Chapter 5

Applied Geophysics: Exploring The Shallow Subsurface By Burger, Sheehan & Jones
Prepared By A. Fadil For FSSM UCA Graduate & Undergraduate Students

https://www.researchgate.net/profile/Abdelali_Fadil
1
Chapter 5
ELECTRICAL RESISTIVITY
2
5.1 INTRODUCTION
As we move from seismic topics to a discussion of electrical methods for exploring
the shallow subsurface, two striking differences are the diversity of electrical methods
and their relative complexity in deriving quantitative interpretations from field data as
compared to seismic methods.
Although a wide range of methods can be included under the heading of this chapter,
these methods group naturally into two categories: (1) those in which current is
applied to the earth and (2) natural energy sources.
5.1.1 Applied Currents

In electrical resistivity (ER) methods direct current or low-frequency alternating
current is applied at the ground surface, and the potential difference is measured
between two points. Variations in resistance to current flow at depth cause distinctive
variations in the potential difference measurements, which provide information on
subsurface structure and materials.
When the current used in electrical resistivity surveying is switched off, the voltage
between the two points used to determine potential difference does not decline
immediately to zero. Rather, a small residual voltage slowly decreases. A similar
effect occurs when the current is switched on, suggesting that the current flow has
electrically polarized the ground. The character of this induced polarization (IP) also
reveals information about the subsurface.
A third method using applied currents is electromagnetic (EM) surveying. In EM

surveying, a primary electromagnetic field is produced by passing alternating current
through a coil. Conducting bodies beneath the surface generate secondary
electromagnetic fields that are detected by a receiver coil. Differences between the
primary and secondary fields once again provide information about the presence and
characteristics of subsurface conducting bodies.
5.1.2 Natural Currents

The flow of charged particles in the ionosphere due to solar emissions is responsible
for alternating currents that flow through the upper regions of the earth. This natural
current flow (referred to as telluric currents) is altered by the varying conductivity
properties of rocks. The telluric method takes advantage of these natural current
variations by measuring potential differences at the surface and interpreting these
differences in terms of subsurface materials. The magnetotelluric method is similar to
telluric surveying but measures the magnetic field as well as the electrical field.
The self-potential or spontaneous potential (SP) approach uses natural

electrochemical activity. If an ore body is in contact with solutions possessing
different compositions, resultant chemical reactions cause a flow of ions. This leads to
potential differences and thus to a flow of current. A simplified view is a system
behaving as a battery. Because current is naturally produced, we need employ only
two probes to measure potential differences at points on the surface.
5.1.3 A Brief History

The history of electrical prospecting essentially dates from the 1830s, when Robert
W. Fox experimented with natural currents associated with sulfide ore deposits at
3
Cornwall, England. Most of the work during the 1800s dealt only with these types of
currents. In the early part of the second decade of 1900s Conrad Schlumberger in
France and Frank Wenner in the United States applied current to the ground and
measured the resulting potential difference. These investigations established the direct
current resistivity method.
In 1914 Schlumberger found a rich ore deposit in Serbia using the self-potential
method, and in 1917 the electromagnetic method was introduced by H. R. Conklin.
Telluric currents were first studied by O. H. Gish and W. J. Rooney in the United
States in the early 1920s. This work was continued and expanded by the
Schlumberger group beginning in 1934. Virtually all this early work quite naturally
was associated with ore deposits and their exploration. One of the earliest
noncommercial applications was mapping of high-resistivity bedrock by I. B. Crosby
and E. G. Leonardon in 1928 during an investigation of a proposed dam site.
With the exception of the magnetotelluric method, essentially all electrical methods
had been investigated in some fashion by 1930. Since then, of course, progress has
consisted of instrument refinement, development of a solid theoretical basis, and
improvement in interpretative methods, primarily, if not entirely, due to enhanced
computer capabilities.
5.1.4 Chapter Goals

Introducing electrical methods in a text with our objectives presents a number of
problems. First, as you now are aware, several electrical methods exist, and portions
of each contain unique theoretical aspects as well as operational procedures. Second,
all methods are complex theoretically, and quantitative interpretations are not nearly
as easy to achieve compared to the seismic methods with which you are familiar.
Covering electrical methods thoroughly, therefore, would take more room and time
than we devoted to seismic methods even though electrical methods are not as widely
utilized. A superficial coverage of each method would result primarily in confusion,
not mastery. In view of these concerns and constraints, we will concentrate on one
method only, electrical resistivity. Electrical resistivity is probably the most common
method applied to shallow subsurface investigations, especially in groundwater
studies. At the end of this chapter we provide an outline of the major additional
methods in terms of procedure and applications. We hope that mastery of one method
and knowledge of the major applications of the others provide a reasonable solution in
discussing electrical methods in an introductory text.
5.2 BASIC ELECTRICITY

The electrical resistivity method measures potential differences at points on the
earth’s surface that are produced by directing current flow through the subsurface.
This leads to the determination of resistivity distribution in the subsurface and to an
interpretation of earth materials. Because we are working with resistance, current
flow, and potentials, a brief review of basic electrical concepts seems warranted.
Figure 5.1 illustrates a basic electrical circuit containing a battery, connecting wires,
and a resistor. The battery maintains a potential difference between two points: its
positive terminal and its negative terminal. The battery thereby functions as a power
source in moving charges through the circuit, much as a pump moves water through
4
pipes. The convention adopted in this text is to define current flow as the movement
of positive charges. To accomplish this flow, the battery must move positive charges
from a low potential at the negative terminal to high potential at the positive terminal.
The work done in this potential change requires that a force be applied. This force is
known as electromotive force or emf, and the unit of emf is the volt.
A 9-volt battery maintains a potential difference of 9 volts between its terminals and
thus has a certain potential for doing work. As noted, the movement of charges
through the conducting wire is termed current. Specifically:
where i is current in amperes, q is charge in coulombs, and t is time in seconds.
Another important concept in electrical resistivity surveying is the current density j.

Current density is defined as the current divided by the cross-sectional area A of the
material through which it is flowing:
Because current is defined as the movement of charges across a given cross-sectional

area in a unit of time, maintaining a constant current and reducing the cross-sectional
area through which it flows must cause reduced spacing of charges and, therefore, an
increased current density. Figure 5.2 illustrates this relationship diagrammatically
using arrows to represent movement of charges.
5
You recognize, of course, that copper wire, wood, aluminum, and rubber possess
varying resistances to the flow of current. Copper has very low resistance, whereas
rubber has an extremely high resistance. Resistance is quantified in the following
way: one ohm of resistance allows a current of one ampere to flow when one volt of
emf is applied.
Ohm’s law states that current is directly proportional to voltage V and inversely
proportional to resistance R:
Consider Figure 5.1. If the battery supplies 9 volts and the resistor has a value of 10
ohms, the current measured by the ammeter will be 0.9 ampere. If resistance is
increasing, it will take an increasing voltage to maintain the same current.
Because various geologic materials can be expected to have different resistances to

current flow, it might seem fairly straightforward to measure current and voltage to
calculate resistance and determine the material in the subsurface. One immediate
complication is that resistance depends not only on the material but also on its
dimensions. Consider Figure 5.3. The sketches illustrate two resistors with different
lengths l and cross-sectional areas A. If these resistors are constructed from the same
material, it seems intuitively obvious that they will not have the same resistance to
current flow. Remembering that current flow is the movement of charged particles in
a given unit of time, we can use a water analogy. Consider an open pipe in which one
section is filled with gravel. A pump creates a pressure difference in the pipe and
water flows. The gravel causes a resistance to flow relative to the open portion of the
pipe. If we keep everything the same but increase the length filled with the gravel, the
resistance to flow increases and the flow rate of the water is reduced. If, however, we
increase the diameter, the resistance to flow is reduced and more water flows.
This behavior suggests that the resistances of the resistors in Figure 5.3 depend on
their length and cross-sectional areas and also on a fundamental property of the
material used in their construction, which we term resistivity and denote by ρ. Based
on our discussion, we can say that:
6
Resistivity units are resistance-length, which commonly is denoted by ohm.m (Ω.m).
Conductance is the inverse of resistance, and conductivity is the inverse of resistivity.
Copper has a resistivity of 1.7 x 10-8 Ω.m. What is the resistance of 20 m of copper
wire with a cross-sectional radius of 0.005 m? Quartz has a resistivity of 1 x 1016
Ω.m. What is the resistance of a quartz wire of the same dimensions?
5.3 CURRENT FLOW IN A HOMOGENEOUS ISOTROPIC EARTH

5.3.1 Point Current Source
Because the resistivity method consists of applying current and measuring potentials,
we begin by considering the potential at a point P1 when current is applied at a point
source C1. We place the return current electrode at a very great distance and assume
material of uniform resistivity, ρ. Because air has infinite resistivity, no current flows
upward. Thus current flows radially outward through the earth equally in all
directions so as to define a hemispherical surface (Figure 5.4).
Because current distribution is equal everywhere on this surface that is at a distance r

from the current electrode C1, the potential also is equal. These surfaces are known as
equipotential surfaces. If we define a very thin shell of thickness dr and employ
Equations 5.3 and 5.4, we can define the potential difference across the shell to be:
We now use Equation 5.6 to determine the potential at P1. In determining the potential
at a point, we compare it to the potential at a point infinitely far away, which by
convention is arbitrarily defined to equal zero. The most direct way to determine V is
to integrate Equation 5.6 over its distance D to the current electrode to infinity, or:
Equation 5.7 is the fundamental equation in our electrical prospecting discussion.

Before using it to develop more practical relationships, let’s examine what
information we can glean from it. Assuming a resistivity and current, we can map the
7
potential at any point in our section view of Figure 5.4. If we plot enough points, we
can draw lines through the points possessing the same potential. This defines the
equipotential surfaces. Because current flow must be perpendicular to the
equipotential surfaces, we can determine the direction of current flow. Of course, in
this case we already know what the result will be; but in the next step the pattern will
not be so evident.
5.3.2 Two Current Electrodes

Our next step is an attempt to determine the current flow in a homogeneous, isotropic
earth when we have two current electrodes. In this case the current must flow from
the positive current electrode (the source) to the negative current electrode (the sink).
The path of the current is not as obvious as in our previous discussion, so we once
again determine the potential at a point. From this information we determine the
equipotential surfaces and then the current flow. For simplicity we present the
derivation only for the case when the current electrodes and our potential points lie on
the same plane (Figure 5.5). The more general case is quite similar and can be
examined in advanced textbooks.
The potential at point P1 is determined by using Equation 5.7. The effects of the
source at C1 (+) and the sink C2 (-) are both considered, and therefore:
Expressing r1 and r2 in terms of the x-z coordinate system illustrated in Figure 5.5, we
rewrite Equation 5.8 as:
8
9
Now we need only calculate VP1 at many points in the x-z plane, draw contours
through points of equal potential to define the equipotential surfaces, and, finally,
draw current flow lines perpendicular to the equipotential surfaces.
Table 5.1 is a dynamic table based on Equation 5.9. Parameters in the table are related
to Figure 5.5. If you examine the values of the potential at the points in the table, you
should be able to sense the distribution of the equipotential surfaces. Figure 5.6 is a
simple contour plot of the values in Table 5.1with symbols also keyed to Figure 5.5.
The patterns of equipotential surfaces likely are fairly similar to what you would
predict. Note that the current electrodes are displaced slightly from the centers of the
hemispheres. Many contour lines close to the current electrodes were omitted for
clarity, as you can tell if you examine contour line values and the potential values in
Table 5.1.
We know that the current flow lines are perpendicular to the equipotential lines, but
we do not know how the current is distributed. The mathematical analysis to
determine the current distribution is fairly complicated but results in a simple
equation that provides current distribution as a fraction of the total current. Along a
vertical plane midway between the two current electrodes, the fraction of the total
current if penetrating to depth z for an electrode separation of d is given by:
10
The dynamic Table 5.2 uses Equation 5.10, so we can investigate current distribution
for various current electrode separations. If you try several values for current
electrode separation, you should observe that the values in columns two and three in
Table 5.2 do not change. A graph of these relationships is illustrated in Figure 5.7.
Examining these relationships allows us to arrive at certain conclusions. Fifty percent

of the current is confined above a horizontal plane with a depth of one-half the current
electrode separation. Seventy percent of the current is confined above a depth equal to
the electrode separation (dashed lines in Figure 5.7). Clearly, the greater the electrode
separation, the greater the depth to which a given percentage of current penetrates.
Make a firm mental note, however, that this relationship is for a homogeneous,
isotropic subsurface. Once we introduce a discontinuity, we must reexamine the
current distribution question.
Now we can take the information that Equation 5.10 supplies and combine it with
Figure 5.6 to produce the current flow lines that are diagrammed in Figure 5.8. We
will return several times to these current flow lines, their spacing, and their
distribution, so be sure to grasp what they represent and how they were derived.
5.3.3 Two Potential Electrodes

In electrical resistivity surveying our goal is to measure the potential difference
between two points just as we often make this measurement in electrical circuits.
Examine Figure 5.9 and compare it to Figure 5.5. Figure 5.9 illustrates two potential
electrodes, P1 and P2 that are located on the surface, as are the current electrodes.
Because we already derived an equation to determine the potential at a point due to a
source and a sink, we obtain the potential difference by determining the potential at
one potential electrode P1 and subtracting from it the potential at P2. Using Equation
5.8 we determine that:
11
12
Therefore, the potential difference ΔV equals:
In the resistivity method current is entered into the ground, potential difference is
measured, and resistivity is determined.
Because resistivity is the unknown quantity we normally hope to determine, we solve

Equation 5.14 for ρ and obtain:
Perhaps we should test our understanding of Equation 5.15 by applying it to a known

situation. Let’s assume we can place potential electrodes anywhere along the surface
as illustrated in Figure 5.9. Further, we will use the values in Table 5.1 for our test.
Figure 5.10a presents one possible measurement, and Figure 5.10b another.
Substituting the values in Figure 5.10a produces Equation 5.16, which results in a
resistivity value of 50 ohm.m. A glance at the model values used to produce Table 5.1
confirms that the resistivity was 50 ohm.m.
Equation 5.17 uses the values illustrated in Figure 5.10b. As expected, evaluation of
the equation gives 50 ohm.m as the resistivity.
These calculations confirm that if we produce a current, measure the electrode

spacing, and determine the potential difference, we can arrive at a value for the
resistivity of the subsurface materials. However, until this point we have assumed that
the subsurface has a constant resistivity. Just as in our development of the seismic
method, it is now time to introduce a horizontal surface that separates materials with
different resistivities.
13
5.4 A SINGLE HORIZONTAL INTERFACE

We already noted that derivation of equation for many electrical resistivity problems
is mathematically quite complex and will not be attempted in this text. Our strategy in
presenting the necessary equations varies. In some cases we simply introduce an
important equation and you may consult advanced textbooks if you are interested and
have the mathematical background to follow the derivation. In others cases we present
a simplified case or present only part of the derivation if we feel this will help explain
the concept. And finally, we may present only a qualitative explanation or couple
such an explanation with an equation or a partial derivation.
5.4.1 Current Distribution

An important goal in this section is to gain a qualitative understanding for the pattern
of current distribution in the subsurface when a single horizontal interface separate
materials of different resistivities. Our first step toward this goal is to employ an
equation that tells us the fraction of the current that penetrates below the interface.
This current fraction is given by:
14
where
Although this equation is relatively complex, none of the quantities should surprise
us: current fraction if, electrode spacing a, depth of interface z, and resistivities of the
materials above and below the interface ρ1 and ρ2.
15
Note that in this equation electrode spacing a refers to the spacing between each
electrode in the array (Figure 5.11a), whereas previously we usually referred to the
spacing between current electrodes. The relationships that we can deduce from this
equation form the basis for a qualitative interpretation of field measurements, so let’s
take some time to gather information.
Equation 5.18 is sufficiently straightforward to permit entry into a spreadsheet for

evaluation, and Table 5.3 is the result. The appearance of this table is slightly
different than those we have seen previously because the equation being evaluated is
broken down into several components due to its complexity. The major variables in
the equation are placed in the upper left corner. The result is given as percentage of
current and is located in the upper right. The components are:
16
The numbers in the result column are added together (see the summation symbol in
the Equation 5.18) and multiplied by C. This gives the current fraction, which is then
presented as the percentage of current.
If we select a depth for the interface and for electrode spacing (perhaps in the ratio
z=a) and vary values for ρ1 and ρ2 to produce k-values ranging from -1.0 to 1.0, we
can create a curve such as one of the three illustrated in Figure 5.11. All three of these
curves were produced using values calculated by Table 5.3. If ρ1 equals ρ2, then we
have the previous case of the homogeneous subsurface. The result should be the same
as predicted by Equation 5.10 and illustrated in Figure 5.7. If you insert equal
resistivities into Table 5.3, place the depth of the interface at 24 m, and elect an
electrode spacing of 8 m (which is a current electrode spacing of 24 m), the result in
the table tells us that 30 percent of the current penetrates beneath the level of a plane
located at a depth of 24 m (see curve z=3a in Figure 5.11b for k=0). Figure 5.7
confirms this by illustrating that 70 percent of current flow is above a plane at a depth
equal to the current electrode spacing.
Now let’s examine the situation when ρ2>ρ1. The material below the interface has a
higher resistance to current flow (it is a poor conductor relative to the layer above).
The k-values will be positive for this case. The three curves in Figure 5.11b all have
the same form, so we can generalize our comments. If we select any positive k-value,
we see that substantially less current penetrates below the level of the interface
compared to the homogeneous case (k-value=0). In other words, current flow tends to
avoid a poor conductor in favor of a good conductor. Perhaps a water analogy will
make this phenomenon clearer. Visualize a homogeneous sequence of coarse sand
with water flowing through. If at some depth z we introduce a relatively low-
permeability layer, less water will flow below that depth level than in the
homogeneous case. Of course, the extreme case is when the layer below the interface
is perfectly insulating (or completely impermeable), and no current (or water)
penetrates below that level.
If ρ2<ρ1 the material below the horizontal interface is less resistant to current flow
than the material above the interface and is, therefore, a better conductor. As Figure
5.11b demonstrates, more current will penetrates below the interface than in the
homogeneous case. Using the water analogy once again, we replace our homogeneous
material with coarse gravel that has very high permeability. More water now will flow
below the level of the interface and through the gravel. Note that a substantial
percentage of the current flows through the material above the interface even if the
material below the interface possesses very low resistivities (is almost perfectly
conducting).
In our previous discussion of current flow in a homogeneous, isotropic subsurface we

learned that as current electrode spacing is increased, the percentage of current
17
flowing at depth is increased. For example, only 30 percent of the current penetrates
below a depth z when the electrode spacing is equal to one-third the depth (z=3a).
However, 50 percent of the current penetrates deeper than z when the electrode
separation is increased to two-thirds the depth (z=3a/2), and 63 percent of the current
penetrates below z when the electrode spacing equals the depth (z=a). Remember that
we are using electrode spacing as diagrammed in Figure 5.11a. Thus, when the
electrode spacing is equal to the depth of our plane of reference (z=a), the current
electrode separation is three times the depth (3z=3a=d).
Figure 5.11b indicates, as expected that current penetration increases as electrode

separation increases. However, the percentage of current penetrating below an
interface is controlled by the relative magnitudes of ρ2 and ρ1 as well as by electrode
separation. This relationship plays an important role in data acquisition design, so it is
a good idea to thoroughly understand the implications of Figure 5.11.
5.4.2 Current Flow Lines and Current Density

The previous discussion gives us sufficient information to qualitatively assess current
flow lines and, more important, current density distribution when a horizontal
interface is present. As a first step in this process, we must investigate what happens
to the orientation of flow lines and equipotentials when they cross a boundary
separating regions of differing conductivities or resistivities. The flow lines follow a
tangent relationship such that:
where θ and ρ are as defined in Figure 5.12a. If the resistivity ρ2 of the deeper material
is greater, the flow lines bend in toward the normal to the interface (Figure 5.12b) and
as a consequence are more widely spaced. However, if the reverse is true as in Figure
5.12c, the flow lines bend away from the normal, become oriented more parallel to
the interface, and are closer together.
If we take this relationship and the information we developed to produce Figure 5.11,
we can assess the effect of the horizontal interface. Figure 5.13a illustrates the pattern
of current flow lines for a homogeneous subsurface (ρ2=ρ1). If we increase the value
of ρ2 (Figure 5.13b), more current flows above the interface, the current flow lines are
spaced more closely, and the current density is greater in the region above the
interface relative to the case of the homogeneous subsurface. If ρ2<ρ1, the opposite
effect takes place. A greater percentage of current flows beneath the interface (Figure
5.13c), the current flow lines are spaced more widely in the material above the
interface, and the current density there is reduced.
Our real purpose is to determine, qualitatively at the moment, how the presence of the
interface affects ΔV readings at the surface. We are just about ready to take this step,
but first we need to introduce the notion of apparent resistivity.
18
19
5.4.3 Apparent Resistivity
When we derived Equation 5.15, we assumed a homogeneous, isotropic subsurface.
As demonstrated previously, any combination of electrode spacing and current results
in a potential difference that provides the correct value for the resistivity of the
subsurface (as of course should be the case if our equation is correct). Once the
subsurface is nonhomogeneous, the value determined for the resistivity is extremely
unlikely to equal the resistivity of the material in which the electrodes are inserted.
Equation 5.15 thus defines a different quantity, which is termed the apparent
resistivity, ρa. Inasmuch as non-homogeneity is the rule, we now write:
What does this equation tell us? How do we interpret apparent resistivity values in
terms of subsurface geology?
If we examine Equations 5.2, 5.3, and 5.5, it becomes straightforward to demonstrate

that the potential is proportional to current density:
When we measure the potential difference between the potential electrodes, the values
are proportional to the current density in the cylinder of material near the surface that
extends between these two electrodes. If we examine Equation 5.20 and remember the
relationship expressed in Equation 5.21, we conclude that, all else being equal,
variations in current density near the surface will result in variations in apparent
resistivity. Let’s return to Figure 5.13.
In the case of Figure 5.13a we already know that Equation 5.20 will yield a value of
ρ2=ρ1. In the case represented by Figure 5.13b the current density is increased in the
upper layer relative to the homogeneous case, so ρa will be greater than ρ1.
Conversely, in Figure 5.13c the current density is decreased due to the lower-
resistivity material at depth, and therefore the value of ρa will be less than ρ1. Notice
that the electrode spacing remains constant in this example.
5.4.4 Qualitative Development of the Resistivity Pattern over a Horizontal Interface

By now you should have a good grasp of how the horizontal interface affects current
flow, current density, and the resistivity that we measure at the surface (the apparent
resistivity). But how do we use this information to explore the subsurface as we did
using seismic methods to determine the depth to the interface and the nature of the
geologic materials above and below the interface? As a first step we continue our
qualitative approach, but shortly we will develop quantitative methods.
If a horizontal interface is deep relative to electrode spacing a or current electrode

spacing d as defined in Figure 5.11a, it is unlikely that the presence of the interface
will have much effect on the current density between the potential electrodes.
Therefore, ρa will equal or be very close to ρ1.
This relationship is diagrammed in Figure 5.14, where dashed black lines represent
current flow if no interface is present and solid gray lines represent current flow due
to the interface. In Figure 5.14a both sets of lines coincide. If we simply move this
20
electrode array about on the surface, all reading will be the same assuming the ρ1
layer is laterally homogeneous.
However, if we expand the electrode spacing, more current will flow at depth, and the
current density will be affected by the ρ2 layer. The apparent resistivity value will be
greater than ρ1 but still will be substantially less than ρ2 (Figure 5.14b). If electrode
spacing is increased still more, the effect of the higher-resistivity material below the
interface is even greater, as illustrated in Figure 5.14c. The value of apparent
resistivity continues to increase and gradually approaches the ρ2 value while
becoming increasingly greater than the ρ1 value.
21
In many resistivity investigations electrode spacing is plotted against apparent

resistivity. If we sketch a curve based on our previous analysis, it should have a form
similar to that in Figure 5.15a. Of course, the actual form will depend on the
resistivity contrast, the depth of the interface, and the electrode spacings. At the
moment it is important only that you understand why the general shape of the curve is
as illustrated. If we reverse the situation so that the material below the interface has a
lower resistivity than the material above, the curve assumes the form illustrated in
Figure 5.15b. Various curves for other subsurface situations are illustrated in
advanced textbooks, and we encourage you to study them.
Following the form of the previous discussion, describe why the curve in Figure 5.15b
appears as it does.
22
5.4.5 Quantitative Development of the Resistivity Pattern over a Horizontal Interface
There are a number of approaches to solving the quantitative problem of the single
horizontal interface. The simplest approach is one using an optical analog, which is
suitable for solving a number of the problems we are interested in. We will partially
develop the optical analog method so you can gain some insight into one quantitative
approach.
As a first step in the optical analog approach, we assume a planar boundary that
behaves as a semitransparent mirror. If we place a light source at point C1 and an
observation point at P1, part of the light will travel directly to P1 and part will be
reflected from the mirror (Figure 5.16). The total light reaching P1 will be the sum of
the two. The amount reflected depends on a property of the mirror referred to as the
reflection coefficient (k) and is equal to the reflection coefficient times the intensity of
the light source. In subsequent development of this approach it is more direct to
construct an image of point C1 on the other side of the mirror. This is the same
approach we used in deriving equations for reflected rays in Chapter 4. Recall that C1
and its image are equidistant from the boundary (mirror) and lie on a line that is
23
perpendicular to the boundary. Recall also that the distance r2 is equal to the reflected
path. If our observation point is at P2 rather than P1, the amount of light reaching P2 is
only that which is transmitted through the mirror. This amount is due to the intensity
of the light source, the transmission coefficient (1-k), and the distance of P2 from C1.
To develop the electrical case, we place a current source at C1 with intensity i and
develop equations for the potential at P1 and P2 following the optical analog as just
outlined. The planar boundary separates materials with resistivities ρ1 and ρ2. We use
a modified form of Equation 5.7 to determine the potential at P1 and P2. Equation 5.7
is valid for a current source at the surface because the equipotential surfaces are
hemispheres. In Figure 5.16 the current source is within the ρ1 material, and the
equipotential surfaces are spheres. For such a case the equation takes the form:
where D is the distance from the current electrode to the point where the potential is
being measured.
Remembering that the potential at P1 is due to a direct and a reflected component, we

write:
The potential at P2 (due to transmission) is:
The presence of the boundary constrains current flow across it. If P1 and P2 are both
located on the boundary at the same point (Figure 5.16b), then the two potentials must
be equal. In this case r1=r2=r3, so we set Equations 5.23 and 5.24 equal and simplify to
obtain the following:
This derivation is useful for two reasons. First, we now have sufficient information to
map the equipotentials for this two-medium boundary problem. Also, we know how
to determine the reflection coefficient. You may recall that this same factor appeared
in Equation 5.18 when we first began our qualitative consideration of the horizontal
interface problem. Although we accepted Equation 5.18 without explaining its
derivation, our present inquiry is beginning to shed light on various components of
that equation.
We will not pursue this case further. Instead we now turn to the actual problem at
hand, that of the horizontal interface. If you inspect Figure 5.17 and compare it to
Figure 5.16, you will see that we must add another surface to consider both the effect
of the earth’s surface and the buried interface. To derive the potential equations we
must consider both surfaces, but we proceed essentially as before.
24
A current source C1 of intensity i is placed within the ρ1 layer. The point P1 at which
we want to know the potential is placed a distance r from C1 at the same depth. The
placement of C1 and P1 is for convenience; they will be placed on the surface later in
the derivation. The upper boundary (the earth’s surface) separates ρ0 from ρ1. Our
object of interest, the horizontal interface, serves as the lower boundary and separates
ρ2 from ρ1. The reflection coefficient for the upper boundary is designated by k1,0 and
follows the form of Equation 5.25 (using ρ1 and ρ0), as does the reflection coefficient
25
for the lower boundary k1,2. Image sources of C1 are referred to as such with a number
in parentheses indicating the layer in which they reside.
Referring to Figure 5.17a, we proceed as in the previous example. The potential at P1

due to C1 is simply:
If we consider a single reflection from the upper boundary (equivalent to placing an

image of C1 in the upper material), the intensity of the current is i times the reflection
coefficient k1,0, and the contribution to the potential at P1 is:
The qualifiers associated with the potential symbol V signify that the image is in the
upper material and is the first image constructed there. It is demonstrated in Figure
5.17 that we also can have a single reflection from the lower boundary. This
contribution to the potential is:
Reverting to the optical analog for a moment, we can consider a ray emanating from
C1 reflecting from the lower boundary, reflecting from the upper boundary, and then
arriving at P1. As illustrated in Figure 5.17b, this is equivalent to placing a second
image in the upper layer (C1image(0)’), which is located at a distance 2z from C1.
Because this path involves reflections from both boundaries, both reflections
contribute to a reduction in intensity. Current density, therefore, is multiplied by the
product of both reflection coefficients. The equation for the potential due to the
second image in the upper material is:
We can continue this process and consider as a next step the path that is similar to that
in Figure 5.17b but reflects twice off the lower boundary and twice off the upper
boundary. The equation for this path is:
Note that now each reflection coefficient is squared. We can continue this process
producing more images in the upper medium that contributes less to the potential at
P1. If you examine Equations 5.29 and 5.30, you should note the similarity. Equations
incorporating more of this group of images would continue to have this form. This
suggests that these equations can be expressed as the infinite series:
Up to this we have four equations that represent contributions to the potential at P1

(Equations 5.26, 5.27. 5.28, and 5.31). In addition to these three other infinite series
must be included. In arriving at Equation 5.31 we considered only an even number of
reflections that could be related to an image in the upper medium. Other combinations
are (1) an odd number of reflections with images in the upper medium, (2) an even
26
number of reflections with images in the lower medium, and (3) an odd number of
reflections with images in the lower medium. The total potential at P1 is the sum of all
of these.
The next step is to place C1 and P1 on the surface. In this case, because air has infinite
resistivity, k1,0=1 and m=0. These substitutions are made in all the equations we
added together to determine the potential at P1. We rearrange these to arrive at:
However, we ultimately are interested in determining the potential difference due to a

source and a sink in a configuration such as that illustrated in Figure 5.11a. Equation
5.32 provides the potential at P1 due to the source at C1, which are a distance (a in
Figure 5.11a or r in Figure 5.17) apart. Deriving an equation for the potential at P2 due
to C1 is trivial because the only difference is the separation of the two points, which is
now 2a (Figure 5.11a). Thus the equation for the potential difference ΔV is:
Remember that our ultimate goal is to develop an equation similar to Equation 5.20
that provides values of apparent resistivity for various electrode spacings in the case
of the horizontal interface. If we use the electrode spacings in Figure 5.11a, we
rewrite Equation 5.20 as:
and
Equation 5.33 does not consider the contribution of the electrode at C2, but this
simply doubles the potential difference ΔV. We make this adjustment and then
rewrite Equation 5.35 using the expression for ΔV given in Equation 5.33 to obtain:
The form of Equation 5.36 is valid only for the electrode configuration in Figure
5.11a. Other electrode configurations require modifications of Equation 5.36. These
other configurations will be considered in a later section. It’s been a long path to
arrive at this point, and a number of steps have been neglected to keep the
27
development to a minimum. We hope this gives you sufficient insight into the optical
analog approach to obtaining a quantitative solution to the single interface problem.
Let’s place Equation 5.36 into a dynamic table (Table 5.4) to investigate the variation
of apparent resistivity with electrode spacing. This table is designed much like Table
5.3. The major variables are in the upper left corner. The result is apparent resistivity
and is located in the upper right. The components are:
28
Finally, we have a quantitative approach to determining the variation of apparent

resistivity with electrode spacing as defined in Figure 5.11a (all electrodes spaced at
equal intervals). If we select the subsurface model as it appears in Table 5.4 and vary
the electrode spacing, we can generate the curve in Figure 5.18b. The general form of
the curve should come as no surprise. It is similar to our qualitative prediction (Figure
5.15a). What may be a surprise, however, is the magnitude of the electrode spacing a
necessary for the apparent resistivity to approach the resistivity of the second layer.
29
As electrode spacing is increased, more current penetrates to a greater depth, so, as
you already know, we sample material at increasingly greater depths. However, the
shallow material always is included in the measurement. Therefore, for the measured
apparent resistivity to approach the resistivity of the second layer, a great volume of
the subsurface must be sampled. As you might imagine, this seriously limits the
resistivity method because it requires sufficient space in which to place electrodes.
5.4.6 Using RESIST

Although we could continue to use Table 5.4 and a graphing application to construct
apparent resistivity curves, it certainly is easier and more elegant to use a computer
program written especially for this purpose. RESIST is such a program. It accepts
resistivities and thicknesses for up to 10 layers, calculates apparent resistivities, and
plots apparent resistivity against electrode spacing. Henceforth, for convenience,
whenever we mention electrode spacing (or a-spacing as is sometimes the practice),
we mean the spacing between individual electrodes, which is identified by the letter a
in Figure 5.11a. Later in this chapter other variations will be discussed and labeled
differently.
Figure 5.18b illustrated a curve drawn by RESIST with the same input parameters
that appear in Table 5.4. RESIST uses a different mathematical approach for solving
the single horizontal interface problem, but the results are the same as those the
optical analog produces. Actually, the results of Equation 5.36 (which is used in Table
5.4) are plotted as small, filled circles in Figure 5.18b. The line passing through these
points is the RESIST curve from Figure 5.18c, which is overlaid on these data points
to emphasize that both approaches produce identical results.
Although we have not yet discussed procedures to interpret apparent resistivity curves
in terms of subsurface geology, it is not too early to begin to assemble some basic
properties of such curves. Apparent resistivity curves for the single horizontal
interface have one of the two forms presented in Figure 5.15. Clearly, the exact form
of the curve will depend on the resistivity values of the layers and the depth of the
interface. Before proceeding further (and without examining Figure 5.19), try to
sketch curves on the same set of axes for the following three subsurface models:
The important aspect of these curves is their similarities and differences, not the
absolute values of the apparent resistivities and a-spacing.
Now examine Figure 5.19, which was constructed by overlying the three curves
computed by RESIST for the three sets of values. You should be able to explain the
general shape and relationship of the curves from the basic principles we developed in
our discussion of the qualitative approach. All three curves have similar shapes, but
curves for the deeper interfaces are progressively offset to the right. This makes sense
if we remember the principle of current penetration, electrode spacing, and current
density as affected by poor conductors. The 5 m curve begins to rise earlier because
the effect of the shallow interface is felt even at small electrode spacings. In the case
of the 20 m curve an electrode spacing of almost 10 m (current electrode spacing of
30
30 m) is necessary to reveal the presence of a higher-resistivity layer at depth. The 5
m curve begins to flatten noticeably at an a-spacing in the vicinity of 100 m, but the
20 m curve still is rising steeply at this point. It should be obvious that a 5 m thick
layer constitutes only a small fraction of the volume of material included in a
measurement with an a-spacing of 200 m. Therefore, the apparent resistivity for this
a-spacing is quite close to the 200 ohm.m true resistivity of the deeper layer. This is
not true when the interface is at 20 m, and so the ρa value is only about 80 ohm.m.
Sketch a set of three curves following the layout of Figure 5.19. Use the same
interface depths but reverse the resistivities so the resistivity of the shallow layer is
200 ohm.m and that of the deeper layer is 10 ohm.m. Be sure you can defend the
relative positions of the curves.
Before we proceed to the next step, try to sketch three curves for which the interface
remains the same, and the resistivity of the shallow layer is constant at 10 ohm.m, but
the resistivity of the deeper layer takes on values of 100 ohm.m, 200 ohm.m, and 500
ohm.m.
As a final exercise before we continue our general analysis by considering the effect
of multiple interfaces, let’s investigate one further set of relationships for the single
interface case. In the preceding example resistivities remained constant and only the
depth of the interface varied. Now let’s hold constant the resistivity of the layer above
the interface and the interface depth while we vary the resistivity of the layer below
the interface. Figure 5.20 provides a comparison for five apparent resistivity curves
drawn under the stated constraints. Dashed lines are included in the Figure 5.20 to
facilitate a visual comparison of the latter portion of each curve and the true resistivity
31
of the deeper layer for which the curve was computed. The form and relationship of
the curves should come as no surprise. As the resistivity of the deeper layer increases,
larger electrode spacings are required before the slope of the curve begins to decrease
and approach the resistivity value of that deeper layer.
The form of the curves in Figures 5.19 and 5.20 suggests that we should be able to
devise a scheme to interpret apparent resistivity data in terms of resistivities and
interface depths. We soon will see that this indeed is the case; but before we consider
interpretations of field data, it seems prudent to consider more complicated models.
5.5 MULTIPLE HORIZONTAL INTERFACES

Based on our development of a quantitative solution to the single interface problem, it
should be evident that we could continue in this direction for investigating the effect
of more than one interface. However, the complexity of the image theory approach
becomes extreme, so that in practice more advanced mathematical solutions usually
are utilized. For our purposes it is sufficient to develop apparent resistivity curves
using RESIST and to qualitatively analyze these curves based on the principles we
discussed previously. To keep this discussion relatively brief, we consider only two
horizontal interfaces (three layers). You are encouraged to use RESIST to conduct
your own analysis of the three-interface (four layers) case.
Because we are dealing with three layer resistivities in the two-interface case, four
possible curve types exist: ρ1>ρ2>ρ3 (Q), ρ1>ρ2<ρ3 (H), ρ1<ρ2>ρ3 (K), and ρ1<ρ2<ρ3
(A). The letters used to refer to the various curves are those typically used in electrical
resistivity literature. We use them here only for convenience in referring to specific
curve types. The four curves are illustrated in Figure 5.21. Interface depths are the
same for all curves and are defined in Figure 5.21a. Resistivities used are 1, 10, and
100 ohm.m. The ρ1, ρ2, and ρ3 values for each curve are self-evident. Curves for
ρ1>ρ2>ρ3 (Q) and ρ1<ρ2<ρ3 (A) are contained in Figure 5.21b. These appear very
32
similar to two layer curves and easily could be misinterpreted as such, so it seems
sensible to inquire further about this similarity.
33
Figure 5.22 contains one of these three-layer curves (Q with ρ1=100 ohm.m, ρ2=10
ohm.m, ρ3=1 ohm.m, z1=10 m, and z2=20 m), which is drawn with the solid line. The
three dashed-line curves were computed using ρ1=100 ohm.m and ρ2=1 ohm.m and
horizontal interface depths of 10 m, 15 m, and 20 m. Although the two- and three-
layer curves are not the same, it could be virtually impossible to distinguish between
them with field data that are acquired at discrete intervals and almost always contain
some noise.
Figure 5.21c illustrates the remaining two curve types (H and K). These clearly
indicate the presence of a low-resistivity layer with a sequence of higher resistivities
or the presence of a high-resistivity layer lying between layers with lower resistivities.
The form of curve K can be explained using the principles we developed previously.
At small electrode spacings current density is affected only by the low resistivity (1
ohm.m) of the shallow layer, which leads to small values of apparent resistivity. As
electrode spacings increase, the higher resistivity (100 ohm.m) of layer 2 affects
current density, and apparent resistivities rise. As electrode spacing continue to
increase, a greater percentage of current flows in the deepest layer, which has a lower
resistivity (10 ohm.m) than the intermediate layer. This also affects current density
distributions so that apparent resistivities begin to decrease. Note that even though the
thickness of the second layer is substantial (20 m), apparent resistivities never exceed
13 ohm.m even though the resistivity of the layer is 100 ohm.m. Be sure you
understand why curve H in Figure 5.21c has the particular form that it does.
How many basic curve types exist for the four-layer case? How many of these are
unique in the sense that they clearly arise from four layers?
Although complexities and, therefore, identification and interpretation difficulties

increase as the number of interfaces increases, the basic principles determining
apparent resistivity values versus electrode spacing remain consistent. We will return
34
to this subject during our discussion of interpreting field values; but we turn now to a
different subsurface model, that of the vertical contact.
5.6 VERTICAL CONTACT

A rapid lateral change in resistivity can be modeled as a vertical contact. Such
geometry often is taken to represent a vertical fault, but sudden changes in
composition due to depositional processes or even a pronounced elevation gradient
(vertical cliff) also fit this model. The equations for this situation are no more
complicated than for multiple horizontal interfaces. The method of images works well
for the vertical contact. If you return to the section where we first considered the
quantitative approach to a single horizontal interface and rotate Figure 5.16 90o, you
should appreciate the similarity of the vertical contact problem to the single horizontal
interface problem.
Complications arise, however, as electrode spacing is increased (or the entire spread
is moved) because eventually one or more electrodes cross the contact (which is not a
consideration when the interface is horizontal). This changes the boundary conditions,
and a different potential equation must be applied. We need five equations to satisfy
all possible variations in electrode placements as illustrated in Figure 5.23a. We have
chosen not to derive or present the relevant equations here; rather, we concentrate on
the apparent resistivity patterns associated with measurements taken during traverses
that cross a vertical contact.
When discussing apparent resistivity patterns associated with horizontal interfaces,

we increased the electrode spacings to achieve increased current penetration and,
therefore, changes in the observed apparent resistivity. However, two different
approaches are feasible when a vertical contact is present. When dealing with such an
abrupt lateral change, we can detect the contact either by fixing the center of the
electrode spread and systematically increasing electrode spacing (expanding-spread
traversing) or by holding electrode spacing constant and systematically moving the
entire spread along a traverse line (constant-spread traversing). Expanding-spread
traversing also is referred to as depth profiling, electrical drilling, or resistivity
sounding, whereas constant-spread traversing often is called electrical mapping or
resistivity profiling.
5.6.1 Constant-Spread Traverse

Let’s first examine a constant-spread traverse across a vertical contact. Figure 5.23a
diagrams the essential elements of the traverse. The distance from the spread center to
the contact is x, which is positive to the right of the contact and negative to the left.
All electrode are equally spaced a distance a apart. The entire spread is moved at
equal increments along a line perpendicular to the strike of the contact. Table 5.5
contains results for a given selection of resistivities, electrode spacing, and traversing
increment. Casual examination of the apparent resistivities in the table reveals that the
apparent resistivities do not vary smoothly as was the case for horizontal interfaces. A
plot of Table 5.5 values (Figure 5.23c) confirms this impression. Actually, there are
four points on the curve where the value and direction of slope suddenly change.
In trying to develop a sense of why these slope reversals occur, it helps to plot
electrode positions relative to the contact on a map view. Figure 5.23b illustrates
several important arrangements. Note that the traverse typically is along a single line.
35
36
The positions are staggered in the diagram for clarity. Refer frequently to Figures
5.23a and 5.23b. As expected, when the electrode spread is to the far left (x=-5) or far
right (x=5), apparent resistivities approximate the actual resistivities. As the spread is
moved closer to the contact (x=-2), apparent resistivities decrease relatively quickly
because the spread is approaching a lower-resistivity medium. Slope reversals occur
at x=-1.5, -0.5, 0.5, and 1.5. These values represent spread positions where an
electrode crosses from one side of the contact to the other. Therefore, we should be
able to locate the contact quite closely by noting the positions of spread electrodes
when reversals take place. Remember that Figure 5.23 is idealized, and it is very
unlikely that field values of a and spread-center positions will locate the contact as
closely as in our model.
Why is the change in apparent resistivities much more pronounced between x=-3 to
x=-2 than from x=3 to x=2? Hint: Note that the negative x values are in the higher-
resistivity medium. Use the current density distribution model or a water flow analogy
to arrive at a qualitative explanation.
We strongly suggest that, if possible, you work with Table 5.5 to produce curves of
your own for various values of the quantities in boldface. For best results use x
increments that are smaller than the value you select for electrode spacing. The
equations used are not defined at |x|=a and |x|=3a/2. If you select values that result in
these equalities, the words ‘not defined’ will appear instead of an apparent resistivity.
To avoid this, we suggest selecting an a-spacing such as 2.01 instead of 2.00.
Of course, when we are working in the field, it would be extremely fortuitous for a
traverse to be oriented perpendicular to the trace of a vertical contact. Fortunately, the
main difference in an apparent resistivity curve for a traverse at angle to the contact is
a reduction in the size of the slope reversals. These slope reversal points, or cusps,
become less pronounced as the angle between the traverse and contact trace
decreases. If a traverse is parallel to the contact, apparent resistivity values remain
constant. Their magnitude depends on the distance of the traverse line from the
contact.
5.6.2 Expanding-Spread Traverse

In an expanding-spread traverse the electrode center is fixed and electrode spacings
are systematically increased. Similarly to the previous example, we link the values in
a dynamic table (Table 5.6) to a specific diagram (Figure 5.24). In this example the
spread center is at x=-2.0, and the traverse is perpendicular to the contact. The results
in Table 5.6 are plotted in Figure 5.24c, which reveals two sharp slope discontinuities.
Various expanding electrode spreads are illustrated in Figure 5.24b. Once again the
slope discontinuities can be correlated with electrodes crossing the contact.
Comparison of the a=1.0 and a=2.0 spread geometries suggests that the slope
discontinuity for an a-spacing of 1.25 correlates with a current electrode moving
across the vertical contact. For an a-spacing of 4.0 m, a potential electrode crosses the
contact that corresponds to the second slope discontinuity. Although perhaps not as
dramatic as the resistivity curve obtained by constant-spread traversing, the
expanding-spread curve certainly is sufficiently distinctive both to establish the
presence of a rapid lateral change in resistivity and to establish reasonably accurately
the location of the change.
37
38
A traverse oriented parallel to a vertical contact and located close to the contact
produces a curve that is remarkably similar to a two-layer curve. How can we explain
this behavior when a constant-spread traverse taken in the same orientation produces
constant apparent resistivity values? This seeming contradiction is easily explained if
we remember that equipotential surfaces in a homogeneous, isotropic material are
hemispherical. When the electrodes are close to the contact but are spaced very close
together relative to the contact distance, the effect of the material on the opposite side
of the contact will be minimal. The situation is analogous to that illustrated in Figure
5.14a. However, as electrode spacing is increased, a greater percentage of current
flow not only penetrates more deeply but spreads laterally as well. If a material with
different resistivity is encountered in this lateral direction, current density will be
affected. Thus apparent resistivity values will change also.
If we use the configuration in Figure 5.24a, the resistivity to the right of the contact is
low. Assume we place our electrodes as shown but rotate them 90o so they are parallel
to the contact. An initial reading for a small a-spacing would be close to 100 ohm.m.
As we increase electrode spacing, apparent resistivity will decrease as more current
flows in the lower-resistivity material and current density between the potential
electrodes decreases. This behavior continues as electrode spacing increases. The
extent to which apparent resistivity values approach the resistivity of the material
across the contact depends on the distance of the traverse from the contact. The final
curve for such a traverse looks much like that in Figure 5.15b.
5.7 TWO VERTICAL CONTACTS, HEMISPHERICAL STRUCTURES, AND

DIPPING INTERFACES
Unfortunately, in resistivity surveying it seems that all basic structural configurations
require derivation and application of separate equations. Our goal in this section is
simply to point out some of these other subsurface geometries and to emphasize that
the material you have just absorbed equips you to predict apparent resistivity curves
for these geometries, if only in a generalized way.
5.7.1 Two Vertical Contacts

First let’s briefly consider the structure illustrated in Figure 5.25a. This feature could
just as easily be termed a vertical dike, but we prefer to emphasize that it also is
adequately described as two vertical contacts. Use of the word dike conjures up
igneous terrain, but this feature could represent a rapid lateral variation in sediment
type or a saturated, high-porosity zone of brecciated rock. More important, however,
is the emphasis on the idea of two vertical contacts. Because we already are familiar
with the effects of one vertical contact, we might ask if we can predict the effect of
two vertical contacts on apparent resistivity curves. If the width of the feature
delimited by the vertical contacts is large compared to electrode spacing, then as we
traverse the contacts with a constant-spread electrode arrangement, we should (1) see
a curve with features similar to that in Figure 5.23c if we are proceeding from high-
resistivity material to low-resistivity material, and (2) observe a mirror image of the
first portion of the curve as we proceed from low-resistivity material to high-
resistivity material (Figure 5.25b).
Of course, this is much too simplified an approach to yield any quantitative results if
such a vertical feature actually is traversed. However, if at some time you observe a
curve with cusps that appear to be mirror images, you at least should recognize the
39
possibilities and then seek a more theoretical treatment in the electrical resistivity
literature.
Our point here is that resistivity curves come in a seemingly endless variety of shapes,
and you cannot begin to remember more than a few basic features. But if you grasp
the essential features and behavior of a few curve types (single horizontal interface,
multiple horizontal interfaces, vertical contact), you certainly will be more effective in
the field when required to make decisions concerning electrode spacing, traverse type,
and traverse direction.
5.7.2 Hemispherical Structures

A number of subsurface configurations are better approximated by a hemispherical
form than vertical contacts (Figure 5.26). An abandoned river channel filled with
gravel incised into floodplain deposits of silts and clays and a sinkhole in limestone
filled with dissimilar material are two common examples. Let’s assume we are
sufficiently fortunate to cross the center of such a form with a constant-spread
traverse and an a-spacing that is substantially smaller than the diameter of the
hemispherical form. Can you sketch a representative curve for such a traverse?
40
Actually, for rough approximation, we can think of the hemispherical form as two
vertical contacts that do not extend far downward. As we cross each contact, we
expect to see curves somewhat similar in form to vertical contact curves. However,
the effect of the ρ2 material is tempered by the presence of the ρ1 material beneath it.
Nevertheless, the curve shape is similar to that produced by two vertical contacts. If
the traverse cuts the hemispherical form but crosses nearer to its edge, the bilaterally
symmetric curve form still will be evident, but the cusps will be more rounded and
less pronounced.
5.7.3 Dipping Interfaces

Because of the number of permutations involved with traverses across dipping
interfaces, we will confine this brief discussion to one specific comparison.
Our standard will be the vertical contact (or vertical interface) curve in Figure 5.23c,
which is reproduced in Figure 5.27b. How will this curve change if we lower the dip
to 15o as diagrammed in Figure 5.27a?
If the traverse proceeds from left to right, apparent resistivity values approaching ρ1
(100 ohm.m) will be encountered first. As electrodes cross the contact separating ρ1
material from ρ2 material, we should expect similar behavior as observed for the
vertical contact.
However, due to the 15o dip, only a thin wedge of ρ2 material is present near the
contact. This will reduce the effect of this lower-resistivity material so that apparent
resistivity values will be higher relative to those for the vertical contact.
Also, as the electrode spread moves farther to the right away from the contact, the
decrease in apparent resistivity values will not occur as quickly because ρ1 material
still is not far from the surface.
If we sketch a qualitative representation of this behavior, our curve should appear

somewhat similar to that in Figure 5.27b. Cusps still are present due to electrodes
crossing a boundary between materials of different resistivities, but the total variation
in apparent resistivity from cusp to cusp is less than for the vertical contact. This may
cause greater difficulty in defining the cusp position or its existence and, therefore, in
locating the position of the contact and the amount of its dip.
41
5.8 FIELD PROCEDURES

Until now we have concentrated on developing a sufficient theoretical background for
understanding the shape of qualitative and quantitative apparent resistivity curves
connected to a variety of simple subsurface geometries. We have not as yet addressed
the question of proceeding from actual field data to an interpretation of the
subsurface. Before doing so we need to briefly address resistivity equipment, possible
electrode geometries to use in data collection, and some strategies for acquiring useful
resistivity data.
5.8.1 Equipment
Equipment for a typical resistivity survey consists of an ammeter, a voltmeter, a
power source, electrodes, and connecting wire (Figure 5.28). Electrodes are metal
stakes (copper, steel, or aluminum) with a cap to facilitate driving them into the
ground. Wire to connect the electrodes to the power source and meters must be
durable, light, of low resistance, and well insulated. The wire is placed on reels for
ease in winding and unwinding as electrode spreads are increased; a typical length on
an individual reel is rarely less than 100 m. Power is provided either from a motor-
driven generator or by batteries connected in a series. A number of considerations
dictate the exact source selected. Batteries are best for portability, whereas a generator
often is used when depths of investigation, and therefore current demands, are large.
42
The exact design of the meters and allied circuitry varies and depends on the type of
current used (direct current or alternating current) and also on portability
requirements. Direct current causes potentials to develop at the electrodes due to
electrochemical reactions between electrodes and soil. These potentials are included
in potential measurements unless special electrodes (porous pot) are used.
Commutators, which frequently reverse the polarity of the direct current, alleviate this
problem, but a common trend in modern instruments is to use low-frequency
alternating current.
The circuits to measure potential and current sometimes are in separate units, which
increase flexibility but reduces portability. A typical unit for standard shallow
exploration exercises includes not only the power source but both ammeter and
voltmeter. Obtaining a reading is straightforward with such instruments. Electrodes
43
are placed in a selected pattern and connected to the instrument. A few preliminary
adjustments are made as directed by the manufacturer, and current is then applied.
Typically a meter needle is deflected from a zero or null position. Rotating a dial (that
provides a direct readout in ohms) brings the needle back the zero position. The
ohmmeter reading is recorded, and electrodes are moved to the next position. Figure
5.29 pictures an instrument of this type. Readings can be taken quickly, but moving
electrodes and wire for the 10 or more reading required for any individual traverse is
time-consuming.
5.8.2 Electrode Configurations

The electrode patterns used in resistivity surveying almost always are the Wenner,
Schlumberger, or dipole-dipole. Many other arrangements have been tried, but these
three are the survivors. The Wenner electrode geometry is illustrated in Figure 5.30a.
Spacing between all electrodes is equal and conventionally is denoted by the letter a.
This configuration should appear familiar because it is the pattern we have used in all-
previous examples in this chapter. In conducting an expanding-spread Wenner survey,
we move all electrodes along a straight line after every reading so the spacing
between electrodes remains equal and takes on certain preselected values. Typically
we try to achieve values of a that are equally spaced on a logarithmic scale and ideally
number six points per decade (for example, 10.0, 14.7, 21.5, 31.6, 46.4, and 68.1). If
you used RESIST, you may have noted that these are the intervals for which apparent
resistivity values are calculated. RESIST, as is discussed later in this chapter, assumes
field readings are acquired at these same intervals. The main reason for such spacings
is that shallow materials always are included in any resistivity reading. Larger
electrode spacings are therefore necessary to acquire a reasonable electrical sample of
material at depth. Using the recommended electrode intervals produces equally
spaced points on a graph of the field data.
A symmetric distribution of current and potential electrodes about a central point also
is employed for the Schlumberger array (Figure 5.30c), but the potential electrodes
are spaced much more closely that the current electrodes. Spacings are selected to
maintain the relationship 2L>5MN and also follow the same numbering scheme as the
Wenner array. Because of this particular geometry, meter sensitivity is exceeded after
the current electrode spacing is increased several times. At this point the potential
electrode spacing MN is increased to the next larger value in the numbering scheme
(for example, 14.7 m if the previous value was 10.0 m) and the current electrode
spacing L is reduced by two intervals (such as to 100 m if the previous reading was
215 m). This recording procedure produces overlapping curve segments on a plot of
apparent resistivity versus electrode spacing (Figure 5.31).
The final common geometry is the dipole-dipole (Figure 5.30d). In this arrangement
the potential electrodes and current electrodes function independently. Both sets tend
to be fairly closely spaced with a significant distance between the sets. The potential
electrodes are placed relative to the current electrodes in one of several basic patterns
(not all along the same line). Because cable lengths between the electrodes are short,
it is much easier to place the potential electrodes at large distances from the current
electrodes, thereby facilitating deep investigations. At the same time larger currents
are necessary to reach these depths. This geometry is used much less in North
America than the Wenner and Schlumberger spreads, although use in the Soviet
Union, especially for petroleum exploration, has been significant.
44
45
The Wenner electrode array primarily has been used in North America surveys
because Frank Wenner employed that geometry early in the development of the
resistivity method. Most European workers use the Schlumberger spread, as do many
American investigators. Each geometry has advantages and disadvantages. One
advantage of the Wenner array is that the larger potential electrode spacing places less
demand on instrument sensitivity. A second advantage is the simplicity of the
apparent resistivity equation (Equation 5.20) when electrodes are equally spaced
(Equation 5.35). Prior to the present accessibility and capability of computing
facilities, this simplicity reduced the complexity of many equations and therefore
demanded much less effort in quantitative approaches to resistivity interpretation
(which is why we used it in previous sections). Frequent use led to many examples
and techniques in the literature, which furthered more use.
The Wenner geometry has several shortcomings. Because all electrodes must be
moved for each reading, which is not the case with the Schlumberger method, the
Wenner array requires more field time. Perhaps more serious is that it is more
sensitive to local, near-surface lateral variations than the Schlumberger geometry. Of
course, Wenner data can be analyzed directly; but Schlumberger requires some pre-
interpretation processing because the overlapping curve segments must be smoothed
and new values obtained from the smoother curve. On the other hand, the curve
segments follow a specific pattern of overlaps, which is clearly illustrated in Figure
5.31. Increasing MN spacing when apparent resistivity values are rising results in
offset of the new values downward (positions 1 and 2 in Figure 5.31). When apparent
resistivity values are decreasing, an increase in MN results in offset of the new values
upward (positions 3 and 4 in Figure 5.31). If the overlaps depart from this pattern,
lateral inhomogeneity is indicated and, if not too severe, can be corrected for in the
smoothing process.
The susceptibility of the Wenner electrode geometry to near-surface lateral variations

increases the likelihood that we may interpret these as variations at depth. One
approach to reducing the possibility of such an erroneous conclusion is to place a
third potential electrode at the center point of the electrode array. We take readings as
usual, but in addition to the normal potential difference reading at P1P2, we also take
readings for P1PL and PLP2. This approach usually is referred to as the Lee
modification. All apparent resistivities are calculated using the Wenner equation. If no
lateral variations are present, the Lee apparent resistivity curves should exhibit the
same form and should mimic the Wenner curve. The Lee values will plot at half-
values of ρa, which makes visual interpretation easier. Note: When using the Lee
partition array independently, change 2π to 4π in the Wenner equation (Equation
5.35). Departures from lateral homogeneity should be evident from departures in the
similarity of the Lee values. Once such variations are identified, we can make
appropriate adjustments to the exploration plan. Figure 5.32 illustrates a case where
the Lee modification values are very similar and almost perfectly mimic the Wenner
apparent resistivity values. Hence lateral variations are not significant.
Because quantitative interpretations of Wenner and Schlumberger curves are very

similar, we generally will continue to provide examples based on the Wenner array.
RESIST provides options to use either array, so you may select either depending on
the category of field data you have available. If you gather data on your own, we
46
suggest you take the time to use both arrays to better judge their strengths and
weaknesses.
5.8.3 Surveying Strategies

As in the case of seismic exploration, resistivity survey goals should be specifically
defined before fieldwork begins. Equipment must be selected that is adequate for the
task at hand. And, if at all possible, independent geologic controls should be
available. It is difficult to derive detailed geologic information from resistivity
surveys unless such control exists. The next step is to locate sampling sites and to
decide whether to use constant-spread or expanding-spread traverses.
If the goal of the effort is primarily to seek information about apparent resistivity
variations with depth to define the location of a high-resistivity-contrast interface,
then expanding-spread traverses are the obvious choice. An example is a survey
interested in mapping depth to a bedrock surface overlain by saturated silty sands. Of
course, either Wenner or Schlumberger can be used. In either case, you always should
run another traverse at right angles to the first to establish whether lateral variations
are present. Remember that an expanding-spread traverse oriented parallel to a
vertical contact can produce a curve almost identical to that for a two-layer case,
whereas a traverse perpendicular to the contact reveals cusplike discontinuities that
establish the presence of the contact without doubt.
If survey goals are to map lateral variations in resistivity, perhaps due to the presence
of a gravel bar, then constant-spread traversing is required. Normal practice dictates
that two or three electrode spacings are used at each station to develop three-
dimensional information. Apparent resistivities for each electrode spacing are placed
at each array center and contoured. If three spacings are used at each site, three
separate contour maps are produced. Results of constant-spread surveys are difficult
to interpret quantitatively, especially without some vertical control. For this reason at
least one (and ideally more) fully implemented expanding-spread survey is included
to develop a sense of resistivity variation with depth and to provide a basis for
quantitative analysis. If you have a good grasp of the principles covered to this point
and well-defined survey goals, traverse design should be reasonably straightforward.
A major question that often arises is the largest electrode spacing to use. In other
words, what is the largest spacing that will assure adequate sampling of the
subsurface at the depth of interest? In the past many explorationists incorrectly
equated electrode spacing directly with depth. Even at this stage you should realize
that such a relationship is not correct. This topic is important, so we return briefly to
some diagrams discussed previously to review several relationships. First let’s
reexamine Figure 5.11. This graph demonstrates, for the case of a single horizontal
interface, that the amount of current penetrating beneath the interface (and therefore
providing a sufficient sample of the lower material) depends not only on electrode
spacing but also on the relative values of ρ1 and ρ2. In cases where the depth of the
interface is equal to electrode spacing (z=a) and ignoring extreme resistivity contrasts,
the percentage of current penetrating beneath the interface can vary from 20 to 80
percent.
Similar lessons are available from Figures 5.19 and 5.20. Figure 5.19 demonstrates
that increased electrode spacings are required to completely define curves for
47
interfaces at increasing depths. Resistivity contrast is held constant for each curve.
Graphs such as this suggest that three decades of readings (from 1 m to 1000 m) are
sufficient to define curves for most shallow interfaces. Now study Figure 5.20.
Interface depth is constant for each curve. At an a-spacing of 100 m the curve for
ρ2=20 ohm.m is completely defined. At this same spacing, however, the curves for
ρ2=200 ohm.m and 500 ohm.m are barely distinguishable and still are rising at such a
rate that they offer no insight into the resistivity value of the lower layer except that it
is high. However, many experienced explorationists find the rule that a-spacing is
equal to depth to be useful in a variety of geologic settings. It is perfectly acceptable
to use this rule to gain a feel for the meaning of curve form during fieldwork as long
as you remember this approximation is imperfect and eventually subject your field
data to detailed computer analysis.
The best advice we can offer with regard to electrode spacing decisions is to plot
apparent resistivity values as readings are taken. Unless you are positive that the form
of the curve is sufficient to meet the goals of your survey, keep acquiring data until
you reach the limits of your instrument, cable length, or available room. Of course, if
many traverses are required, then time end expense may constrain the number of
readings you can accumulate. More often that not, we find that we run out of room or
cable before obtaining as many readings as we would like.
5.8.4 Other Considerations

A number of factors can affect resistivity measurements in addition to the resistivities
of the rocks and sediments beneath the surface, and you should be aware of the most
significant of these to aid in field survey design. Some factors are obvious. Buried
pipelines or cables may be good conductors and, if so, will affect current flow.
Electrode contact must be good, and resistances should be low and uniformly
maintained. If dry surface materials are present, wetting the ground around the
electrode often is all that is required.
Topography affects apparent resistivity measurements because it affects current flow.

Perhaps the most obvious example of this is the vertical contact. Check Figure 5.23
and replace the ρ1 material with air, which we assume to have infinite resistivity. This
provides the example of the vertical cliff, which admittedly is an extreme case, but
which illustrates that current flow would be affected and that apparent resistivities
measured near the cliff would also be affected. Slight undulations and elevation
changes are not a problem, but rugged topography should be avoided if at all possible.
No topographic corrections are available as there were for seismic work.
Finally, many rocks and some sediments often possess different resistivities
depending on the measurement direction. An obvious example is a rock unit
containing a single, very well oriented fracture set filled with water. Traverses parallel
to the strike of the fractures will measure lower resistivities than traverses normal to
the fractures. Although this electrical anisotropy is not as serious an interpretation
problem as buried objects and rugged topography, you should be aware that it can
affect apparent resistivities. You generally attempt to collect data along two
perpendicular traverses. Variations between curves for the two traverses normally
reflect lateral variations in materials; but keep in mind that anisotropy in a given unit
may be responsible in some cases.
48
5.9 QUANTITATIVE INTERPRETATION OF APPARENT RESISTIVITY
CURVES
An exclusively quantitative interpretation of apparent resistivity curves often is
difficult. This arises from the wide variations in resistivity possessed by geologic
materials and the difficulty in developing theoretical expressions for apparent
resistivities of all but the simplest geometries. Quantitative interpretation is best
developed for cases in which layering is horizontal, so it is to these situations that we
will confine our analysis. First, however, we consider the resistivities of sediments
and rocks.
5.9.1 Electrical Resistivities of Geologic Materials

In the shallow subsurface electricity is conducted almost entirely by the fluid present.
Thus the resistivities of sediments and rocks in this environment are controlled by the
amount of water present and its salinity. Because this conduction is electrolytic and is
related primarily to ions moving through the fluid, the cation exchange capacity of
clay minerals increases conductivity. Although clay minerals are the most important
in this regard, all fine-grained minerals possess an exchange capacity to some degree.
This means that increasing silt or clay content in poorly sorted rocks or sediments will
reduce resistivities. In saturated materials increasing porosity will reduce resistivities.
Well-sorted materials have greater porosities than poorly sorted ones, but in well-
sorted rocks or sediments with well-rounded grains, grain size alone does not
influence porosity (Fetter, 1980, 60-62). In addition, irregularly shaped grains will
pack less tightly than sphere-shaped grains and consequently will have higher
porosities. In general, finer-grained sediments will have higher porosities (and lower
resistivities) than coarse-grained sediments. Based on these comments, however, it
seems prudent to apply this relationship only in the most general way.
All the following reduce resistivities: increasing water content, increasing salinity of
water, increasing clay content, and decreasing grain size. Assuming that water is
available to fill voids, resistivity is lowered by increasing porosity, increasing number
of fractures, and increasing weathering. Conversely, resistivities are raised by
increasing compaction and lithification.
Because these factors vary so much in the natural environment, it is not surprising that
resistivities vary greatly and that it will be difficult to correlate resistivities with
source materials in the absence of other geologic information. Indeed, Zohdy, Eaton,
and Mabey (1974, 9) note, ’No other physical property of naturally occurring rocks
or soils displays such a wide range of values’. Although direct correlation will not be
as straightforward as it was with seismic velocities, some generalized relationships
often may be established. Bedrock almost always has higher resistivities than the
saturated sediments lying above. Generally, the unsaturated sediments above the
water table will have higher resistivities than the saturated sediments below the water
table. However, the only reliable way to correlate resistivities with local geology is by
using independent geologic information (such as an expanding-spread traverse located
near a carefully logged drill hole). In this way you can establish correlations and carry
them forward into unknown terrain.
The list of resistivities shown here is based on personal experience and is adjusted to
include values provided by numerous sources. These values are supplied primarily to
provide a starting point for interpretation, but beware. For every observation that falls
49
within the values on this list, there probably is an observation that provides an
exception.
A thorough discussion of factors controlling resistivities and a good presentation of

resistivity ranges for common geologic materials is presented in Ward (1990, 148-51).
5.9.2 Empirical Methods

At least two empirical methods are cited frequently in texts dealing with
interpretations of expanding-spread apparent resistivity curves. These are the Moore
(1945) cumulative resistivity method and the Barnes (1952) layer method. Neither
method should be used except for the most preliminary analysis or when the
analytical method described in the following sections cannot be utilized because the
required materials are not available.
When using the Moore method, we collect apparent resistivity data at electrode
spacings that are evenly spaced on a linear scale using a Wenner spread. For each a-
spacing a cumulative resistivity value is calculated that is the sum of all apparent
resistivities collected at all previous electrode spacings. These cumulative resistivities
are plotted on the vertical axis; a-spacings are plotted on the horizontal axis. Straight
lines are fit to the data points. The horizontal coordinates of the points of intersection
of these lines are assumed to give the depths to horizontal interfaces.
The Moore method has succeeded in some locales and failed in others. It is possible
to check this approach by using a program such as RESIST to generate apparent
resistivity values for known models and then to interpret the data using the Moore
method. If curiosity leads you to such an investigation, note that RESIST plots data
that are equally spaced on a logarithmic scale, whereas the Moore method requires
data equally spaced on a linear scale. Zohdy, Eaton, and Mabey (1974, 46) report that
the method appears to work well for two-layer cases where the resistivity contrast is
not too great. If this condition does not hold, the depth to the interface can be in error
by as much as 50 percent. These investigators regard the application of the Moore
method to data from three or more layers as highly questionable. In addition, the
method does not provide resistivities for the derived layering.
You realize that each time electrode-spacing is increased; the volume of earth is
increased that affects an apparent resistivity determination. A basic premise of
Barnes’s method is that each measured apparent resistivity represents the average
resistivity in a layer that extends from the surface to a depth equal to the electrode
spacing. This premise leads to the following: when the electrode spacing is increased
in equal intervals, the depth representing the average resistivity also increases in equal
50
intervals. This assumption is not accurate, as you should realize at this point. Once
again, you could test this method by generating data using RESIST, plotting and
interpreting them using the Barnes procedure, and comparing the results with your
initial model. Like the Moore method, the Barnes method works for certain
subsurface configurations but not for others.
Both methods function best when data are collected at equal electrode-spacing
intervals on a linear scale. The analytical methods to be discussed next assume that
data are collected at electrode-spacing intervals that are equally spaced on a
logarithmic scale. Because these analytical methods are preferred for quantitative
interpretation, using the Moore or Barnes method will require additional field time to
collect data or to interpolate existing data.
5.9.3 Analytical Methods – Curve Matching

In discussing curve-matching techniques, we restrict examples to the case of the
single horizontal interface and an expanding-spread traverse. However, these
techniques can be extended to any situation for which the appropriate equations exist
(such as Equation 5.36). Curve-matching techniques as described here essentially
have been supplanted by the computer techniques that are described in the following
section. Nevertheless, we believe that understanding this approach and working with
and observing many different types of curves provide useful exposure and practice.
Some authors recommend curve matching as a preliminary step before using the
computer.
Consider the following. If we have a field curve for which apparent resistivity is
plotted against electrode spacing and can match this curve with a curve calculated by
RESIST for a given model data set, we then have a solution to our field data. The
actual subsurface geology must be very close to the model on which the theoretical
curve is based. Therefore, given a field curve, all we have to do is find a similar
theoretical curve and a solution is at hand. Unfortunately, because we are working
with three variables (ρ1, ρ2, and the depth of the interface z1), there exist an infinite
number of combinations that create an infinite number of curves through which to
search.
However, there is an elegant but simple solution to this problem. If we choose a

model and use RESIST to calculate apparent resistivity data and then divide ρa values
by ρ1 and electrode-spacing values by z1, we create a curve such as that labeled as k =
-0.4 in Figure 5.33a. If we now choose a new model with different values of ρ1, ρ2,
and z1, but constrain the resistivity values so we maintain the condition k = -0.4, our
new curve plots in the same position as our first curve! In fact, any two-layer curve
for which k = -0.4 plots at this same position. Therefore, we can create a family of
curves, termed master curves, for the two-layer case. Four of many possible master
curves are illustrated in Figure 5.33a.
Once master curves are available, they can be used to analyze field data. Field data,
such as those in Figure 5.33b, are plotted on transparent paper at the same scale as the
master curves. The transparency then is placed over the master curves and moved
about, keeping the graph axes parallel, until the field data lie on one of the master
curves (Figure 5.33c). The lines labeled ‘resistivity index’ and ‘depth index’ are
located on the master curve sheet, and their intersections with the field data graph
51
axes are noted. These intersections provide values for ρ1 and z1 (see Figure 5.33c). We
also note the k-value (see equation 5.25) of the curve that best fits the field data. In
cases where a field curve does not lie exactly on a master curve representing an
integer k-value, the k-value is interpolated. Because we now know ρ1 and because:
it is straightforward to determine ρ2. Figure 5.33c provides values of ρ1 = 100 ohm.m,

k = -0.4, and z1 = 10 m. Equation 5.37 then provides a value of 42.86 ohm.m for ρ2.
52
Generating master curves for three, four, or five layers basically follows the same
procedure but is more complicated. Many sets of curves are produced, and more
searching is required to find proper fits between master curves and field data. Several
sets of master curves have been published and are available at libraries for those
interested in working with this approach.
Two specific problems occur when we try to interpret field data based on comparisons
with theoretical curves representing three or more layers. These problems are known
as equivalence and suppression. Equivalence essentially means that various
subsurface models can generate curves that are very similar (equivalent). In other
words, a given interpretation is not necessarily unique. You already observed one case
of equivalence in Figure 5.22, where a three-layer curve is similar to a two-layer
curve. Another case is documented in Figure 5.34a, in which two three-layer models
generate similar curves. Given the noise inherent in field data, one of these likely
could not be distinguished from the other.
Suppression occurs when the presence of a thin layer in a multilayer sequence cannot
be recognized on an apparent resistivity curve. Suppression is controlled by the
relative thickness of a layer (thickness of the layer ÷ depth of the layer) and the
resistivity contrast between the thin layer and adjacent layers. Figure 5.34b illustrates
curves produced by a three-layer model and a five-layer model that are very similar.
Your major observation should center on the form of the Model 1 curve. Even though
the model consists of five layers, the curve essentially is identical in form to a
common type of three-layer curve.
Note that the problems of equivalence and suppression also are present when we fit
model curves to field data by computer. We now turn to this approach.
53
5.9.4 Analytical Methods – Automated Curve Matching

As we noted at the beginning of our discussion of curve matching, computer analysis
of field curves largely has supplanted the curve-matching technique. In performing a
computer-based analysis we enter field observations and an initial model that we
believe is responsible for the observed values. The initial model should produce a
curve of the same type and as close as possible to the observed curve. Hence, if we
are inexperienced in interpreting resistivity curves, a preliminary step using curve
matching often saves computer time and results in more realistic solutions.
54
After we enter field data and an initial model, the computer program calculates an
apparent resistivity curve based on the model and compares this curve to the observed
curve. Depending on the variation between the two, adjustments are made to the
model according to the mathematical procedure used in the computer algorithm, and a
new model curve is calculated. This process continues until both curves are as similar
in shape as specified by the person guiding the analysis or until the computer program
is unable to make the curves any more similar. Based on the results of the automated
interpretation, we may accept the final model or initiate another analysis by
specifying a new model. The program RESIST that is included with this text follows
this general procedure. It performs analysis for either Wenner or Schlumberger data.
It is possible to fix certain model parameters so they cannot be changed by the
computer program during the analytical procedure. Full details about RESIST are in
Appendix C.
A slightly different computer-based approach is suggested by Zohdy (1989). Zohdy

notes that apparent resistivity curves tend to be out of phase with true resistivity-depth
curves and that the values of observed apparent resistivities are always less than or
equal to the actual resistivities. These and other properties of apparent resistivity
curves for horizontally stratified materials lead to the following iterative approach.
The observed apparent resistivities and electrode spacings are used as the beginning
model (that is, the apparent resistivities are treated as true resistivities and the
electrode spacings are used as depth). Of course we realize this is not the case, but we
make this assumption to generate an initial curve. A theoretical curve is computed for
this model. The theoretical curve is compared to the observed curve. If the agreement
is not good (say within 2 percent), model depths are uniformly decreased by 10
percent, and a new theoretical curve is calculated. This continues until good
agreement is reached. Resistivities used in the models then are adjusted by a similar
interactive procedure until observed and computed curves agree to within set
conditions. The final model represents a solution to the observed data.
Doubtless other computer-based procedures will appear. Because of equivalence and

suppression and because of the many factors that affect curve details, a thorough
knowledge of resistivity principles and as much experience as possible must go hand-
in-hand with computer-derived solutions for field measurements.
5.10 APLICATIONS OF ELECTRICAL RESISTIVITY SURVEYING

In terms of the material discussed until this point, you may wonder if electrical
resistivity surveying is viable tool for exploration. Although resistivity interpretation
is not particularly straightforward, especially when compared to the techniques
presented for seismic work, resistivity surveying can be extremely valuable. In some
cases it is possible to determine reasonably accurate depths to bedrock using
resistivity. This may be a more cost-efficient procedure than investing in relatively
expensive seismic equipment. However, resistivity applications most often are used
when other geophysical methods simply cannot supply the desired information. We
emphasize these cases in the following descriptions.
5.10.1 Applications Related to Aquifers

Many communities that depend on confined aquifers for their water supplies
obviously want to prevent contamination of this supply or to have some idea of the
extent of the aquifer for planning purposes. In many cases a buried aquifer (sand and
55
gravel) is overlain by a clay layer that prevents most infiltration of solutions from
above. Therefore any information about the lateral distribution of the clay-gravel
contact, its depth, and the clay thickness is valuable. Refraction seismic methods often
cannot map the contact because of a velocity inversion, and reflection methods, if
they can detect the contact, may be too expensive or elaborate for use.
In Chapter 2 we introduced the Whately study and noted that resistivity methods also
were used. In this case resistivity was effective in mapping bedrock depths, which
tended to correlate with the location of a buried aquifer. Figure 5.35a illustrates a
typical resistivity curve from the Whately area. The steep downward segment of the
curve is due to a thick clay layer (approximately 50m), and the last upward segment
reveals the presence of bedrock. Analysis of the curve yields a bedrock depth of 71 m,
which is abnormally deep for the local area, but a well less than 1 km away along the
trend of the buried aquifer gives a bedrock depth of 75 m. Although the gravel aquifer
cannot be detected on the curve, the bedrock depth suggests it should be present.
Easthampton, Massachusetts, also depends for most of its water supply on a confined
aquifer. The aquifer is a glacial sand and gravel deposit that rests for the most part on
Triassic sedimentary rocks. The sand and gravel deposit is overlain by glacial lake
clays that vary in thickness, thinning to zero in the recharge area where the aquifer is
exposed at the surface. In 1981 a group of Smith College students and faculty
conducted a study, supported by funds from the Shell Oil Company Foundation, to
outline the thickness and extent of the aquifer. The study used both seismic refraction
and resistivity methods and accomplished its objectives. Figure 3.35b is an
expanding-spread profile from this study. This particular profile yields a clay
thickness of 31 m, which, when compiled with many other such determinations,
produced a contour map for clay thickness and distribution.
In some situation it might be of interest to map the depth of the water table. If the
surface deposits are homogeneous, then the lower resistivity of the water–saturated
material will impart a sufficiently great resistivity contrast for analysis. However, if
the near-surface materials are not homogeneous, it will then be difficult or impossible
to map the water table with any precision. Moreover, the extent to which the capillary
fringe is developed also will affect how closely the actual depth of the water table can
be determined.
Resistivity surveys have been used successfully at a number of sites to map buried
stream channels. Most of these surveys utilized both constant-spread and expanding
spread traversing to gather sufficient information to locate a buried stream channel.
Shallow reflection work might be able to produce superior results but would entail at
least equivalent field time as well as more expensive equipment. Figure 5.36a is
apparent resistivity map from a study in the Penitencia, California, area, which is
northeast of San Jose (Zohdy, Eaton, and Mabey 1974, 47-50). This map was
produced from many constant-spread traverses using an a-spacing of 6.1 m. Contours
of equal apparent resistivity values define a zone of high resistivity oriented
approximately east-west. A geologic section (Figure 5.36b) based on four expanding-
spread traverses, the apparent resistivity profile along the section line BA, and three
boreholes reveals that the high-resistivity trend on the apparent resistivity map is a
zone of gravel and boulders that defines the location of a buried stream channel.
56
57
5.10.2 Applications Related to Contamination
Some of the most important successes of the resistivity method are based on
contamination of normal groundwater. Usually contaminated groundwater, whether
affected by leachate from a landfill or by saltwater intrusion, has greater conductivity
than ordinary groundwater. Therefore, materials containing the contaminated water
will posses lower resistivities than materials containing unaffected groundwater. If the
water table is fairly shallow and the subsurface is relatively homogeneous, it is
possible to map the extent of the contamination.
Ideally, traverses are implemented before contamination occurs. Such data document
resistivity variations due to lateral and vertical variations in materials. Deviations
from these values, based on traverses taken after contamination, provide a good basis
for delineating the horizontal and vertical extent of the problem. Figure 5.37
summarizes how resistivity can define the leachate plume associated with
contaminated waters leaking from landfills. It is based on patterns observed in studies
of several landfills located at various sites throughout the United States. The
resistivity contours are determined from variations in apparent resistivities compared
to normal values associated with the landfill site.
5.10.3 Applications in Mapping Karst and Geologic Structures

Hubbert (1932 and 1944) published two classic studies demonstrating the application
of constant-spread traversing in defining karst relationships and mapping fault
locations. Figure 5.38 illustrates a constant-spread profile over karst topography in
Hardin County, Illinois (Hubbert 1944). Low apparent resistivity values are
associated with clay at the surface, whereas somewhat higher apparent resistivity
values are associated with masses of limestone at shallow depths that are overlain by
clay. The highest apparent resistivity values correlate with limestone at the surface or
large voids (caverns), which have infinite resistivity.
Hubbert (1932) also located faults between blocks of sandstone and limestone and a
shear zone in sandstone by a constant-spread traverse (Figure 5.39). Note the
similarity of the apparent resistivity curves in the vicinity of faults to the theoretical
curves of vertical contacts (Figure 5.23 and Figure 5.25). If the 30 m station interval
was reduced, the increased station density likely would result in curves with even
more similarities to theoretical curves. Both of Hubbert’s studies confirm that, given
sufficient resistivity contrast and relatively straightforward geology, constant-spread
traversing is capable of locating vertical geologic contacts and similar features of
interest.
5.10.4 Other Applications

Resistivity surveys have been used to outline new geothermal fields in areas of known
geothermal potential. Geothermal waters are better conductors than normal waters due
to their greater content of dissolved salts.
Because of the typically high resistivity contrast between rock and sediment,
resistivity surveys often can delineate buried stone walls and foundations. Surveys
also are sensitive to differences in moisture content of materials, so they often can
differentiate between excavated and unexcavated ground. For both these reasons
resistivity has been a frequent tool in archaeological studies.
58
59
The complexities inherent in resistivity interpretations and the nonuniqueness of
potential solutions to observed data have tended to obscure the value of resistivity
surveys. Although the method does have these drawbacks, when combined with
seismic surveys or the other electrical method detailed next, it can be an effective
exploration tool. However, when we use resistivity to explore subsurface geologic
relationships, independent control from well logs is mandatory. Its capability as an
inexpensive monitoring tool has not been developed as fully as possible, but this
likely will change as our society faces more contamination and needs to locate and
remedy the problem.
5.11 OTHER ELECTRICAL METHODS

As noted at the beginning of this chapter, the following descriptions of additional
electrical and electromagnetic methods are purposely brief. We include them mainly
to provide a sense of applications, limitations, advantages, and how they may be
similar to or different from electrical resistivity methods.
5.11.1 Induced Potential IP

Explorationists initially involved with the development and application of electrical
resistivity surveying noticed that the potential difference does not immediately
decline to zero after current is turned off. Rather, a large percentage of the potential
disappears immediately, but the remainder gradually decreases over an interval,
usually several seconds. Subsequent work demonstrated that certain subsurface
materials are capable of electrical polarization and function like a rechargeable
battery. Applied current induces a buildup of charges that, when the current is
removed, gradually discharge, producing a flow of current and potential differences.
This decay of potential with time is referred to as induced potential in the time
domain and is the phenomenon we discuss here. We also can investigate the effect of
different frequencies of alternating current on apparent resistivity values. This
approach is termed IP in the frequency domain. A common measure of the effect of
IP in the time domain is chargeability, which is defined as the ratio of the area under
the decay curve to the potential difference measured before the current was turned off.
It is possible to demonstrate a theoretical relationship between apparent resistivity and
chargeability. This is extremely valuable. It permits us to derive quantitative
relationships between subsurface geometries and apparent chargeabilities if we know
the appropriate equations for apparent resistivity, such as those that were presented
for the horizontal interface.
Because of this specific relation between apparent resistivity and apparent

chargeability, field procedures and data interpretation techniques are similar.
Electrode arrays typically are the Schlumberger or dipole-dipole, and apparent
chargeability is plotted against electrode spacing. Curve matching is commonly
employed. Historically, IP has found much more application in exploration for base
metals rather than in typical electrical resistivity applications such as groundwater
studies. One reason is that equipment tends to be bulkier and data acquisition slower.
Fieldwork is more expensive, and considering the relative lack of interpretive
techniques, IP tends not to be the cost effective unless the economic potential is
substantial.
60
Equipment for conducting induced potential studies is similar to electrical resistivity
equipment with two major exceptions. A timing circuit is required to switch the
current on and off, and a recording unit is necessary to preserve the decrease of
potential with time. Standard electrical resistivity equipment cannot, therefore, be
used for IP studies; but equipment designed for IP work can be used to measure
apparent resistivities.
Induced potential studies can be valuable assets to electrical resistivity studies. Based
on our discussion of interpretive procedures for resistivity, it is apparent that curves
that look like a single horizontal interface actually may contain the signature of
several layers but cannot be differentiated. One or more of these hidden layers may
possess a different chargeability. If apparent resistivity and apparent chargeability
curves plotted on the same graph, such distinctions may permits more informed, and
more correct, interpretations of the subsurface.
5.11.2 Spontaneous Potential SP

Natural potential differences at the surface may arise from electrochemical reactions
at depth that cause currents to flow. The most common example of this type of
phenomenon is associated with ore bodies, part of which are above the water table.
Oxidation takes place above the water table, whereas reduction takes place below it.
These processes result in different charge concentrations that cause current flow.
Sulfide and graphite bodies are by far the most common source of significant
potentials of this sort, so it is not surprising that most SP surveys are conducted to
locate ore bodies, especially sulfide ones.
Equipment demands are fairly straightforward. Two electrodes, wire, and a

millivoltmeter constitute the normal setup. In this case porous-pot electrodes must be
used because metal electrodes themselves produce a self-potential. A porous-pot
electrode consists of a ceramic container with holes. A copper tube is surrounded by a
solution of copper sulfate that slowly leaks through the holes into the ground. This
solution disseminates the charges that otherwise would build up if the ground was in
direct contact with a copper electrode.
Several characteristics conspire to negate the possibility of wide application of the

self-potential method. Although surveys are quick and efficient and equipment
demands are minimal, quantitative interpretation is difficult and depths of sources are
difficult to model. Self-potentials are easily masked by other sources such as telluric
currents. Because the water table is an important component of most sources, its
typically shallow depth limits surveys to the same shallow terrain. And finally, of
course, sources are limited to the specific type mentioned. SP exploration, therefore,
is not a common shallow method except where we desire a relatively fast, inexpensive
tool in searching for shallow sulfide bodies.
5.11.3 Telluric and Magnetotelluric Methods

Natural electric currents flow in the earth that are induced by the flow of charged
particles in the ionosphere. At any given moment these currents follow uniform
patterns over large areas. However, currents are flowing at depth, and based on what
you know from this Chapter, it seems reasonable to expect these currents to describe
patterns based on the conductivity and geometry of the materials through which they
flow. Naturally, we measure the potential differences produced by this current flow.
61
Because telluric currents vary with time, field operations include a base station at
which two pairs of electrodes are places at right angles. Voltage is continuously
monitored, and we move two additional pairs of electrodes along a grid. Comparison
between the base station readings and readings taken at grid locations yields a
potential difference that is due to the subsurface effect on current density.
Magnetotelluric methods combine the measurement of magnetic field intensity

variations with voltage measurements. Both sets of data are used to compute apparent
resistivity functions, which then are compared with theoretical curves to determine
thicknesses and true resistivities. A major requirement of this approach is a
magnetometer capable of measuring weak magnetic field variations that fluctuate
frequently.
Most surveys of this type have been in Europe and the Soviet Union. Survey
objectives typically are directed at more regional objectives than we have mentioned
to date, such as determining the geometry of sedimentary rock sequences in major
depositional basins.
62

Chapter 5

Uploaded by

Copyright:

Available Formats

Chapter 5

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Chapter 5

Uploaded by

Copyright:

Available Formats

Applied Geophysics: Exploring The Shallow Subsurface By Burger, Sheehan & Jones

Prepared By A. Fadil For FSSM UCA Graduate & Undergraduate Students

5.1.1 Applied Currents

A third method using applied currents is electromagnetic (EM) surveying. In EM

5.1.2 Natural Currents

The self-potential or spontaneous potential (SP) approach uses natural

5.1.3 A Brief History

5.1.4 Chapter Goals

5.2 BASIC ELECTRICITY

where i is current in amperes, q is charge in coulombs, and t is time in seconds.

Another important concept in electrical resistivity surveying is the current density j.

Because current is defined as the movement of charges across a given cross-sectional

Because various geologic materials can be expected to have different resistances to

5.3 CURRENT FLOW IN A HOMOGENEOUS ISOTROPIC EARTH

Because current distribution is equal everywhere on this surface that is at a distance r

Equation 5.7 is the fundamental equation in our electrical prospecting discussion.

5.3.2 Two Current Electrodes

Examining these relationships allows us to arrive at certain conclusions. Fifty percent

5.3.3 Two Potential Electrodes

Therefore, the potential difference ΔV equals:

Because resistivity is the unknown quantity we normally hope to determine, we solve

Perhaps we should test our understanding of Equation 5.15 by applying it to a known

These calculations confirm that if we produce a current, measure the electrode

5.4 A SINGLE HORIZONTAL INTERFACE

5.4.1 Current Distribution

Equation 5.18 is sufficiently straightforward to permit entry into a spreadsheet for

In our previous discussion of current flow in a homogeneous, isotropic subsurface we

Figure 5.11b indicates, as expected that current penetration increases as electrode

5.4.2 Current Flow Lines and Current Density

If we examine Equations 5.2, 5.3, and 5.5, it becomes straightforward to demonstrate

5.4.4 Qualitative Development of the Resistivity Pattern over a Horizontal Interface

If a horizontal interface is deep relative to electrode spacing a or current electrode

In many resistivity investigations electrode spacing is plotted against apparent

Remembering that the potential at P1 is due to a direct and a reflected component, we

The potential at P2 (due to transmission) is:

Referring to Figure 5.17a, we proceed as in the previous example. The potential at P1

If we consider a single reflection from the upper boundary (equivalent to placing an

Up to this we have four equations that represent contributions to the potential at P1

However, we ultimately are interested in determining the potential difference due to a

Finally, we have a quantitative approach to determining the variation of apparent

5.4.6 Using RESIST

5.5 MULTIPLE HORIZONTAL INTERFACES

Although complexities and, therefore, identification and interpretation difficulties

5.6 VERTICAL CONTACT

When discussing apparent resistivity patterns associated with horizontal interfaces,

5.6.1 Constant-Spread Traverse

5.6.2 Expanding-Spread Traverse

5.7 TWO VERTICAL CONTACTS, HEMISPHERICAL STRUCTURES, AND

5.7.1 Two Vertical Contacts

5.7.2 Hemispherical Structures

5.7.3 Dipping Interfaces

If we sketch a qualitative representation of this behavior, our curve should appear

5.8 FIELD PROCEDURES

5.8.2 Electrode Configurations

The susceptibility of the Wenner electrode geometry to near-surface lateral variations

Because quantitative interpretations of Wenner and Schlumberger curves are very

5.8.3 Surveying Strategies

5.8.4 Other Considerations

Topography affects apparent resistivity measurements because it affects current flow.