Contributions to Geophysics and Geodesy

Vol. 42/2, 2012 (161180)

Determination of dipping contacts using

electrical resistivity tomography
1 , Ivan DOSTAL
1 , David KUSNIR

1
Rene PUTISKA
AK
1

Department of Applied and Environmental Geophysics, Faculty of Natural Sciences

Comenius University, Mlynsk
a dolina, pav. G, 842 48 Bratislava, Slovak Republic
e-mail: putiska@fns.uniba.sk; dostal@fns.uniba.sk; kusnirak@fns.uniba.sk

Abstract: Generally, all electrode arrays are able to delineate the contact of two lithostratigraphic units especially with very high resistivity contrast. However, the image resolution for the location of vertical and dipping structures is dierent. The responses of
dipole-dipole (DD), Wenner alpha (WA), Schlumberger (SCH) and combined pole-dipole
(PD) arrays have been computed using the nite dierence method. Comparison of the
responses indicates that: (1) The dipole-dipole array usually gives the best resolution
and is the most detailed method especially for the detection of vertical structures. This
array has shown the best resolution to recognize the geometrical characterisation of the
fault. (2) The pole-dipole has shown the second best result in our test. The PD is an
eective method for detection of vertical structures with a high depth range, but the
deepest parts are deformed. (3) Wenner alpha shows a low resolution, inconvenient for
detailed investigation of dip structures. (4) The Schlumberger array gives a good and
sharp resolution to assess the contact between two lithological units but gives poor result
for imaging geometry of dipping contact.
Key words: electrical resistivity tomography, L2 norm inversion method, L1 norm
inversion method, vertical and dipping contact

1. Introduction
The resistivity tomography is a common method to indicate vertical structures (Caputo et al., 2003; Wise et al., 2003; Rizzo et al., 2004; Nguyen et
al., 2005 and Fazzito et al., 2009) and others, but what is the possibility
of the method to determine angle of the dipping contact? For investigation
of this problem we dened 2D subsurface model with three dierent angles
60 , 90 and 120 which were investigated through a numerical modelling. A
variety of electrode arrays are available for exploration using the resistivity
tomography. Comparison of the responses of dipole-dipole (DD), Wenner
doi: 10.2478/v10126-012-0007-6

161

Unauthenticated
Download Date | 11/29/15 8:17 AM

Putiska R. et al.: Determination of dipping contacts using . . .

(161180)

alpha (WA), Schlumberger (SCH), combined pole-dipole (PD) arrays have

been computed using the nite dierence method.
The choice of the best array for a eld resistivity survey depends on
the type of structure to be mapped, the sensitivity of the resistivity meter
and the background noise level. Each type of combination has advantages
and limitations in terms of lateral resolution and vertical penetration for
instance.

2. The synthetic test

The main goal of this study is the ability of electrical tomography to make
a visual estimation for direction and slope of vertical contact. In order to
assess the trend of dip, the following has been carried out to compare the
eect of the inversion method.
The synthetic data are computed using the forward modelling program
RES2DMOD (Loke and Barker, 1996), with no Gaussian noise added to
the data sets. The synthetic models used in this work represent the three
geometrical settings of the contact of two geological environments (a vertical
90 and a dipping 60 and 120 ) and they represent a simplied geological
and structural sketch along the prole (Fig. 3a, 6a, 9a). These models have
been saved as the data of the apparent resistivity without topographic information. Synthetic data sets were generated for dipole-dipole (DD), Wenner
alpha (WA), Schlumberger (SCH), combined pole-dipole (PD) arrays with
an electrode spacing of 5.5 m.
Firstly, three models with a dierence in the thickness and number of
the model layers were compared (Fig. 1b, c, d). The RES2DINV (Loke and
Barker, 1996) program generated size and positions of the model blocks
(Loke, 1997). The model (Fig. 1a) has subdivided the subsurface into
rectangular blocks with varying resistivity (Loke and Barker, 1995), the
overview of the main parameters of models is compiled in Table 1.
For inversion process the software RES2DINV among others use inversion methods as L2 norm, L1 norm with standard horizontal and vertical
roughness lter and L1 norm with diagonal roughness lter.
The L2 norm inversion method gives optimal results where the subsurface geology exhibits a smooth variation, such as the diusion boundary of
162

Unauthenticated
Download Date | 11/29/15 8:17 AM

Contributions to Geophysics and Geodesy

Vol. 42/2, 2012 (161180)

Table 1. Model parameters of the three models used in the synthetic test

a chemical plume. However, in cases where the subsurface consists of bodies

that are internally homogeneous with sharp boundaries (such as an igneous
dyke), this method tends to smear out the boundaries. The L 1 norm or
blocky optimisation method tends to produce models that are piecewise
constant (Ellis and Oldenburg, 1994). The data sets have been inverted
using the L2 norm inversion method (Fig. 1h, i, j), L 1 norm with standard
horizontal and vertical roughness lter (Fig. 1e, f, g), and L 1 norm with diagonal roughness lter (Fig. 1k, l, m).
The results show that for the L1 norm inversion method it is better
to use a model where the number of model cells exceeds the number of
data points (Loke et al., 2003), what eliminates the eect of the block
(Fig. 1g, m). The robust inversion with wide model blocks gave images that
were too blocky (Fig. 1e, f, k, l). In case of the L 2 norm inversion method
very wide cells reduce the eect of smoothness (Fig. 1h, i, j).
In the next work phase a model C has been used (Fig. 1d, Table 1).
During further analysis the eect of three dierent factors able to inuence
the result was tested:
1. resistivity contact of two dierent lithological units with varying angle
of the dip (60, 90 and 120 degrees),
2. type of electrode array to investigate the imaging capabilities of these
electrode congurations,
3. eect of the inversion method (L2 norm inversion method, L1 norm
with standard horizontal and vertical roughness lter and L 1 norm
with diagonal roughness lter) to dene the dip of tectonic fault (60,
90 and 120 degrees).
163

Unauthenticated
Download Date | 11/29/15 8:17 AM

Putiska R. et al.: Determination of dipping contacts using . . .

(161180)

Fig. 1. The eect of models with dierent methods to subdivide the subsurface into
rectangular prisms on the inversion result: a) geometry of synthetic model; b, c, d)
arrangement of model blocks and apparent resistivity data points; e, f, g) inverse model
using the L1 norm with standard horizontal and vertical roughness lter; h, i, j) inverse
model using the L2 norm inversion method; k, l, m) inverse model using the L1 norm
with diagonal roughness lter.

164

Unauthenticated
Download Date | 11/29/15 8:17 AM

Contributions to Geophysics and Geodesy

Vol. 42/2, 2012 (161180)

The results of the work are shown in Fig. 311. The pictures show a set of
inversion models with dierent arrays conguration and dierent inversion
parameters. The inversion models are calculated from synthetic data with
vertical contact (90 degrees Fig. 3, 4, 5) and dipping contact (60 degrees
Fig. 6, 7, 8 or 120 degrees Fig. 9, 10, 11).

3. Electrical arrays
Generally, all tested electrode arrays (Fig. 2) are able to delineate the contact
of two lithostratigraphic units especially with very high resistivity contrast.
However, the image resolution for the location of vertical and dipping structures is dierent.
Based on the numerical simulations, it is possible to summarize the main
advantages and disadvantages of these arrays when assessing the dip of the
fault. These electrode congurations were also investigated using robust
(L1 norm) inversion and smoothness-constrained least-squares (L 2 norm)
inversion (Farquharson, 2008) for the three synthetic models.
Dipole-Dipole (DD) is the most detailed method especially for the detection of vertical structures. This array has shown the best resolution to
recognize the geometrical characterisation of the fault. The depth range of
this array is about 1/5 of the maximum C2P2 distance used. The eective
depth range is strongly limited by a rapid decrease of the measured potential
at larger dipole distance. Articial electric noise causes additional significant limitation on the use of this method. The highest resolution allows

Fig. 2. The arrays used in resistivity test and their geometric factors.

165

Unauthenticated
Download Date | 11/29/15 8:17 AM

Putiska R. et al.: Determination of dipping contacts using . . .

(161180)

the maximum possible dierentiation of deeper situated structures, but the

deeper structures may be deformed (Fig. 3b11b).
The combined Pole-Dipole (PD) array has shown the second best result
in our test. The PD is an eective method for detection of vertical structures with a high depth range, but the deepest parts are deformed. Section
coverage is about 1/3 of the used length of the electrode array. The accuracy of positions in section decreases (side shift) with depth Fig. 3c11c.
This array requires the installation of an external current electrode C2 (C1
in case of a reverse way) called the innite electrode. The place of the
innite electrode must be at least at the distance of vefold of the maximum
length of the used electrode array. Its optimum position should be in the
perpendicular direction from the electrode array. The long distance of the
innite current electrode requires a maximum power of the transmitter and
a careful installation of such an electrode (or even electrode nest) to reach
its lowest possible ground resistance.
The Wenner alpha (WA) array has a low depth range (about 1/6 of the
maximum used C1C2 distance) and a low side covering. The results of this
array show a low resolution, inconvenient for detailed investigation of dip
structures because all models show more or less vertical contact without the
geometry of dip Fig. 3d11d.
Schlumberger (SCH) is suited especially for depicting horizontal and subhorizontal (declined) layers. Detection of larger inhomogeneities of various
shape and direction like wider crackles, tectonic zones and contacts of layers
with dierence of resistivities is also eective. This array gives a good and
sharp resolution to assess the contact between two lithological units but a
poor result of imaging geometry of dipping contact (Fig. 3e11e). Section
covering of this array is about 1/5 of the maximum used C1C2 distance
and the resolution is medium sucient, rather for detailed investigation of
shallow structures.

4. Test results
The L2 norm smoothness-constrained optimization method produces a model with a smooth variation of resistivity values. After the comparison of
DD, PD, WA and WS results, it is possible to say that the contact between the lithological units is not sharp enough to assess the exact dip
166

Unauthenticated
Download Date | 11/29/15 8:17 AM

Contributions to Geophysics and Geodesy

Vol. 42/2, 2012 (161180)

Fig. 3. Inverse model resistivity sections from synthetic data set (using the L2 norm
inversion method): a) geometry of the synthetic model, b) dipole-dipole array, c) poledipole array, d) Wenner-alpha array, e) Schlumberger array.

167

Unauthenticated
Download Date | 11/29/15 8:17 AM

Putiska R. et al.: Determination of dipping contacts using . . .

(161180)

Fig. 4. Inverse model resistivity sections from synthetic data set (using the L1 norm with
standard horizontal and vertical roughness lter): a) geometry of the synthetic model,
b) dipole-dipole array, c) pole-dipole array, d) Wenner-alpha array, e) Schlumberger array.

168

Unauthenticated
Download Date | 11/29/15 8:17 AM

Contributions to Geophysics and Geodesy

Vol. 42/2, 2012 (161180)

Fig. 5. Inverse model resistivity sections from synthetic data set (using the L1 norm
with diagonal roughness lter): a) geometry of the synthetic model, b) dipole-dipole
array, c) pole-dipole array, d) Wenner-alpha array, e) Schlumberger array.

169

Unauthenticated
Download Date | 11/29/15 8:17 AM

Putiska R. et al.: Determination of dipping contacts using . . .

(161180)

Fig. 6. Inverse model resistivity sections from synthetic data set (using the L2 norm
inversion method): a) geometry of the synthetic model, b) dipole-dipole array, c) poledipole array, d) Wenner-alpha array, e) Schlumberger array.

170

Unauthenticated
Download Date | 11/29/15 8:17 AM

Contributions to Geophysics and Geodesy

Vol. 42/2, 2012 (161180)

Fig. 7. Inverse model resistivity sections from synthetic data set (using the L1 norm with
standard horizontal and vertical roughness lter): a) geometry of the synthetic model,
b) dipole-dipole array, c) pole-dipole array, d) Wenner-alpha array, e) Schlumberger array.

171

Unauthenticated
Download Date | 11/29/15 8:17 AM

Putiska R. et al.: Determination of dipping contacts using . . .

(161180)

Fig. 8. Inverse model resistivity sections from synthetic data set (using the L1 norm
with diagonal roughness lter): a) geometry of the synthetic model, b) dipole-dipole
array, c) pole-dipole array, d) Wenner-alpha array, e) Schlumberger array.

172

Unauthenticated
Download Date | 11/29/15 8:17 AM

Contributions to Geophysics and Geodesy

Vol. 42/2, 2012 (161180)

Fig. 9. Inverse model resistivity sections from synthetic data set (using the L2 norm
inversion method): a) geometry of synthetic the model, b) dipole-dipole array, c) poledipole array, d) Wenner-alpha array, e) Schlumberger array.

173

Unauthenticated
Download Date | 11/29/15 8:17 AM

Putiska R. et al.: Determination of dipping contacts using . . .

(161180)

Fig. 10. Inverse model resistivity sections from synthetic data set (using the L1 norm with
standard horizontal and vertical roughness lter): a) geometry of the synthetic model,
b) dipole-dipole array, c) pole-dipole array, d) Wenner-alpha array, e) Schlumberger array.

174

Unauthenticated
Download Date | 11/29/15 8:17 AM

Contributions to Geophysics and Geodesy

Vol. 42/2, 2012 (161180)

Fig. 11. Inverse model resistivity sections from synthetic data set (using the L1 norm
with diagonal roughness lter): a) geometry of the synthetic model, b) dipole-dipole array,
c) pole-dipole array, d) Wenner-alpha array, e) Schlumberger array.

175

Unauthenticated
Download Date | 11/29/15 8:17 AM

Putiska R. et al.: Determination of dipping contacts using . . .

(161180)

of the fault, even if the number of model cells exceeds the number of data
points. The L2 norm smooth inversion method is not optimal for such data
sets (Fig. 3, 6, 9). The L2 norm inversion method gives the optimal results
where the subsurface geology shows a smooth variation, such as the diusion boundary of a chemical plume. However, in cases where the subsurface
consists of bodies that are internally homogeneous with sharp boundaries
(such as an igneous dyke), this method tends to smear out the boundaries
(Loke et al., 2010).
In this case the L1 norm with standard horizontal and vertical roughness lter (Fig. 4, 7, 10) gave dierent results for each of the arrays used.
For DD (Fig. 4b, 7b, 10b) and PD (Fig. 4c, 7c, 10c) the model contains sharp
dipping and vertical interface with some deformations on the bottom. The
WA (Fig. 4d, 7d, 10d) and SCH (Fig. 4e, 7e, 10e) result show a sharp vertical
interface but these two arrays are not able to recognize the geometry of dip.
In case of the L1 norm with diagonal roughness lter the nite-element
method with triangular elements was used for the forward modelling calculation (Fig. 5, 8, 11). On the Fig. 1 are shown models with dierent number
of blocks but with the same number of data points (Table 1).
The RMS error resulting from the inverse model calculation for all three
mentioned inverse methods has been used as a quantitative tool to compare
the suitability of the inversion method (Fig. 12). An RMS error parameter
has been normalized to 1000 points of the input model.
The model A arrangement shows 206 numbers of blocks. This model has
been inverted using the L2 norm inversion method (Fig. 1h), the L 1 norm
with standard horizontal and vertical roughness lter (Fig. 1e) and the
L1 norm with diagonal roughness lter (Fig. 1k). Figure 1k shows the
resulting model, when the diagonal nite elements were given more importance than the horizontal and vertical dierences (Fig. 1e).
The third model arrangement shows 1506 numbers of blocks. In this case
the number of model cells exceeds the number of data points. The dierences between the L1 norm with standard horizontal and vertical roughness
lter (Fig. 1g) and the L1 norm with diagonal roughness lter (Fig. 1m) are
not signicant both methods give very similar results (Fig. 4, 5, 7, 8, 10, 11).
From the results of the test we can state several outputs:
1. each of the involved electrical arrays is able to identify resistivity contact between two lithological layers, but with increasing depth the reso176

Unauthenticated
Download Date | 11/29/15 8:17 AM

Contributions to Geophysics and Geodesy

Vol. 42/2, 2012 (161180)

Fig. 12. RMS error comparison resulted from the inversion of the models. The error
is normalised per 1000 points of the model for all electrical arrays involved in test
the Dipole-Dipole (DD), combined Pole-Dipole (PD), Wenner alpha (WA) and WennerSchlumberger (SCH). All three contact variants with 90 degrees, 60 degrees and 120
degrees angle of the dip were calculated using L2 norm and L1 norm with standard
settings and also with diagonal lter components (L1 norm, DFC).

lution of the inverse model does not allow to recognize sharp boundary
between the layers. The eect is caused by the lower density of the
measured points in deeper parts of the image.
2. reliable quantitative estimation of the angle between the lithological
layers combined Pole-Dipole (PD) and Dipole-Dipole shows best results
unlike WA and SCH do.
3. eect of the selected inversion method is signicant to the nal inversion model as the L1 norm with diagonal lter components gives in
most of the cases best results (Fig. 12).
177

Unauthenticated
Download Date | 11/29/15 8:17 AM

Putiska R. et al.: Determination of dipping contacts using . . .

178

Unauthenticated
Download Date | 11/29/15 8:17 AM

(161180)

Contributions to Geophysics and Geodesy

Vol. 42/2, 2012 (161180)

5. Conclusions
A synthetic test allows us to compare the capabilities of the selected electrode array and inversion settings to identify a dipped contact of two geological layers and estimate the real angle between them. The data have
been calculated by the L2 norm inversion method, the L1 norm with standard horizontal and vertical roughness lter and the L 1 norm with diagonal
roughness lter. The 2D inversion result of the resistivity prole correlated
with synthetic models. From these synthetic models it is possible to summarize the main advantages or disadvantages of these arrays for 2D resistivity
imaging compiled in Table 2.

Acknowledgments. The authors are grateful to the Slovak Research and Development Agency APVV (grant nos. APVV-0194-10) and the Slovak Grant Agency VEGA
(grant nos. 1/0095/12, 2/0067/12, 1/0747/11) for the support of their research.

References
Caputo R., Piscitelli S., Oliveto A., Rizzo E., Lapenna V., 2003: The use of electrical
resistivity tomographies in active tectonics: examples from Tyrnavos Basin, Greece.
Journal of Geodynamics, 36, 1935.
Ellis R. G., Oldenburg D. W., 1994: Applied geophysical inversion. Geophysical Journal
International, 116, 511.
Farquharson C. G., 2008: Constructing piecewise-constant models in multidimensional
minimum- structure inversions. Geophysics, 73, K1K9.
Fazzito S. Y., Rapalini A. E., Cortes J. M., Terrizzano C. M., 2009: Characterization of
Quaternary faults by electric resistivity tomography in the Andean Precordillera of
Western Argentina. Journal of South American Earth Sciences, 28, 217228.
Loke M. H, Barker R. D., 1995: Least-squares deconvolution of apparent resistivity pseudosection. Geophysics, 60, 16821690.
Loke M. H., Barker R. D., 1996: Rapid least-squares inversion of apparent resistivity
pseudosections by a quasi-Newton method. Geophysical Prospecting, 44, 131152.
Loke M. H., 1997: Rapid 2D resistivity inversion using the least-squares method, RES2
DINV Program manual, Penang, Malaysia.
Loke M. H., Acworth I., Dahlin T., 2003: A comparison of smooth and blocky inversion
methods in 2D electrical imaging surveys. Exploration Geophysics, 34, 182187.

179

Unauthenticated
Download Date | 11/29/15 8:17 AM

Putiska R. et al.: Determination of dipping contacts using . . .

(161180)

Loke M. H., Wilkinson P. B., Chambers J. E., 2010: Fast computation of optimized
electrode arrays for 2D resistivity surveys. Computers and Geosciences, 36, 11,
14141426.
Nguyen F., Garambois S., Jongmans D., Pirard E., Loke M. H., 2005: Image processing of
2D resistivity data for imaging faults. Journal of Applied Geophysics, 57, 260277.
Rizzo E., Colella A., Lapenna V., Piscitelli S., 2004: High-resolution images of the faultcontrolled high agri valley basin (Southern Italy) with deep and shallow electrical
resistivity tomographies. Physics and Chemistry of the Earth, 29, 321327.
Wise D. J., Cassidy J., Locke C. A., 2003: Geophysical imaging of the quaternary Wairoa
North Fault, New Zealand: a case study. Journal of Applied Geophysics, 53, 116.

180

Unauthenticated
Download Date | 11/29/15 8:17 AM