-
Geophysical Prospecting, 2015 doi: 10.1111/1365-2478.12248
Evidence for a new regional NW–SE fault and crustal structurein
Tunisia derived from gravity data
Mohamed Arfaoui1,3∗, Alan Reid2,4 and Mohamed Hedi
Inoubli31Office National des Mines, BP no. 215, 1080 Tunis,
Tunisia, 2Reid Geophysics Ltd, 7 Keymer House, Michel Grove,
Eastbourne, BN211JZ, U.K., 3Unité de Recherche de Géophysique
Appliquée aux Matériaux et aux Minerais, Université Tunis-El
Manar, Faculté des Sciences deTunis, 2092 El Mana, Tunis, Tunisia,
and 4School of Earth & Environment, University of Leeds, Leeds,
LS2 9JT, U.K.
Received July 2013, revision accepted January 2015
ABSTRACTA new Tunisian gravity map interpretation based on the
Gaussian filtered residualanomaly, total horizontal gradient, and
Moho discontinuity morphology establishedfrom gravity data exhibit
a new regional northwest–southeast fault extending fromEastern
Kairouan to Ghardimaou (Algeria–Tunisia Boundary). It presents a
horizon-tal gradient maximum lineament that terminates the
north–south Jurassic structures inthe Kairouan plain. Further, this
interpretation reveals other known fault systems andcrustal
structures in Tunisia. The new regional northwest–southeast fault
constituteswith the north–south axis and Gafsa–Jefara faults the
deepest faults coinciding withthe Moho flexures, which had an
important role in their initiation. They constitutethe border
intra-continental crust faults of the Mesozoic rift. The newly
recognizeddeep fault has critical implications for mineral and
petroleum perspectives.
INTRODUCTION
Crustal strike-slip faults are of considerable structural
im-portance and have an indirect economic implication becausethey
are of significance for recognition of sedimentary struc-ture
evolution. Crustal faults play a significant role in theformation
of:
- folds and faults, which commonly constitute structural trapsto
accumulate hydrocarbons;
- economic mineral deposits by controlling the migration
ofhydrothermal fluids.
The delineation of economic reserves of valuable mineral
de-posits and hydrocarbon accumulations include generally thestudy
of the relation between basement morphology and sed-imentary basins
and the relation between shallow and deepfaults.
Examination of Tunisian deep structures and their evo-lution has
required a combination of geological studies andlarge-scale
regional geophysical surveys. Thus, Tunisia has
∗E-mail: [email protected]
been the focus of several regional geophysics
investigationssince 1940. The first gravimetric campaigns were
completedin 1947, and a regional aero-magnetic survey covered all
ofTunisia in 1964. The requirement for a gravity map coveringthe
entire country was fulfilled by compiling data from vari-ous
gravimetric campaigns carried out by oil companies, theresults of
this work mainly discussed the structure of the litho-sphere
(Jallouli and Mickus 2000; Mickus and Jallouli 1999).Knowledge of
deep structures in Tunisia was enormously en-riched by the results
of a seismic survey carried out by GeoTraverse in 1985 (Buness et
al. 1992). Thus, regional geophys-ical surveys have contributed
greatly to the characterization ofdeep structures in Tunisia.
However, some interpretations areaffected by poor quality data, and
others are still hypotheticalor have been based on a single profile
model; therefore, theycannot be extended to the entire Tunisian
surface.
In spite of the poor data quality in certain profiles andthe low
sampling density, the airborne magnetic survey resultsreveal the
shape and depth of a magnetic basement situated ata depth of 2 km
in the extreme North of Tunisia, which may bein concordance with
Archaean granitic basement outcrops inEastern Algeria. In the
Atlas, the magnetic basement appears
1C⃝ 2015 European Association of Geoscientists &
Engineers
-
2 M. Arfaoui, A. Reid and M. H. Inoubli
with a convexity at a depth of 4 km, which separates
twodifferent tectonic zones in the South. The magnetic
basementincreases to a depth of 4 km in the Eastern Mednine
regionand south of Chotts, where its structure is interpreted as
ananticline with NW–SE axis.
Several authors have discussed the results of the
seismicreflection/refraction profile observed by the European
Geotra-verse Project for lithosphere recognition. Morelli and
Nicol-ich (1990) presented a cross section of the lithosphere
alongthe European Geotraverse southern segment from the Alps
toTunisia, which showed a Moho that deepens from a depthof 20 km on
the north coast to 35 km in the extreme northand 40 km in the
Atlas. Buness et al. (1992) proved in theirseismic model of an N–S
profile segment of Tunisia that thicksediment reached 14 km in the
Tunisian furrow and 12 kmin the Gafsa furrow. The authors presented
a tentative Mohocontour map of the surveyed area of Tunisia, in
which thecrust thickness was approximately 22 km in the north and
37km in the Atlas.
Mickus and Jallouli (1999) and Jallouli and Mickus(2000)
remodelled the N–S profile based on Tunisian gravi-metric data and
provided new information about the deepstructure that remained
associated with the profile, which can-not be generalized
elsewhere. In those studies, it was revealedthat:
(i) a thickened crust in the South of the Saharan flexure (32–38
km thick) coupled with a reduction of the Paleozoic andMesozoic
sediment thicknesses (2–3 km) and a crust thick-ening in the Atlas
of 35 km that decreases towards the north(26 km) contributing to
thick Palaeozoic, Mesozoic, andCenozoic sediments (12 km);
(ii) a thick Cenozoic filling of the basins in the Tunisian
furrow(1–2 km);
(iii) the Triassic is significantly thicker in the north than
thesouth and begins to increase in the north.
A residual anomaly model presented by Jallouli and Mickus(2000)
shows sediments and upper crust thickness variationsfrom the north
to the south of Tunisia without any indicationof the gravity effect
from major faults. None of the Tunisiagravity studies discuss the
large Bouguer anomaly gradientlineaments.
None of the geophysical studies used for the recogni-tion of
Tunisian deep structure indicate the presence of deepfaults,
although the major fault systems that affect the crustsuch as the
Jeffara–Gafsa, Teboursouk, and N–S faults havebeen the subject of
many geological studies since the workof Glangeaud (1951),
Dubourdieu (1956), Castany (1951),
Jauzein (1967), Zargouni (1985), Boukadi (1996), Martinezand
Truillet (1987) and Bouaziz et al. (2002).
In this paper, we use Woollard’s (1959) empirical
relationbetween crust thickness and Bouguer anomaly to determinethe
depth, shape, and flexure of Moho discontinuity over allof Tunisia.
Then, horizontal gradient grid peaks of Bougueranomaly and optimal
upward continuation have been per-formed to determine,
respectively, superficial and deep faultsystems.
TUNIS IAN GEOLOGY SETTING
Sited in North Africa, Tunisia includes geologic outcrops
thatextend in age from the first era to the recent quaternary.
Out-crop diversity results from the juxtaposition of two
differentAfrican domains sharply separated by the Jefara–Gafsa
faultsystem: the Alpine domain, including the Tell and Atlas
zonesstructured essentially at the Mesozoic and Cenozoic, and
thestable Saharan platform that belongs to the old African
base-ment. The nearest Archaean granitic basement outcrop ap-pears
in Eastern Algeria, which may be in concordance withthe basement of
Northern Tunisia.
The tectonic setting of Tunisia is an integral part of
NorthAfrican tectonics recognized mainly by:
(i) The Tethyan extensional event corresponding to EarlyMesozoic
rifting of the North Africa margin (Snoke,Schamel, and Karasek
1988) and massive subsidence dur-ing the Jurassic, Triassic and
early Cretaceous. This exten-sion has largely controlled
sedimentation by the activation ofNE–SW, NW–SE, and N–S deep
faults; limited extensionalevents were recognized equally at Late
Cretaceous and Ter-tiary times (Martinez and Truillet 1987).
(ii) The compressional tectonic event associated with theAlpine
orogeny remained active from the late Cretaceousto recent time. It
caused the inversion of normal faults, reac-tivation of uplifts,
and intrusion of Triassic evaporate diapirs(Ben Ferjani, Burrolet,
and Mejri 1990; Bouaziz et al. 2002).
Tectonic movements have controlled sedimentation bythe
activation of NE–SW, NW–SE, and N–S deep paleo-faultsand fashioned
structures by erecting anticlines, synclines, andcollapsed troughs.
Thus, five major structural zones can bedistinguished in Tunisia by
their historical geology and con-stituents, known from north to
south as: the nappe zone,diapiric zone, Atlas zone, the N–S Axis,
eastern platform, andSaharan platform (Fig. 1a). The boundaries of
these zones aregenerally marked by faults.
C⃝ 2015 European Association of Geoscientists & Engineers,
Geophysical Prospecting, 1–12
-
New regional fault and crustal structure 3
Figure 1 (a) Geological map of Tunisia. (b) Major structural
provinces.
In the extreme north, the nappe zone proposed by Jauzein(1967)
and Rouvier (1977) is characterized by allochthonousunits formed by
Numidian clays and sandstones. Local out-crops of acid volcanic
rocks and basalts occur in this zone(Rouvier 1977; Ouazaa
2000).
A diapiric zone formed in the Tunisian furrow dur-ing the
Mesozoic (Bolze, Burollet, and Castany 1952) andlargely deformed
during the compressional tectonic eventis distinguished by Triassic
evaporite outcrops that occupyanticline cores and follow major
faults (Perthuisot 1978,1981).
The Tunisian Atlas comprises thin platform sediments(Burollet
1956), which have been folded into large NE–SW an-ticlines and
synclines. Thereafter, they have been cut by NW–SE border faults of
Mio-Plio-Quaternary troughs (Jauzein1967; Turki 1985).
The N–S axis composed of N–S structures correspondsto a
sedimentary limit between two domains, which have dif-ferent
paleogeographic settings, i.e., the Atlas zone and theEastern
platform. This alignment formed a paleogeographicfeature since the
Jurassic as the reductions or condensations ofthe sedimentary
sequences often show gaps and discordances.
The Eastern platform, extending east of the N–S axis,is
characterized by a slow subsidence during the Mesozoicand more
active subsidence during the Cenozoic (Burolletand Byramjee 1974).
At surface, the eastern platform presentsfolding tectonics marked
by large Mio-Plio-Quaternary foldsand appears unaffected by
faulting but, in depth, shows astructure of horsts and grabens,
which are the result of NE–SW, NW–SE, and N–S fault activities
(Fig. 1b) (Haller 1983).
The Saharan platform zone, fringed to the north by theGafsa
fault and the huge endorheic salted basins of El Jerid and
C⃝ 2015 European Association of Geoscientists & Engineers,
Geophysical Prospecting, 1–12
-
4 M. Arfaoui, A. Reid and M. H. Inoubli
El Fejej Chotts, remained stable during the whole Mesozoicand
Cenozoic, and was affected only by epeirogenic move-ments (Bishop
1975).
A review of the vicinity zones of Tunisian major faultsystems
reveals important contrasts and gaps in sedimentaryseries,
strike-slip structure, overlapping, flexures, unconfor-mities, and
facies changes. In some cases these faults haveconstituted an
important paleogeographic limit. Other faultsextend by Triassic
salt structures (such as the Teboursoukfault) or associate with
volcanism and hydrothermalism. Fol-lowing their directions, these
major faults can be grouped asfollows.
NE–SW major faults essentially dominate the Alpine do-main. The
Zaghouane fault shows an overlapping to thesoutheast. This fault
was primarily active in the Jurrassic.The Tunis–Elles fault
continued to Bakaria in Algeria, witha length of 250 km (Jauzein
1967). This fault presents gaps,thinning, and local overlapping
caused by reverse faults. TheTeboursouk fault extends by the
Triassic salt structures atLansarine, Fej Lahdhoum, Kebouche, and
is characterizedby the overlapping of the Tebousouk. The
Guardimaou–CapSerat fault separates the Algerian and Tunisian blocs
and ismarked by volcanic rock manifestations and diapirism.
NW–SE major faults appear in the southern part ofTunisia. They
include the Jefara–Gafsa fault system that sep-arates the Atlas
zone from the Saharan platform. This sys-tem contains the Gafsa and
Mednine faults. The Mio-Pliocenebasin border faults also follow the
NW–SE direction (Fig. 2).
T U N I S I A N G R A V I T Y M A P
Between 1948 and 1967, oil companies conducted many sur-veys
over their permits, e.g., SEREPT, i.e., Sahel (1948), CapBon (1950)
and Kairouane (1967) permits; AGIP, i.e., Sudand Bir Aouine permits
(1951); and Mobil, i.e., Gafsa andGabès permits (1951). There are
many more. To establish theregional Tunisian gravity map from these
data, Petroleum Ac-tivity Tunisia Enterprise (ETAP, 1982),
completed a gravitysurvey in the extreme north and compiled all the
data to es-tablish the gravity map with approximately one station
per25 km2.
The compilation operation was used to homogenize allthe data.
Field gravity data were corrected for Earth tides andinstrumental
drift and converted to absolute gravity values.The international
gravity Formula of 1967 is used to reducethe field data. Then, the
usual free air, Bouguer, and ter-rain corrections were applied; the
reduction density used was2.67 g/cm3.
Figure 2 Tectonic Tunisia map adapted from Bouaziz et al.
(2002)and Tunisia geological map.
The Tunisian gravity map used in this paper was estab-lished
using a grid generated from the digitized points of theBouguer
anomaly map presented by ETAP (1982). This gridis accepted after a
cross-validation of data based on a sta-tistical analysis of the
differences between digitized Bougueranomaly and calculated grid
values, which gives an averageof the differences of 0.01 mGal and a
standard deviation of1.06 mGal. The values of the Bouguer anomaly
vary between−85 mGal in the western part of central Tunisia and 40
mGalin the northern part of Tunisia. They are generally compa-rable
with those of the map established in 1982 (−86 mGal
C⃝ 2015 European Association of Geoscientists & Engineers,
Geophysical Prospecting, 1–12
-
New regional fault and crustal structure 5
Figure 3 Bouguer anomaly map of Tunisia.
and 47 mGal). The Bouguer anomaly map adopted from thisstudy is
dominated by:
- long-wavelength negative anomalies in the Tunisian Atlas,the
Chotts Zone, and the Oriental Erg Basin;
- long-wavelength positive anomalies associated with the
Tell,the Sahel, and the southeastern coastal zones. Other
positiveanomalies of maximum amplitude of -10 mGal occupy
theAlgerian anticlinorium (Fig. 3).
High gradients of the Bouguer anomaly are associated
withsignificant structural alignment. It is clearly observed in
thelevel of the N–S axis, the South of the Chott El Jerid, and
be-tween the Algerian anticlinorium and the Oriental Erg Basin.The
most outstanding high-gradient zone is the Kairouane–Ghardimaou
NW–SE alignment, which appears affected inthe north by other
alignments. This high gradient constitutesthe gravity effect of a
major fault that can be expected to af-fect the upper and lower
crusts. We discuss this alignment inmore details later.
The regional aspect of the Tunisian Bouguer anomalymap is
characterized by long wavelengths that reflect only thegravity
effects of the thick sedimentary units, combined withmajor faults
and density contrast effects inside the lithosphere.Thus, the
Bouguer anomaly map is largely correlated with themajor structural
province map of Tunisia (Jallouli and Mickus2000). The gravity
anomalies of the Tell and the TunisianAtlas are mainly due to
variations in the thickness and densityof the sedimentary
sequences.
The positive anomaly associated with the Sahel is limitedto the
West by the high gradient defining the North–Southaxis. This axis
is observed throughout the Sahel. It is not sim-ply a response to a
change in density of Mio-Plio-Quaternaryoutcrops dominating this
region. It is due probably to the ris-ing of the basement towards
the east and to a thinning of thecrust.
The Oriental Erg Basin presents an anomaly of −60 mGalcrossing
the image of the Tunisian part of the Illizi Basin,which extends
from the Southeast of Algeria in the Westthrough to Libya in the
east (Jallouli and Mickus 2000). TheAlgerian Anticlinorium is
represented by a relatively positiveanomaly of −10 mGal.
According to the above discussion, we identify fiveanomalous
zones on the Bouguer anomaly map where bound-aries are indicated by
at least one high gradient zone. Thesezones are: the Tell, the
Tunisian Atlas, the Sahel, the AlgerianAnticlinorium, and the
Oriental Erg Basin.
To better identify the major outcrop structures, a Gaus-sian
regional/residual filter with a standard deviation of0.007 km−1, is
applied in the wavenumber domain to theBouguer anomaly to determine
the residual anomaly associ-ated with relatively shallow
sources.
The Gaussian regional/residual filter is a low-pass fil-ter and
a high-pass filter characterized by a smoother cutoffprocess.
Residual and regional results from Gaussian filter-ing are similar
to high-pass and low-pass filters, respectively,in that they
attenuate low- and high-wavenumber compo-nents. Mathematically, a
Gaussian regional filter anomaly isobtained by convolution of the
Bouguer anomaly with a Gaus-sian function in the space domain. The
regional and residualGaussian filters are implemented in the
wavenumber domainand are given, respectively, in the following:
Regional filter : L(k) = e
(k2
2k20
)
, (1)
Residual filter : L(k) = 1 − e−
(k2
2k20
)
(2)
C⃝ 2015 European Association of Geoscientists & Engineers,
Geophysical Prospecting, 1–12
-
6 M. Arfaoui, A. Reid and M. H. Inoubli
Figure 4 Residual anomaly map after Gaussian filtering. A white
lineindicates the Triassic outcrops boundary.
where L(k) is the Gaussian filter operator in spectral domain,k
is the wavenumber in cycles km−1, and k0 is the standarddeviation
of the Gaussian function in km−1.
The Gaussian residual anomaly appears to be the mostuseful to
explain Tunisian geological outcrops. The residualand geological
maps present significant similarities.
The residual anomaly shows short-wavelength anomaliesthat show
values ranging from −15 mGal to 9 mGal, with thepresence of strong
gradients (Fig. 4). This map reveals certaindominant anomaly
directions associated with the principalstructural zones: the NE–SW
direction characterizes northernTunisia and the Sahel; the NW–SE
tendency dominates thenorth part of the Atlas and the Saharan
flexure; the E–Wdirection characterizes the Oriental Erg and the
Chotts zone.
Thus, the structural zones of Tunisia are clearly iden-tifiable
on the residual anomaly map (Fig. 4). The extreme
northern zone is occupied by NE–SW anomalies having gen-erally
positive amplitudes. They are limited towards the southby an E–W
band of negative anomalies, which also consti-tute the northern
limit of a succession of negative and positiveNE–SW anomalies,
translating the effects of the structuresof the Tunisian furrow.
The spatial shift of these anomaliesfollowing the NE–SW direction
is probably due to densitydiscontinuities, which particularly
affect the NW–SE and N–Sdirections. The majority of the Triassic
outcrops of NorthernTunisia are in relation to the anomaly
boundaries occupiedby NE–SW discontinuities (Fig. 4).
The NW–SE negative anomalies of the northern partof the Atlas
coincide with Mio-Plio-Quaternary basin fill-ings; they are
separated by excess mass anomalies associatedwith uplift
structures. The Sahel, dominated by Mio-Plioceneand Quaternary
outcrops (Fig. 1b), is represented by deficitsand excesses in mass
anomalies that are oriented NE–SW and,respectively, associated with
collapse and uplift structures af-fected by faults. The Sfax area
is occupied by a relativelybroad anomaly of 1 mGal. In the west, an
N–S anomaly witha maximum amplitude of 9 mGal is associated with
the N–Saxis structures. This feature disappears towards the north
inthe level Kairouan area; it appears to be stopped by a
NW–SEnegative anomaly in relation to the zone of a high
Bougueranomaly gradient described previously (Figs 3 and 4).
The Chotts zone is represented by E–W anomalies, gen-erally
uncorrelated with the salt nature of the outcrops. Thus,except for
the Chott El Gharsa, which is superimposable withan E–W negative
anomaly of minimum of −6 mGal, Chott ElJerid and Chott El Fejej are
related partially or completely topositive anomalies.
The northern part of the Chott El Jerid is superimpos-able on
three negative anomalies organized on a flattened arcshape, which
is continuous with the southern part of ChottEl Fejej. These
anomalies could be explained as basins sep-arated by faults, which
is an interpretation that is in agree-ment with the asymmetrical
syncline under the Chott (WesternPetroleum Corporation 1967;
Guederi 1980). The southernpart of Chott El Jerid is superimposed
on a positive anomalyindicating a depth structure signal different
to that observedin the northern part. Chott El Fejej is associated
with a posi-tive anomaly reaching a maximum of 5 mGal according to
aneroded anticline substratum of the Chott (Western
PetroleumCorporation 1967; Guederi 1980).
Starting from Gabès, the coast of southern Tunisia showsa broad
positive residual anomaly uncorrelated with outcrop-ping Quaternary
deposits. It confirms the presence of the densestructures
associated with the Medenine fault. The anomalous
C⃝ 2015 European Association of Geoscientists & Engineers,
Geophysical Prospecting, 1–12
-
New regional fault and crustal structure 7
arc of a deficit of mass on the level of the northern part of
theChott El Jerid, as previously discussed, continues towards theSE
simultaneously with this anomaly. The Triassic outcropin this zone
is superimposed on a negative anomaly, and itappears to be
continuous in depth under the Sebkha of OumEl Khielat.
The Oriental Erg is presented by two negative residualanomalies:
an E–W anomaly in the North and an oval formanomaly in the South;
they are the gravity response to twosedimentary basins separated by
an uplift zone.
T U N I S I A N C R U S T A L S T R U C T U R E
Isostatic anomalies
The comparison between Tunisia Bouguer anomaly map andthe
topographic one shows that long-wavelength Bougueranomalies
correlate inversely with long-wavelength topogra-phy: A significant
negative anomaly covers the Tunisian Atlasmountains and a positive
anomalous band along the coastalzones of subdued topography. This
supports the notion ofisostatic compensation. The gravity effect of
the relief is com-pensated at depth by a deficit of mass reflecting
a thick crustwith a density that is relatively low compared with
mantledensity. The effects of the mass deficits at the base of the
crustgenerally mask the gravity signal associated with
shallowercrustal sources.
The regional isostatic anomaly map of Tunisia is estab-lished
using the local compensation Airy model based on to-pography, with
a density of the land of 2.67 g/cm3, a contrastof density through
the Moho of 0.35 g/cm3, and a thicknessof the normal crust of 30
km. This value is averaged fromprevious crust studies and the Moho
map below determinedfrom gravity data (Fig. 6). The regional
isostatic map generallypresents long-wavelength anomalies
associated with densitycontrasts between the mantle and the lower
crust (Fig. 5a). Itshows a broad anomaly of mass deficit on the
Tunisian Atlasreaching −74 mGal, closed in the east at an alignment
coin-ciding with the N–S axis and continuous in the West
(SaharianAtlas in Algeria). This isostatic anomaly is the response
to thesignificant change of crustal thickness in this zone.
AnotherN–S negative anomaly is observed in the south with
relativelylow amplitude (−40 mGal). The remainder of the country
iscovered by a positive regional isostatic anomaly.
The residual isostatic anomaly was calculated by elimi-nating
the gravity effects of large wavelength caused by theisostatic
variations associated generally with the lower crustsources.
Table 1 Difference range between Moho depths determined from
thethree formulas.
Formulas Difference range
Woollard 1959 – Demenitskaya 1958 0.84 – 7.33Woollard and
Strange 1962 – Woollard 1959 0 –3.35Woollard and Strange 1962 –
Demenitskaya 1958 2.19–3.98
The long wavelength of some isostatic residual anomaliesexplains
the gravity effects of the upper crust (Fig. 5b). Thus,the Atlas
trend (NE–SW) appears in the north zone. The Sahel,north zone, and
the Algerian Anticlinorium are occupied bypositive anomalies,
whereas the Chotts Zone and the OrientalErg are associated with
negative anomalies, which reveal thelow density nature of the
sedimentary sequences (Fig. 5b).
Moho morphology
The Bouguer anomaly usually contains the effect of all
litho-spheric density variations. It is therefore normal to seek
theeffects of crustal structure and major deep faults in the
grav-ity data. In this paper, crustal structure relating
essentially tothe depth and morphology of the Moho discontinuity is
ap-proached by the empirical relationship, which is deduced
bycomparing the gravity anomaly and seismically determinedcrustal
thickness proposed by Woollard (1959) and used byLiu and Yen
(1975), Ram Babu (1997), Rivero, Pinto, andCasas (2002),
Demenitskaya (1967), Arslan, Akýn, and Alaca(2010), and many
others.
The formula proposed by Woollard (1959) is one amongvarious
empirical formulas that are generated from the corre-lation between
gravity and Moho depth deduced from seismicobservations in distinct
regions.
We determine the Moho depth using two formulas pro-posed by
Demenitskaya (1958) and Woollard and Strange(1962). We compare
these results with those obtained usingthe proposed formula below
(Woollard 1959) (Table 1).
The Moho depth calculated by the Woollard formulais the closest
to the Moho seismic depth following the N–Srefraction/reflection
profile, from the Geotraverse Europeanproject (Buness et al. 1992),
and we have therefore chosen touse this model for further work.
The Moho discontinuity depth is estimated by thefollowing
empirical Woollard (1959) formula (3) deducedexperimentally from
the linear relation between seismic
C⃝ 2015 European Association of Geoscientists & Engineers,
Geophysical Prospecting, 1–12
-
8 M. Arfaoui, A. Reid and M. H. Inoubli
Figure 5 (a) Regional isostatic anomaly map, (b) Residual
isostatic anomaly map.
measurement depth to the Moho discontinuity and Bougueranomaly
proved in various parts of the world.
Hm = 32 − 0.08g, (3)
where Hm is the Moho depth in km, and g is the regionalgravity
anomaly in mGal.
The depth of Moho varies between 29 km and 39 km. Thecrust has a
maximum thickness in the Tunisian Atlas region,where the Moho
discontinuity deepens gradually towards theWest to reach 39 km.
This thickening of the crust continues inAlgeria (Saharian Atlas).
The Moho depth map also shows arelative thinning of the crust in
the North of Tunisia (29 km)and in the Sahel (31 km) (Fig. 6).
The comparison of the Moho depth estimated by thepresent study
with the Moho depth estimated by Buness et al.(1992) and Mickus and
Jallouli (1999), according to an N–S
profile, shows a similarity between values with a differencesof
2 km in the South and the Atlas and approximately 5 kmin the North.
On the other hand, the values of the Mohodepth presented are
roughly comparable with those of theMoho depth map of Europe based
on seismic measurements(Molinari and Morelli 2011), which show a
variation of Mohodepth between 30 km and 40 km beneath Tunisia.
The Moho map shows a typical morphology character-ized by a
depression in the Tunisian Atlas and in the south.Elsewhere, the
surfaces present an uplift zone (Fig. 6). Thepassage between the
depression and uplift zones is markedby a flexure of the Moho
surface that coincides in the eastand the south with known faults,
namely the N–S axis faultand the Gafsa fault. The north flexure
must mark an un-known fault, which we name the Kairouan–Ghardimaou
fault(Fig. 6).
C⃝ 2015 European Association of Geoscientists & Engineers,
Geophysical Prospecting, 1–12
-
New regional fault and crustal structure 9
Figure 6 Depth of the Moho discontinuity. NDF: new deep
fault,NSAF: North South axis fault, GJF: Gafsa Jeffara fault. The
Gafsaand Mednine faults constitute the Gafsa -Jefara fault.
SUPERFICIAL A N D DE EP DEN SI T YDISCONTINUI T I ES
The horizontal gradient map computed from the Gaussianresidual
grid shows high gradient features, which constitutezones of density
discontinuity (Fig. 7). The Blakely methoduses the horizontal
gradient peaks to delineate abrupt lateraldensity changes. They are
generally interpreted as faults orcontacts (Blakely and Simpson
1986). Taking into accountthe inhomogeneity of the measurement
distribution and thegravity map scale, horizontal gradient peaks
revealed by theBlakely method are likely to be responses to major
deep faultsassociated with a significant density contrast.
Thus, the major known normal fault systems of Tunisiashould be
and are detectable on the horizontal gradient peakmap: the Mednine
fault with an NW–SE direction appears par-allel to the southern
coast and continues beyond the Tunisian–Libyan border. The absence
of the alignment correspondingto the Gafsa fault is due to the poor
gravity data cover of thiszone. The Chotts zone is characterized by
E–W, NW–SE, andN–S alignments related the faults affecting geologic
structuresat depth (Fig. 7).
Figure 7 Horizontal gradient map. The black lineaments
correspondto faults and discontinuities of density. NDF: new deep
fault, NSAF:North-South axis fault, GF: Gafsa fault, MF: Mednine
fault. TheGafsa and Mednine faults constitute the Gafsa Jefara
fault.
The southern part of the Tunisian Atlas is dominated byE–W, N–S,
and NE–SW faults, whereas the northern part ofthis area is marked
by NW–SE discontinuities coinciding withbasin border faults. These
NW–SE faults, cut by the NE–SWfaults in the NW of the Meknassi
area, appear in relationto the discontinuities of the same
direction and alignment inthe Sfax–Mahres area. To the north of the
latter area, NE–SWand NW–SE discontinuities affect the
non-outcropping Creta-ceous. An N–S discontinuity extends in the
West of the Sahel;it corresponds to the N–S axis and constitutes a
structurallimit at which the NE–SW and NW–SE faults of the Atlas
andthe Sahel terminate (Fig. 7).
The N–S discontinuity disappears in the Kairouan plainarea, and
the previously described NW–SE discontinuity fromKairouan to
Ghardimaou takes its place. It represents a
C⃝ 2015 European Association of Geoscientists & Engineers,
Geophysical Prospecting, 1–12
-
10 M. Arfaoui, A. Reid and M. H. Inoubli
Figure 8 (a) Cross-correlations between continuations to two
successive heights versus the height, (b) Deflection (c) versus the
upward contin-uation height.
regional deep fault. This lineament coincides with the
high-gravity-gradient zone in the north of El Kef, which has
con-stituted a paleogeographic boundary during Barremian
time(Chikhaoui 2002; Arfaoui et al. 2011). This
non-outcroppingfault is not described by the geologists; it affects
deep struc-tures and crust (Fig. 7).
Tunisia presents a tectonic regime typically characterizedby
NW–SE and NE–SW faults. The majority of the Triassicoutcrops are
associated with NE–SW discontinuities (Triassicalignments of
Dabadib, Tibar, and Lansarin, among others).
To determine deep faults or crustal faults, we determinethe
horizontal gradient peaks of the upward continuation ofthe Bouguer
anomaly at an optimum height estimate by themethod of Zeng, Xu, and
Tan (2007). This method is basedon the cross-correlation factor r
between Bouguer anomalyupward continuations to two successive
heights. The cross-correlation factor between two Bouguer anomaly
continu-ations g1 and g2 is evaluated by the formula (4) given
byAbdelrahman et al. (1989):
r (g1, g2) =∑M
i
∑Nj g1(xi , yj )g2(xi , yj )√∑M
i
∑Nj g
21(xi , yj )
∑Mi
∑Nj g
22(xi , yj )
, (4)
where M and N are the number of columns and rows ofBouguer
anomaly grids, respectively.
The cross-correlation is plotted as a function of the up-ward
continuation height (Fig. 8a). The deflection definedby the
difference between the cross-correlation curve and thechord joining
its two end points is plotted equally as func-tion of the height.
Optimal upward continuation height cor-responds to maximum
deflection is deduced directly from thedeflection curve (Fig.
8b).
In our case, the upward continuation is calculated forheights
varying from 5 km to 95 km with a step of 5 km.The optimum upward
continuation height estimated by themethod of Zeng et al. (2007)
was 30 km (Fig. 8b).
The different regional anomalies inferred from lesser andgreater
heights, compared with the optimal one, are domi-nated by the
gravity effects of near-surface sources and deep-seated geological
bodies, respectively (Zeng et al. 2007).
The NW–SE Kairouane–Ghardimaou, N–S axis, and theMednine faults
persist and constitute the deep structural pat-tern of Tunisia. The
Kairouane–Ghardimaou fault, revealedin this study, continues into
Algeria to the west and into theMediterranean Sea to the east. The
three faults constitute asignificant limit in the Moho depth map
that separates a thincrust zone in the east and north from a thick
crust zone inthe west, from which it can be deduced that this limit
consti-tutes a Moho flexure origin of these major faults (Fig. 9).
TheMednine fault and associated E–W faults affect the
Palaeozoicbasement (Ben Ayed 1986). The two deepest NW–SE faultsare
equally associated with the Sirte Basin opening.
In order to determine the crustal deformation caused bythe
current collision in the Alpine–Himalaya orogeny, Sok-outis et al.
(2000) built an experimental rheological modelcomposed of three
layers: upper crust with a change in thethickness, lower crust, and
asthenosphere mantle. Then, theysubmitted the model to a shortening
force. Thus, four ma-jor belt systems appeared at different stages
of the shorteningexperiment. The new deep fault
(Kairouane–Gardimaou) cor-responds to the strike-slip fault
affecting the N–S axis in thetectonic Tunisia map presented by
Sokoutis et al. (2000) thatcaused the deviation of the N–S axis
direction to the east ofthe part sited north of Kairouan.
C⃝ 2015 European Association of Geoscientists & Engineers,
Geophysical Prospecting, 1–12
-
New regional fault and crustal structure 11
Figure 9 Horizontal gradient after optimal upward continuation
ofthe Bouguer anomaly (30 km). The black dashed lines indicate
deepfaults.
The superficial effects of the NW–SE deep fault aremarked by the
E–W collapsed structures of the Gaafour–ElAroussa graben and the
Mejarda basin bordered by normalfaults and marked at their southern
borders by structures withno detected anticlinal ending (Figs. 1a
and 2).
The tectonic sketch of Tunisia and the Pelagian Sea pre-sented
by Burollet (1991) shows an offshore of Eastern Tunisiadominated by
NW–SE strike-slip faults and grabens orientedin the same direction
as the new deep fault. Burollet (1991) re-vealed that the NW–SE
strike-slip faults and grabens resultedfrom basement extension
associated with shear faults. Thissuggests and confirms the
presence of the new deep fault.
The new NW–SE fault with the N–S axis and the Med-nine faults
constitute an old conspicuous crustal limit of theNorthern and
Eastern Tunisia margins, which resulted from
early Mesozoic extensional events. They represent equally
thewest border faults or the intra-continental crust faults of
theMesozoic rift whose centre is apparently in an offshore
thinnercrust zone (Molinari and Morelli 2011).
Thus, Mesozoic rifting in the Sahel platform and in thenorth is
confirmed by volcanic rocks found in the Trias-sic, Jurassic, and
Cretaceous series in oil wells and outcrops(Ouazaa 2000). In these
active crust-thinning stages, N–S andNW–SE rifting control faulting
occurs, respectively, in theSahel platform and in Northern Tunisia.
An active post riftsubsidence is recognized during the Mesozoic
north of thenew Kairouane–Ghardimaou deep fault in the Tunisian
Fur-row where more then 5000 m of Cretaceous series are
ac-cumulated (north of El Kef) (Burollet 1956; Jauzein
1967;Chikhaoui 2002). Post-rift subsidence is recognized
mainlyduring Cenozoic time in the Eastern platform (Burollet
andByramjee 1974).
CONCLUSIONS
The major deep faults known in Tunisia as the Mednine andGafsa
faults are recognized by the horizontal gradient peaks,with a new
deep fault oriented NW–SE, which runs betweenKairouan and
Ghardimaou. Crustal structure extended tothe entire Tunisian
territory based on the morphology of theMoho discontinuity shows
conspicuous flexure zones in thesurfaces of the latter, which
coincide with the deepest faults:NW–SE Kairouan–Ghardimaou, N–S
axis, and the Medninefaults. Thus, the new NW–SE
Kairouan–Ghardimaou fault re-vealed by this study is a deep crust
fault; it is an important ge-ological significance with critical
mineral and oil explorationimplications.
ACKNOWLEDGEMENTS
We would like to thank the Editor Prof. Maurizio Fedi, and
theAssociate Editor and anonymous reviewers for their
numerousconstrictive suggestions that improved the final version of
thismanuscript substantially.
REFERENCES
Abdelrahman E.M., Bayoumi A.I., Abdelhady Y.E., Gobashy M.M.and
El-Araby H.M. 1989. Gravity interpretation using correlationfactors
between successive least-squares residual anomalies. Geo-physics
54, 1614–1621.
Arfaoui M., Inoubli M.H., Tlig S. and Alouani R. 2011.
Gravityanalysis of salt structures. An example from the El
Kef-Ouargharegion (northern Tunisia). Geophysical Prospecting 59,
576–591.
C⃝ 2015 European Association of Geoscientists & Engineers,
Geophysical Prospecting, 1–12
-
12 M. Arfaoui, A. Reid and M. H. Inoubli
Arslan S., Akýn U. and Alaca A. 2010. Investigation of crustal
struc-ture of Turkey by means of gravity data. Bulletin of the
MineralResearch and Exploration 140.
Ben Ayed N. 1986. Evolution tectonique de l’avant-pays de la
chainealpine de Tunisie du début du Mésozoı̈que à l’actuel. PhD
thesis,Université de Paris Sud, Centre d’Orsay.
Ben Ferjani A., Burrolet P. and Mejri F. 1990. Petroleum geology
ofTunisia. Proceedings of the 3rd Tunisian Petroleum
ExplorationConference, Tunisia, pp. 194.
Bishop W.F. 1975. Geology of Tunisia and adjacent parts of
Algeriaand Libya. American Association Geologists Bulletin 59,
1033–1058.
Blakely R.J. and Simpson R.W. 1986. Approximating edges of
sourcebodies from magnetic or gravity anomalies. Geophysics 51,
1494–1498.
Bolze J., Burollet P.F. and Castany G. 1952. Le Sillon Tunisien,
2ndedn. XIXe Congrès Géologique International.
Bouaziz S., Barrier E., Soussi M., Turki M.M. and Zouari H.
2002.Tectonic evolution of the northern African margin in Tunisia
frompaleostress data and sedimentary record. Tectonophysics 357,
227–253.
Boukadi N. 1996. Schéma structural nouveau pour le Nord de
laTunisie. Proceedings of the 5th Tunisian Petroleum
ExplorationConference, Tunisia, Expanded Abstracts, 91.
Buness H., Giese P., Bobier C., Eva C., Merlanti F., Pedone R.
et al.1992. The EGT’85 seismic experiment in Tunisia: a
reconnaissanceof deep structures. Tectonophysics 207, 245–267.
Burollet P.F. 1956. Contribution à l’étude stratigraphique de
laTunisie centrale. Annales des Mines et de la Géologie,
Tunis.
Burollet P.F. 1991. Structures and tectonics of Tunisia.
Tectono-physics 195, 359–369.
Burollet P.F. and Byramjee R. 1974. Evolution
géodynamiquenéogène de la Méditerranée occidentale. Comptes
Rendus del’Academie des Sciences 278, 1321–1324.
Castany G. 1951. L’orogenèse de l’Atlas tunisien. Bulletin de
la Sociétégéologique de France 6, 701–720.
Chikhaoui M. 2002. La zone des diapirs en Tunisie: Cadre
structuralet évolution géodynamique de la sédimentation
méso-cénozoı̈queet géométrie des corps triasiques. PhD thesis,
Es Sciences Universitéde Tunis el Manar.
Demenitskaya R.M. 1958. Planetary structures and their
reflection inBouguer anomalies. Soviet Geology 8, 312–319.
Demenitskaya R.M. 1967. Crust and Mantle of the Earth.
Nedra,Moscow, Russia.
Dubourdieu G. 1956. Etude géologique de la région de
l’Ouenza(confins algéro-tunisiens). PhD thesis, Faculté des
Sciences d’Alger.
Glangeaud L. 1951. Interprétation tectono-physique des
caractèresstructuraux et paléogéographiques de la Méditerranée
occidentale.Bulletin de la Société géologique de France 6,
735–762.
Guederi M. 1980. Géochimie des sels et des saumures du chott
ElJerid (Sud tunisien). PhD thesis, Université de Toulouse.
Haller P. 1983. Structure profonde du Sahel tunisien
interprétationgéodynamique. PhD thesis, Fac. SC. et Tech. de
l’Université deFranche-Comté.
Jallouli C. and Mickus K. 2000. Regional gravity analysis of
thecrustal structure of Tunisia. Journal of African Earth Sciences
30,63–78.
Jauzein A. 1967. Contribution à l’étude géologique des
confins de ladorsale tunisienne (Tunisie septentrionale). Annales
des Mines etde la Géologie No. 22, Tunis.
Liu C.C. and Yen T.P. 1975. Bouguer anomaly, surface
elevationand crustal thickness in Taiwan. Petroleum geology of
Taiwan 12,97–105.
Martinez C. and Truillet R. 1987. Evolution structurale
etpaléogéographique de la Tunisie. Mémoire de la Société
géologiqued’Italie 38, 35–45.
Mickus K. and Jallouli C. 1999. Crustal structure beneath the
Telland Atlas mountains (Algeria and Tunisia) through the analysis
ofgravity data. Tectonophysics 314, 373–385.
Molinari I. and Morelli A. 2011. EPcrust: A reference crustal
modelfor the European plate. Geophysical Journal International
185,352–364.
Morelli C. and Nicolich R. 1990. A cross section of the
lithospherealong the European Geotraverse Southern Segment (from
the Alpsto Tunisia). Tectonophysics 176, 229–243.
Ouazaa Laaridhi N. L. 2000. Étude minéralogique et
géochimiquedes épisodes magmatiques mésozoiques et miocènes de
la Tunisie.PhD thesis, Es Sciences Université de Tunis el
Manar.
Perthuisot V. 1978. Dynamique et pétrogenèse des extrusions
tri-asiques en Tunisie septentrionale. PhD thesis, Ecole
NormaleSupérieure, Paris.
Perthuisot V. 1981. Diapirism in Northern Tunisia. Journal of
Struc-tural Geology 3, 231–235.
Ram Babu H.V. 1997. Average crustal density of the Indian
litho-sphere: an inference from gravity anomalies and deep
seismicsoundings. Journal of Geodynamics 23, 1–4.
Rivero L., Pinto V. and Casas A. 2002. Moho depth structure of
theeastern part of the Pyrenean belt derived from gravity data.
Journalof Geodynamics 33, 315–332.
Rouvier H. 1977. Géologie de l’extrême-nord tunisien:
tectoniqueset paléogéographies superposées l’extrémité
orientale de la chaı̂nenord-maghrébine. PhD thesis, Université
Paris VI.
Sokoutis D., Bonini M., Medvedev S., Boccaletti M., Talbot C.J.
andKoyi H. 2000. Indentation of a continent with a built-in
thicknesschange: experiment and nature. Tectonophysics 320,
243–270
Snoke A., Schamel S. and Karasek R. 1988. Structural evolution
ofthe Jebel Debadib anticline: A clue to the regional tectonic
style ofthe Tunisian Atlas. Tectonics 7, 497–516.
Turki M.M. 1985. Polycinématique et contrôle sédimentaire
associéssur la cicatrice de Zaghouan-Nebhana. PhD thesis, Faculté
des Sci-ences de Tunis.
Western Petroleum Corporation 1967. Chott El Jerid Projet
PotasseTunisien. Unpublished Report ONM.
Woollard G.P. 1959. Crustal structure from gravity and seismic
datameasurements. Journal of Geophysical Research 64,
1524–1544.
Woollard G.P. and Strange W.E. 1962. Gravity anomalies and
crustof the earth in the Pacific basin. Geophysical Monograph 6,
12.
Zargouni F. 1985. Tectonique de l’Atlas méridional de
Tunisie,évolution géométrique et cinématique des structures en
zone decisaillement. PhD thesis, Université Louis Pasteur
Strasbourg.
Zeng H., Xu D. and Tan H. 2007. A model study for
estimatingoptimum upward- continuation height for gravity
separation withapplication to a Bouguer gravity anomaly over a
mineral deposit,Jilin province, northeast China. Geophysics 72,
I45–I50.
C⃝ 2015 European Association of Geoscientists & Engineers,
Geophysical Prospecting, 1–12