was set at 1. X The values Eanalytical/Enumerical for a=R 0:2 at a depth of z=R 0:1 are not shown because the numerical technique did not give a result. X If some of the values were found to be beyond the graphical limits (i.e. they were smaller than 0.6 or greater than 1.4), the curves end at the cube sizes where they

q 2000 European Association of Geoscientists & Engineers, Geophysical Prospecting, 48, 871885

878

S. Szalai and L. Szarka

reach either 0.6 or 1.4. In the cases C2T00R60 and C2T60R90 (Figs 4c and l), the curve fitting outside the rectangles is only indicative. Although the numerical modelling results are not always perfect (the numerical results are less precise where the body is close to either the transmitter or the receiver n 1999, pers. comm.), it can be seen clearly that the analytical and the (Hursa numerical values are in very good agreement for the case of the smallest cubes. If the side of the cube is less than 10% of the transmitterreceiver distance, the function jE analytical =E numerical 2 1j remains, in most cases, less than 0.050.1. (If the cube is situated between the transmitter and the receiver (C1), the greatest deviations between the approximate and the numerical results are observed in the depth range

Figure 2. Parameter-sensitivity maps. (a) Approximate solution and (b) analogue modelling results for axial dipoledipole configuration and a high-conductivity cube. The side of the cube is a 0:2R; its resistivity is zero and the midpoint of the cube is at a depth of z 0:2R: The numbers are expressed as percentages of the electric field measured by a dipole equatorial array over the same homogeneous half-space.

q 2000 European Association of Geoscientists & Engineers, Geophysical Prospecting, 48, 871885

q 2000 European Association of Geoscientists & Engineers, Geophysical Prospecting, 48, 871885

Analytical dipoledipole response Figure 3. Eanalytical/Enumerical electric field ratios as a function of the size (a/R) of the cube for 16 different dipoledipole configurations. The different curves in each subfigure correspond to different burial depths. (Their notation is shown in the bottom left-hand corner.) The position of the cube (C1 in the figure) is at the midpoint of the transmitterreceiver line xC =R 0; yC =R 0:5: 879

880 S. Szalai and L. Szarka

q 2000 European Association of Geoscientists & Engineers, Geophysical Prospecting, 48, 871885

Figure 4. Eanalytical/Enumerical electric field ratios as a function of the size (a/R) of the cube for 16 different dipoledipole configurations. The different curves in each subfigure correspond to different burial depths. (Their notation is shown in the bottom left-hand corner.) The position of the cube (C2 in the figure) is off the transmitterreceiver line, at the point xC =R 0:2; yC =R 0:2:

q 2000 European Association of Geoscientists & Engineers, Geophysical Prospecting, 48, 871885

Analytical dipoledipole response Figure 5. Eanalytical/Enumerical electric field ratios as a function of the size (a/R) of the cube for 16 different dipoledipole configurations. The different curves in each subfigure correspond to different burial depths. (Their notation is shown in the bottom left-hand corner.) The position of the cube (C3 in the figure) is along the transmitterreceiver line, at a distance of d =R 0:2 behind the transmitter xC =R 0; yC =R 20:2: 881

882

S. Szalai and L. Szarka

z=R 0:20:3 and the deviation itself depends strongly on the angle of the transmitter and that of the receiver. If the cube is near to the transmitter (C2), the greatest deviations are observed at smaller depths z=R 0:1; 0.2, 0.3). If the cube is behind the transmitter (C3), the greatest deviation appears in the depth range z=R 0:40:5: As can be seen in Figs 3, 4 and 5, the most important constraint in the application of the proposed approximate solution is obviously the size of the cube. At the same time, in some cases (see e.g. Fig. 4c), the high values of the ratio jE analytical =E numerical 2 1j are probably due to the fact that the cube extends over sensitivity zones of the opposite sign. If the results of the numerical computations are accepted as correct, we can state that the proposed approximate solution can be applied with an error of less than about 10%, provided the side of the cube is not longer than 10% of the transmitterreceiver distance. We should not forget that this error of 10% is expressed in terms of the secondary effect due to the body, and not in terms of the total field. Since the secondary field is usually much less than the total field, this error is in fact much less. If the ratio a/R (the ratio of the side of the cube to the transmitterreceiver distance) is about 20%, the image might still be acceptable in the case of many configurations (see the analogue modelling experiment for an example of this).

Geometry of the buried body It can easily be shown that the final effect of the buried cube provided it is not large does not depend on the rotation angle, so the cube may be rotated in the horizontal plane. Using our approach, the dipole moment for a rectangular prism (having sides ax, ay and az) can also easily be computed. In section (2b) of Table 1, the charge density at face x, for example, is proportional to ay az, and the displacement of the charges at the two corresponding faces is ax. Consequently, the dipole moment in any direction is proportional to ax ay az, i.e. to the volume V of the rectangular prism. In this way the components of the dipole moment induced by the prism are pi C 210 kE i C V ; where i x; y or z:

Since the electric field of the transmitter can be decomposed into any three components, the dipole moment due to a rotated rectangular prism or even due to a parallelepiped can be given in a similar way. It must be mentioned that the results obtained by the approximate solution become worse as the differences between the sides of the body increase. It is probable that some reasonable approximation can be found even for bodies of more complicated geometry, but so far we have only computed the effects due to small cubes.
q 2000 European Association of Geoscientists & Engineers, Geophysical Prospecting, 48, 871885

Analytical dipoledipole response

883

Conclusions An approximate analytical solution for computing the electric dipoledipole response due to a small buried cube is presented. It is based on the assumptions that 1 the electric field at the six faces of the cube is calculated from the computed electric field in a homogeneous half-space, at the coordinates of the centre of the cube; 2 the charge density over each face is considered to be homogeneous; 3 as a consequence of (1) and (2), an equal amount of charge of opposite sign is induced on the opposite faces of the cube, forming an electric dipole. The results are in a satisfactory agreement with 3D numerical and analogue modelling results, provided the cube is small enough, i.e. the side of the cube is not more than one-tenth of the transmitterreceiver distance. Before the advent of correct 3D modelling codes it was not possible to tell whether an analytical solution based on the linearization of the 3D DC problem was correct. As long as we remain within the validity of the modelling domain, such 3D responses can be obtained more easily and much faster than before. This simple method is strongly recommended for computing 3D DC field distortion effects, estimating parameter-sensitivities, or even determining some initial models for further inversions. Acknowledgements We thank our colleagues and the referees for their valuable comments.

Appendix In geophysics, the moment of a direct current dipole is widely accepted as IAB, where I is the current between the ground electrodes A and B. Here we show the relationship between the geophysical definition of the electric dipole moment IAB and the electrostatic one, which is known as QAB, where Q is the static charge. According to the continuity equation, the current is related to the time variation of charges, so that

2Q ; A1 2t H where Q S D dS (D is the electric displacement vector, S is a closed surface and the relationship between H D and Q is known as the integral form of the fourth Maxwell equation), and I S j dS (here j is the current density). In a homogeneous medium j sE and E 1 1 D (s denotes conductivity, 1 denotes permittivity of the medium and E is the electric field). These two constitutive H s s relationships are written as j 1 D; and so I 1 S D dS:
I 2
q 2000 European Association of Geoscientists & Engineers, Geophysical Prospecting, 48, 871885

884

S. Szalai and L. Szarka

In this way, for Q

D dS; the following differential equation holds, A2

s

s 2Q 0: Q 2t 1

The solution of (A2) is Q Q0 e2 1t : Its physical meaning is that a charge Q0 would disappear in a conducting medium very quickly, i.e. with a time constant of T 0 1=s: Comparing (A2) with (A1), it is clear that in order to keep Q constant (i.e. to have 2Q s 2t 0; a continuous current supply I 1 Q is needed. Using the resistivity @ 1=s; if 1 10 ; the charge can be rewritten as Q 10 @ I : A3

Therefore the moment of a direct current dipole in a medium of a resistivity @ will be in a perfect agreement with the electrostatic definition if it is defined as p QAB 10 @I AB: A4

This derivation is more detailed than that given by Kaufmann and Keller (1985). References
Alfano L. 1959. Introduction to the interpretation of resistivity measurements for complicated structural conditions. Geophysical Prospecting 7, 311366. Alpin L.M. 1941. Dipole Electric Sounding. Razvedke Nedr. No. 1. Alpin L.M. 1950. The Theory of Dipole Sounding. Geostoptekhizdat Moscow. Alpin L.M., Berdichevsky M.N., Vedrintsev G.A. and Zagarmistr A.M. 1966. Dipole Methods for Measuring Earth Conductivity. Translated by Keller G.V. Consultants Bureau, New York. Dachnov V.N. 1953. Electrical Exploration for Oil and Gas. Geostoptekhizdat Moscow. . 1989. Parameter sensitivity of underground DC measurements. Geophysical Gyulai A Transactions 35, 209225. Harrington R.F. 1961. Time-Harmonic Electromagnetic Fields. McGraw-Hill Book Co. Honkura Y. 1976. Perturbation of the electric current by a resistivity anomaly and its application to earthquake prediction. Journal of Geomagnetics and Geoelectrics 28, 4757. n G. 1996. On the role of lateral sensitivities in the horizontal electric profiling. Magyar Hursa Geofizika 37, 106117 (in Hungarian). Hvozdara M. and Kaikkonen P. 1998. The integral equations for solutions of the forward D.C. geoelectric problem for a 3-D body in inhomogeneous conductivity buried in the half-space. Journal of Applied Geophysics 38, 95107. Kaufmann A.A. and Keller G.V. 1985. Inductive Mining Prospecting: Part I. Theory. Elsevier Science Publishing Co. Korn G.A. and Korn T.M. 1968. Mathematical Handbook for Scientists and Engineers. McGrawHill Book Co. Li Y. and Oldenburg D.W. 1991. Aspects of charge accumulation in DC resistivity experiments. Geophysical Prospecting 39, 803826. Price A.T. 1973. The theory of geomagnetic induction. Physics of the Earth and Planetary Interiors 7, 227233.

q 2000 European Association of Geoscientists & Engineers, Geophysical Prospecting, 48, 871885

Analytical dipoledipole response

885

Roy A. and Apparao A. 1971. Depth of investigation in direct current methods. Geophysics 36, 943959. Seigel H.O. 1959. Mathematical formulation and type curves for induced polarization. Geophysics 24, 547565. Szarka L. 1992. Comment on ``Aspects of charge-accumulation in D.C. resistivity experiments'' by Y. Li and D.W. Oldenburg. Geophysical Prospecting 40, 823828. Ward S.H. and Hohmann G.U. 1987. Electromagnetic theory for geophysical application In: Electromagnetic Method in Applied Geophysics (ed. M.N. Nabighian), pp. 131311. Society of Exploration Geophysicists, Tulsa. Zhdanov M.S. and Keller G.V. 1994. The Geoelectrical Methods in Geophysical Exploration Elsevier Science Publishing Co.

q 2000 European Association of Geoscientists & Engineers, Geophysical Prospecting, 48, 871885