Geoscience Frontiers 4 (2013) 469e476

Contents lists available at SciVerse ScienceDirect

China University of Geosciences (Beijing)

Geoscience Frontiers

journal homepage: www.elsevier .com/locate/gsf

Research paper

Application of 2-D geoelectrical resistivity tomography for subsurfacecavity detection in the eastern part of Saudi Arabia

Mohamed Metwaly a,b,*, Fouzan AlFouzan c

aKing Saud University, College of Science, Geology and Geophysics Department, P.O. Box 2455, Riyadh 11451, Saudi ArabiabNational Research Institute of Astronomy and Geophysics (NRIAG), EgyptcKing Abdulaziz City for Science and Technology (KACST), Saudi Arabia

a r t i c l e i n f o

Article history:Received 21 April 2012Received in revised form21 November 2012Accepted 12 December 2012Available online 2 January 2013

Keywords:Cavity detectionElectrical resistivity tomographySynthetic modeling

* Corresponding author. King Saud University, ColleGeophysics Department, P.O. Box 2455, Riyadh 114509499610; fax: þ966 1 4676214.E-mail address: mmetwaly70@yahoo.com (M. Metwal

Peer-review under responsibility of China University

Production and hosting by El

1674-9871/$ e see front matter � 2012, China Univerhttp://dx.doi.org/10.1016/j.gsf.2012.12.005

a b s t r a c t

The presence of natural voids and cavities in subsurface karstic limestones causes severe problem for civilengineering and environmental management. The presence of such features hinders the extension ofurbanizations particularly in the new metropolitan. The eastern part of Saudi Arabia contains varioustypes of karstic limestone, sinkholes, solution cavities and voids. In this context, geophysical methodsparticularly electrical resistivity technique is used as a cost-effective solution for investigating subsurfacecaves, voids, and shallow weathered zones. 2-D electrical resistivity data sets have been acquired alongseven profiles in the new urbanization at Al Hassa area. Data processing has been carried out taking intoconsideration the response of synthetic models, which simulates physical models of the most commonkarstic features in the area. The results are very useful to determine the extension of shallow weatheredzones and to locate different cavities underneath them. The hard limestone bedrock can also be detectedand traced along the surveyed profiles.

� 2012, China University of Geosciences (Beijing) and Peking University. Production and hosting byElsevier B.V. All rights reserved.

1. Introduction

The presence of natural voids and cavities in subsurface lime-stone causes severe problems for civil engineering and environ-mental management (Sum et al., 1996). The term subsurface cavityis used to denote all subsurface features cavities, caves, caverns,voids, karst, and sinkhole (Owen, 1983). The most abundant naturalcavities are formed by dissolution processes in carbonates (e.g.,limestone and dolomite) and evaporates (e.g., salt, gypsum, andanhydrite). Karst refers to a characteristic topographic feature thatis developed by dissolution made by downward percolating

ge of Science, Geology and51, Saudi Arabia. Tel.: þ966

y).

of Geosciences (Beijing).

sevier

sity of Geosciences (Beijing) and P

meteoric water (Benito et al., 1995). Sinkholes are common in thekarst areas. It is developed as a result of dissolution of limestonesthrough there action with acidic polluted water (Al-Zarah, 2007).After a certain period, groundwater which, penetrated throughweak zones in limestone will develop channels and voids (Elawadi,2003). This continuous process will reach a critical stage, whereroof of the space will no longer support the weight of the over-burden and other foundations. This will result in the developmentof sinkholes, which have negative and catastrophic effects on bothshallow and deep foundations. The presence of such cavernousfeatures leads to restrictions in land utilization and causes variablegeotechnical hazards like ground surface subsidence, collapsing ofsurface structures and cracks and fissures in the surface buildings(Elawadi, 2003). Detection and delineation of subsurface cavitiesremain a common task in many scientific and environmental fields(van Schoor, 2002; Zhou et al., 2002; Abu-Shariah, 2009). Classicalcommercial operations use drillings to locate cavities. A muchmoreeffective and economic solution is to use of the geophysical surveyto reduce the overall amount of drilling and to improve theplacement of test borings for more efficient target verification andmapping.

The eastern part of Saudi Arabia is covered by limestone as partof the Shedgum Plateau. The Shedgum Plateau is covered by

eking University. Production and hosting by Elsevier B.V. All rights reserved.

M. Metwaly, F. AlFouzan / Geoscience Frontiers 4 (2013) 469e476470

tertiary carbonates and evaporates of the Um er Radhuma, Rus,Dammam, Hadrukh, Dam and Hofuf formations (Al-Zarah, 2007; ElMahmoudi, 2010). The Shedgum Plateau and the adjacent areas andcities like Hofuf, Al Hassah and Al Uyoun are dotted with numerouskarstic features, sinkholes, solution cavities and caves (Edgell,1990a,b; Pint, 2003; Hussain et al., 2006). Due to increase in thedevelopment activities along many areas of Saudi Arabia, includingthe eastern province, many land subsidence and geotechnicalconstruction problems have been reported as a result of hostcarbonates and evaporate rocks solution process (Edgell, 1990b).When the carbonate rocks are wetted either during agriculturalactivities, groundwater exploitation, and waste disposal or evenduring rain storms, subsidence takes place due to either theremoval of salts from the rocks or the rearrangement of soilparticles in loess sediments (van Schoor, 2002).

Delineation of subsurface cavities and shallow weathered zoneusing geophysical methods has gained wide interests in the pastfew decades (Sum et al., 1996; Zhou et al., 2002; Abu-Shariah, 2009;Vachiratienchai et al., 2010). Electrical Resistivity Tomography(ERT) is one of the most frequently used geophysical techniques inexploring the subsurface cavities and weathered zones. It offersa quick and cost-effective imaging of the shallow subsurface withacceptable resolution. Therefore, the main aim of this study is todelineate the extent of weathered surface layers and the possibilityof cavities in the locality of Al Hassa area by applying a kind of non-destructive surface ERT method. Moreover, the most stable and

Figure 1. Location map of the study area

hard limestone bedrock in the area is also detected for supportingthe surface engineering constructions.

2. Study area

The southern part of Al Hassa area is undergoing rapid devel-opment activities in infrastructure and housing facilities (Fig. 1).During primary construction process, few surface and subsurfacecavities have been reported at different locations and depths(Fig. 2). Therefore, conducting the surface geophysical explorationfor the open and non-disturbed areas was highly recommended,particularly along some accessible areas within the new metro-politan. The main objective of survey was investigating the shallowsubsurface weak zones, subsurface cavities and sinkholes. Imagingsuch subsurface hazards is very important for designing thebuilding foundation and for choosing the safe areas and roads.

3. Application of ERT

The 2-D electrical resistivity method has wide applications inenvironmental, engineering and shallow subsurface investigations(van Schoor, 2002). It is based on assumption that various entitieslike minerals, solid bedrock, sediments, air and water filled struc-tures have detectable electrical resistivity contrast relative to thehost medium (Pánek et al., 2010). In this study it was supposed thatconsiderable resistivity contrasts between hard limestone bedrock,

in the eastern part of Saudi Arabia.

Figure 2. Examples of the subsurface cavities documented close to the study area, Al Hassa region, Saudi Arabia.

M. Metwaly, F. AlFouzan / Geoscience Frontiers 4 (2013) 469e476 471

the intercalated cavities and weak subsurface zones can bedetected.

The principle of ERT technique consists of the application ofconstant direct current imposing into the ground via two currentelectrodes and measuring the resulting voltage at two potentialelectrodes. Themethod is based onmulti-electrode andmulti-cablesystem. Each of the electrodes alternatively acts as current andpotential electrode. The whole profile is measured without theintervention of surveyors. The position of current and potentialelectrodes during the measurement is dependent on the chosengeometry of electrode arrays. Most frequently used arrays are thedipoleedipole, Wenner and WennereSchlumberger arrays (Fig. 3).Each electrode configuration has specific advantages and disad-vantages, based on the penetration depth and the horizontalresolution. In this study, the WennereSchlumberger configurationhas been utilized because it provides quite deep penetration, reli-able stability and ability to detect both horizontal and verticalsubsurface features (Dahlin and Zhou, 2004; Candansayar, 2008).The configuration is based on fixed spacing between potentialelectrodes while, the spacing between current electrodes is loga-rithmically increased for number of dipole lengths (Fig. 3). Then thespacing between potential electrodes is increased in order to obtainmore deep penetration.

4. Data acquisition

The 2-D ERT data have been acquired along seven profiles in themost probable cavity affected areas in the newly metropolitanconstructed site (Fig. 1) using multi-electrode system of SYSCAL

Figure 3. Most common electrode arrays including the utilized in the current study.Explanation of indexes: C1, C2dcurrent electrodes; P1, P2dpotential electrodes; a andnaddistance between the electrodes.

Pro. WennereSchlumberger electrode array has been utilized andthe unit electrode spacing was 2 m along all the acquired profiles.To cover the proposed profile length and maintain reasonablesubsurface lateral resolution, roll-along acquisition technique wasperformed. The total profile length was divided into segments. Eachof them has maximum length of about 128 m with 20e30 moverlap. Then the concatenate option from RES2DINV (Loke andBarker, 1996) inversion software has been set to compile differentsegment together in one profile. The data acquisition strategywas (i) to select at least one profile across or very close tocavity feature (Fig. 2) aiming to obtain typical resistivity evidence,and (ii) to situate additional possible profiles through sites withsimilar morphological expression but unknown features. The entiremeasurement was carried out along the possible profiles in thestudy area after scattering some water to ensure a wet surfacecondition, which is favored for getting good coupling betweenelectrodes and ground for signal quality.

Generally the data obtained during ERT field measurements areclassically presented as apparent resistivity pseudo-sections, whichgive an approximate picture of the subsurface resistivity. Before theinversion process, to obtain a true model representing continuousdistribution of calculated electrical resistivity in the subsurface, thedatawere concatenated and the noise and spiky values were edited.Then inversion procedure was utilized using RES2DINV software,which is based on the regularized least-squares optimizationmethod (Sasaki, 1989; deGroot-Heldin and Constable, 1990; Lokeet al., 2003). The blocky optimization method was applied duringthe data inversion as the expected subsurface targets have sharpboundaries (Ellis and Oldenburg, 1994; Metwaly et al., 2008).

5. Results

The surveyed area is almost flat. Therefore, the topographicdifference was not considered in any of the inverted profiles. Afteradjusting the input data and choosing the optimum inversionparameters, a few iterations were enough to get the invertedresistivity models representing the subsurface distribution untilthe depth of about 15.0 m from the ground surface. The obtainedinformation about the shallow subsurface along the resistivitycross-sections was interpreted based on high and low resistivityvalues, and its relation to the known features in the surveyed area.Investigation of the inverted 2-D ERT profiles (Fig. 4) revealed thefollowing features:

(1) The shallow weathered limestone layer was detected withrelatively low resistivity values (less than 24 U m) at variabledepths based on the degree of weathering and/or the surfaceirregularities of the underneath layer (Fig. 4). The low resis-tivity values of the weathered layer referred to the wetconditions from the surface construction activities.

Figure 4. Inverted 2-D resistivity sections along the accessible roads in the study area. (a) and (b) are the P1E and P1W of profile P1, (c) is the profile 3 and (d) is the profile 4.

M.M

etwaly,F.A

lFouzan/Geoscience

Frontiers4(2013)

469e476

472

Figure 5. 2-D resistivity model and the three inverted results using different noise percentages for the cavity and sinkhole in limestone rock. (a) The physical model, (b) model with 0% noise, (c) model with 10% noise, (d) model with20% noise.

M.M

etwaly,F.A

lFouzan/Geoscience

Frontiers4(2013)

469e476

473

M. Metwaly, F. AlFouzan / Geoscience Frontiers 4 (2013) 469e476474

(2) The hard limestone bedrock can be investigated clearly withvariable surface topography. It has a higher resistivity values(>60 U m), and extend almost to the end of the investigateddepths (Fig. 4).

(3) Along some places of the inverted resistivity cross-sections,there are local low resistivity values extending from thesurface weathered layer to the inside of the hard limestonebedrock (Fig. 4). It represents ideal shape for subsurface cavi-ties and sinkholes. These features have no surface manifesta-tions. However, many of these features have been reportedduring the preliminary construction process.

(4) Along some profiles, it is possible to trace shallow weatheredzones inside the massive limestone layer. Such feature is alsocharacterized by low resistivity values (Fig. 4). It is consideredas a preliminary stage of formulating different cavity and/orsinkhole in limestone.

In spite of clear interpretation of most of the karstic anomalousresistivity features in the processed profiles, forward modeling forthe possible features that are traced in the inverted 2-D resistivityprofiles has been constructed. The main purpose of the last step isto clarify ambiguities in the inverted resistivity profiles and to

Figure 6. 2-D resistivity model and the three inverted results using different noise percenta0% noise, (c) model with 10% noise, (d) model with 20% noise.

confirm the location, extension and shapes of the cavernousfeatures without boring or excavations.

6. Synthetic models

Based on the subsurface resistivity distribution and the config-uration of the karstic features that can be traced in the 2-D invertedprofiles, synthetic models have been designed for confirming theinterpretation of the inverted 2-D resistivity profiles. The forwardmodeling has been designed using RES2DMOD software (Loke andBarker, 1996) utilizing the finite difference calculations, whichdivides the subsurface into number of rectangular blocks withspecific resistivity values (Dey and Morrison, 1979; Loke et al.,2003). The first synthetic model shown in Fig. 4d contains lime-stone with the weathered zone at the surface and two mainsubsurface features. One is a cavity and the other is considered asa sinkhole. Simply themodel has two resistivity values. The shallowweathered zone and the subsurface features have a resistivityof 10 U m. The low resistivity values of the weathered zonesare assigned to the increase in moisture content as indicated bythe surface observations and processed resistivity data (Figs. 5aand 6a). The host dry limestone is characterized by relatively high

ges for the weathering effects in limestone rock. (a) The physical model, (b) model with

Figure 7. 3-D view for inverted resistivity sections along the different profiles in the study area.

M. Metwaly, F. AlFouzan / Geoscience Frontiers 4 (2013) 469e476 475

resistivity value (100 U m). The physical parameters of the secondsyntheticmodel are similar with the firstmodel. However, there areonly two resistivity features in this model (Fig. 6a). The dry hostlimestone (100 U m) contains the weathered subsurface layer(10 U m), which extends to some distance into subsurface. More-over, there is a small low resistivity feature (10 U m), representinga small cave along the same horizon.

Firstly, the resistivity responses have been calculated using theWennereSchlumberger electrode array without considering anynoise percentages. Then two different values of Gauss distributedrandom noise with a standard deviation of 10% and 20% respec-tively were added to the calculated responses for both models inorder to simulate the field conditions. The synthetic apparentresistivity data were then inverted using a block inversion schemeand the same parameters are applied for inverting the acquiredresistivity data.

6.1. Inverted ERT of Model 1

The inverted section of the noise free data indicated the locationof the two subsurface features and the shallow weathered zone.The boundaries of the cavities were determined, but the bottom ofthe sinkhole is extended to the end of the investigated depth. Thesefeatures are close to that observed in Fig. 4d, which have almost thesame subsurface structures. With increasing the noise percentageto about 10%, it is possible to locate the same features. However,with 20% noise percentage, the two features are slightly shifted upwards (Fig. 5).

6.2. Inverted ERT of Model 2

Similarly, the second model, which represents the subsurfaceweathered zones, has clear inverted features in the case of noisefree data (Fig. 6b). The extended subsurface weathered zone ischaracterized by relatively low resistivity values as compared to theresistivity values of limestone. In addition, other small cavity can beclearly detected at the same location and depth relative to theoriginal model. Such feature is comparable with the resistivity

profile in Fig. 4. By increasing the noise percentage to be 10% and20%, it is possible to trace and locate the same subsurface featureswith slight smearing due to noise effects (Fig. 6c and d).

7. Generalized fence diagram

In order to visualize the continuity of the shallow weatheredzone and the subsurface hazard features along the study area,a generalized fence diagram has been prepared using the invertedresistivity data sets along all profiles (Fig. 7). Powerful geostatisticalgridding (kriging) method of Geosoft software has been applied tosmoothly interpolate the resistivity data along these profiles. Linearcolor scale has been used to visualize the limited resistivity range(1e100). It is possible to identify the extension of the weatheredzone in subsurface (marked with number 3), which has an impor-tant relation to the designed foundation depth. The second essen-tial feature is the weathered zones underneath the hard limestone(marked with number 2). It is very important to know the locationof those features because they have very little or no surfacemanifestation. Moreover, the occurrence of subsurface cavities canbe traced along the surveyed lines at different places (marked withnumber 1). Such features are extending to deeper depth, and havelittle or no extension at the surface. The depth to the solid andundisturbed limestone bedrock (marked with number 4) can betraced clearly along all the surveyed lines.

8. Conclusions

The eastern part of the Saudi Arabia, which consists mainly oflimestone, has great developing activities. Investigating the thick-ness of the shallow weathered zone and imaging the possibility ofsubsurface cavities are essential process before the infra and surfacefoundation construction. Surface cavities were reported from thenearby sites. The ERT data sets were acquired along seven profilesin Al Hassa region where new metropolitan is being constructed.WennereSchlumberger electrode array was utilized using 2 melectrode spacing. The inverted resistivity data represent the resis-tivity distribution to maximum depth of about 15 m. The shallow

M. Metwaly, F. AlFouzan / Geoscience Frontiers 4 (2013) 469e476476

weathered layer of the limestone was detected with relatively lowresistivity values at variable depths. Underneath the shallowweathered zone there are two important features with low resis-tivity values as well. The first one is the subsurface cavities, whichextend from the shallow weathered zones into the solid limestone.The second is the subsurface weathered zones that are formed dueto the weathering effects of surface and groundwater inside thelimestone. They are formed inside the solid limestone with rela-tively high resistivity values. Moreover, hard limestone bedrock wastraced along themeasured profileswith high resistivity values. Mostof these features were validated based on the modeling techniqueand the inversion process to get similar features.

Acknowledgment

The second author thanks and appreciates KACST for their helpand support during this study.

References

Abu-Shariah, M.I.I., 2009. Determination of cave geometry by using a geoelectricalresistivity inverse model. Engineering Geology 105, 239e244.

Al-Zarah, A.I., 2007. Hydrogeochemical processes of Alkhober aquifer in easternregion, Saudi Arabia. Journal of Applied Science 7, 3669e3677.

Benito, G., Perez del Campo, P., Gutierrez-Elorza, M., Sancho, C., 1995. Natural andhuman-induced sinkholes in gypsum terrain and associated environmentalproblems in NE Spain. Environmental Geology 25, 156e164.

Candansayar, M.E., 2008. Two-dimensional individual and joint inversion of three-and four-electrode array dc resistivity data. Journal of Geophysics and Engi-neering 5, 290e300.

Dahlin, T., Zhou, B., 2004. A numerical comparison of 2-D resistivity imaging with10 electrode arrays. Geophysical Prospecting 52, 379e398.

deGroot-Hedlin, C., Constable, S.C., 1990. Occam’s inversion to generate smooth, two-dimensional models from magnetotelluric data. Geophysics 55, 1613e1624.

Dey, A., Morrison, H.F., 1979. Resistivity modeling for arbitrarily shaped two-dimensional structures. Geophysical Prospecting 27, 106e136.

Edgell, H.S., 1990a. Geological framework of Saudi Arabian groundwater resources.Journal of King Abdulaziz University, Earth Sciences 3, 267e286 (Special issue).

Edgell, H.S., 1990b. Karst in northeastern Saudi Arabia. Journal of King AbdulazizUniversity: Earth Science 3, 81e94 (Special issue): 1st Saudi Symposium onEarth Science.

El Mahmoudi, A. S., Khalied M. A. and Massoud A. M., 2010. 2-D electrical tomog-raphy for mapping of aquifer at the new campus of King Feisal University, AlHassa, KSA. 4th International Conference on Water Resources and Arid Envi-ronments (ICWRAE4), 267e280.

Elawadi, E. A., 2003. Cavity detection by integrated geophysical methods. PhD.Thesis, Kyushu University, Japan.

Ellis, R.G., Oldenburg, D.W., 1994. The pole-pole 3-D DC-resistivity inverse problem:a conjugate-gradient approach. Geophysical Journal International 119, 187e194.

Hussain, M., Al-Khalifah, F., Khandaker, N.I., 2006. The Jabal Al Qarah caves of theHofuf area, northeastern Saudi Arabia: a geological investigation. Journal ofCave and Karst Studies 68, 12e21.

Loke, M., Barker, R., 1996. Rapid least-squares inversion of apparent resistivitypseudosection by a quasi-Newton method. Geophysical Prospecting 44,131e152.

Loke, M.H., Acworth, I., Dahlin, T., 2003. A comparison of smooth and blockyinversion methods in 2D electrical imaging surveys. Exploration Geophysics 34,182e187.

Metwaly, M., El-Qady, G., Matsushima, J., Szalai, S., Al-Arifi, N.S.N., Taha, A., 2008.Contribution of 3-D electrical resistivity tomography for landmines detection.Nonlinear Processes in Geophysics 15, 977e986.

Owen, T.E., 1983. Detection and mapping of tunnels and caves. In: Fitch, A.A. (Ed.),Developments in Geophysical Exploration Methods-5. Applied SciencePublishers Ltd, pp. 161e258.

Pánek, T., Margielewski, W., Tábo�rík, P., Urban, J., Hradecký, J., Szura, C., 2010.Gravitationally induced caves and other discontinuities detected by 2D elec-trical resistivity tomography: case studies from the Polish Flysch Carpathians.Geomorphology 123, 165e180.

Pint, J., 2003. The Desert Caves of Saudi Arabia. Stacey International, London, 120 p.Sasaki, Y., 1989. Two-dimensional joint inversion of magnetotelluric and dipole-

dipole resistivity data. Geophysics 54, 174e187.Sum, C. W., Othman, J., Loganathan, P., 1996. Geotechnical problems in limestone

terrain with emphasis on cavities and sinkholes. Seminar Geologi dan Sekitarn,UKM, Puri Pujangga, Bangi, pp. 102e117.

Vachiratienchai, C., Songkhun, B., Weerachai, S., 2010. A hybrid finite differ-enceefinite element method to incorporate topography for 2D direct current(DC) resistivity modeling. Physics of the Earth and Planetary Interiors 183,426e434.

van Schoor, M., 2002. Detecting of sinkholes using 2D electrical resistivity imaging.Journal of Applied Geophysics 50, 393e399.

Zhou, W., Beck, B.F., Adams, A.L., 2002. Effective electrode array in mapping karsthazards in electrical resistivity tomography. Environmental Geology 42,922e928.