Download Evolving deformation along a transform plate boundary: Example

Click
Here
TECTONICS, VOL. 28, TC5005, doi:10.1029/2008TC002316, 2009
for
Full
Article
Evolving deformation along a transform plate boundary: Example
from the Dead Sea Fault in northern Israel
Ram Weinberger,1 Michael R. Gross,2 and Amihai Sneh1
Received 18 April 2008; revised 19 May 2009; accepted 6 July 2009; published 22 September 2009.
[1] We analyzed geologic structures adjacent to the
Dead Sea Fault (DSF) along the margins of the Sinai
and Arabian plates in northern Israel in order to
investigate the style and sequence of deformation
associated with a transform plate boundary. The field
area, located between the Hula basin in northern Israel
and the Lebanese restraining bend in southern Lebanon,
is divided into distinct structural blocks by a series
of distributed faults that comprise this approximately
N-S trending section of the DSF. Cretaceous and
Tertiary rocks within and adjacent to the structural
blocks are folded into broad anticlines and synclines,
with more intense localized shortening manifested by
tight folds and thrust duplexes. Kinematic analyses of
folds, faults, and veins provide evidence for two
directions of regional shortening: (1) NW-SE
shortening responsible for the formation of NE-SW
trending fold axes and left-lateral strike-slip motion
along N-S trending faults and (2) E-W shortening as
indicated by N-S trending fold axes, N-S striking
thrust faults, and extensional calcite-filled veins that
strike E-W. Crosscutting relations and U-Th ages of
the vein material suggest that the E-W phase of
transform-normal shortening represents the most recent
and presently active phase of deformation. The
structural analysis provides evidence for the transition
from an early (Miocene–lower Pliocene) phase of
pure strike-slip motion to a late (Pleistocene to
Recent) phase of convergent strike slip. The latter
phase is characterized by strain partitioning, which is
manifested by discrete left-lateral strike-slip motion
across weak N-S faults and the development of a
fold-thrust belt in response to transform-normal
shortening. Analogous to the strain partitioning
observed in southern California, we suggest that blind
thrust faults adjacent to the DSF in the study area may
pose a seismic risk to populations in northern Israel and
southern Lebanon. Citation: Weinberger, R., M. R. Gross,
and A. Sneh (2009), Evolving deformation along a transform plate
1
Geological Survey of Israel, Jerusalem, Israel.
Department of Earth Sciences, Florida International University, Miami,
Florida, USA.
2
Copyright 2009 by the American Geophysical Union.
0278-7407/09/2008TC002316$12.00
boundary: Example from the Dead Sea Fault in northern Israel,
Tectonics, 28, TC5005, doi:10.1029/2008TC002316.
1. Introduction
[2] Continental transforms are dominated by strike-slip
motion, often resulting in net horizontal displacements that
exceed 100 km. Oblique plate convergence and divergence
relative to the transform boundaries may lead to localized
regions of contraction or extension, commonly referred to as
transpression or transtension, respectively [Aydin and Nur,
1982; Woodcock and Fischer, 1986; Zoback et al., 1987;
Sylvester, 1988; Teyssier et al., 1995; Dewey et al., 1998].
(We use ‘‘transpression/transtension’’ due to its widespread
usage and acceptance in the geologic literature, though we
note that a term describing deformation rather than stress
is more appropriate.) Along transform plate boundaries,
releasing bends or steps can create large pull-apart basins
such as the Salton Sea in the San Andreas Fault (SAF) and
the Dead Sea and Hula basins along the Dead Sea Fault
(DSF), whereas a restraining geometry can result in uplifted
blocks such as the Transverse Ranges (Big bend of the SAF)
and Lebanon and Anti-Lebanon mountains (Lebanese
restraining bend of the DSF). In addition to their remarkable
effect on landscape geomorphology, these regions have a
profound impact on structural style, seismicity and the
distribution of hydrocarbon, mineral and water resources.
[3] For continental transforms, transpression (more aptly
named ‘‘convergent strike slip’’) is often manifested by a
zone up to 100 km wide of distributed deformation characterized by folds and thrust faults [e.g., Christie-Blick and
Biddle, 1985; Sylvester, 1988; Norris et al., 1990; Woodcock
and Schubert, 1994]. The strike-slip deformation within
these zones deviates from simple shear because of a component of shortening orthogonal to the deformation zone
[Mount and Suppe, 1987; Teyssier et al., 1995; Dewey et
al., 1998]. Strains may be kinematically nonpartitioned
or partitioned in transpression and transtension. In nonpartitioned transpression, the strike slip and shortening components of strain are coupled, resulting in en echelon folds
with axes striking 20– 40° to the main transform [Aydin
and Page, 1984; Sylvester, 1988], strike-slip duplexes
[Woodcock and Fischer, 1986], and maximum shortening
oriented 35 – 45° to the main transform [Tikoff and
Teyssier, 1994]. In contrast, for partitioned transpression
the strike-slip and shortening components are decoupled into
discrete zones of deformation. For complete partitioning, the
strike-slip component is accommodated by the bounding,
TC5005
1 of 19
TC5005
WEINBERGER ET AL.: DEFORMATION ALONG THE DEAD SEA FAULT
TC5005
great detail [e.g., Davis et al., 1989; Butler et al., 1998;
Tavarnelli, 1998; Woodcock and Rickards, 2003], which in
turn may provide insight into the tectonic development of
transform plate boundaries. Regions where transpression is
partitioned into strike-slip and contractional domains are of
particular interest due to the potential for seismicity originating from blind thrust faults adjacent to the transform
[Stein and King, 1984; Davis et al., 1989; Shaw and Suppe,
1996; Dolan et al., 2003]. Thus understanding how faults,
folds and other structures develop in these complex regions
contributes to the assessment of seismic risk along plate
boundary faults.
[5] The main objective of our study is to describe the
geometries, kinematics and temporal sequence of deformation adjacent to the DSF in northern Israel through geological
mapping and mesostructural analysis. The study area is
located at the critical juncture between the Hula Valley
(basin) in northern Israel and the Lebanese restraining bend
(Figure 1), thus providing the opportunity to investigate
effects of contrasting fault geometry and relative plate motion
on deformation within a strike-slip setting. Whereas the
relative motion across the DSF is well constrained for the
relatively ‘‘straight’’ sections south of the Sea of Galilee [e.g.,
Garfunkel, 1981] and for the Lebanese restraining bend
[Gomez et al., 2007b], the nature of the plate boundary in
the transition zone between these regions is less clear, with
strike-slip motion apparently distributed across a broad
network of faults (Figure 1). Further, the variety of structures
exposed both along discrete fault zones and within intervening blocks may provide important geological constraints for
models of transpression and strain partitioning proposed for
continental transforms.
2. Tectonic and Geologic Setting
Figure 1. Map of the main fault segments of the Dead Sea
Fault (DSF) in northern Israel and southern Lebanon. The
study area near the city of Qiryat Shemona is marked by a
dashed rectangle. The inset shows the plate tectonic
configuration resulting in left-lateral motion across the
DSF. Hula western border fault (HWBF) and Hula eastern
border fault (HEBF) mark the edges of the Hula basin. LRB,
Lebanese restraining bend; DS, Dead Sea.
transform-parallel faults and shortening by transform-normal
deformation distributed within the intervening blocks [Mount
and Suppe, 1987; Teyssier et al., 1995; Dewey et al., 1998].
[4] Structural analysis provides the opportunity to characterize the internal deformation of transpressive zones in
[6] The 1000 km long Dead Sea Fault (Transform)
accommodates strike-slip motion between the Arabian plate
and the Sinai plate (or African plate) (Figure 1, inset). It
connects seafloor spreading in the Red Sea with continental
collision at the Eurasian plate. The transform is commonly
divided into four sectors: (1) the 300 km long, NNE-SSW
trending southern section, linking the Red Sea rift to the
Dead Sea basin; (2) the 220 km long, N-S trending section
between the Dead Sea and the Hula Valley; (3) the 140 km
long section in the Lebanese restraining bend defined by the
NE-SW trending Yammunneh fault; and (4) the 340 km
long, N-S trending northern sector from the Ghab basin in
Syria into southern Turkey. A host of stratigraphic, structural and geochronological evidence suggests 105 km of
left-lateral offset across the DSF since the middle Miocene
[e.g., Quennell, 1959; Freund et al., 1970; Garfunkel, 1981;
Joffe and Garfunkel, 1987; Sneh and Weinberger, 2003a].
Estimated slip rates across the DSF vary widely, but generally
fall within the range of 3 –10 mm/a. Estimates based on
tectonic modeling and long-term geological considerations
generally range from 6 to 10 mm/a [e.g., Joffe and Garfunkel,
1987], whereas geomorphologic estimates yield 3 – 7 mm/a
[Ginat et al., 1998; Niemi et al., 2001; Meghraoui et al.,
2003; Daeron et al., 2004]. GPS measurements reveal a
2 of 19
TC5005
WEINBERGER ET AL.: DEFORMATION ALONG THE DEAD SEA FAULT
current slip rate of 3.5 mm/a for the southern and central
sections of the DSF [e.g., Wdowinski et al., 2004].
[7] In the southern and central sections of the DSF
between the Red Sea and Hula Valley, the main transform
is shifted to the west through a series of overlapping, leftstepping, en echelon segments. This geometry resulted in
the initiation of two prominent pull-apart basins, the Dead
Sea and the Hula Valley to its north (Figure 1) [Garfunkel,
1981]. The right-stepping bend in the Yammunneh fault
corresponds to a restraining geometry that uplifted the
Lebanon and Anti-Lebanon mountains [Beydoun, 1977;
Daeron et al., 2004]. The transition of the DSF from the
Hula basin to the Lebanese restraining bend is complex,
with deformation distributed across several faults (i.e.,
Serghaya, Rachaya, and Hasbaya faults to the east of the
Yammunneh fault and the Roum and Margaliyyot faults to
its west; Figure 1). This area consists of deformed structural
blocks [Picard, 1952; Glikson, 1966; Weinberger and Sneh,
2004]. The Hula basin is bounded to the east and west by
two N-S striking faults (Figure 1); the Hula eastern border
fault (HEBF) and the Hula western border fault (HWBF).
Farther to the north, the Margaliyyot and Qiryat Shemona
faults branch off the Hula western border fault. The Qiryat
Shemona fault branches into the Roum and Yammunneh
faults in southern Lebanon (Figure 2) and is considered the
main strand of the DSF in this area [Sneh and Weinberger,
2003a]. It subdivides the area into two main blocks; a
western block, locally known as the Misgav Am block,
between the Qiryat Shemona and Margaliyyot faults; and an
eastern block, locally known as the Metulla block, between
the Qiryat Shemona and Tel-Hay faults (Figure 2). These
two deformed blocks are the main focus of this study. A less
deformed block, locally known as the Naftali Mountains
block (Figures 1 and 2), extends west of the Hula western
border fault and south of the Margaliyyot fault (Figure 2).
The block consists of Cretaceous beds that dip gently
westward beneath the thick Tertiary formations that comprise the open Yir’on-Nabatiya syncline (Figure 1). Another
block extends to the east of the Tel-Hay fault but is
extensively covered by Pleistocene basalt flows and travertine units that do not allow access to the deformed blocks
underneath.
3. Results of Geologic Mapping
TC5005
eastern limb and vertical beds with tight parasitic folding in
the core of the syncline (see details below).
3.2. Eastern Block (Metulla Block)
[9] The eastern block is 2 km wide and, like the
western block, it is divided into several moderate to highly
deformed subblocks by approximately NE-SW to N-S
striking faults, including the Har Zefiyya, Metulla, Muftalah,
and Kefar Giladi faults (Figure 2). Strata exposed in this
block consist of the Cretaceous Kamon and Deir Hanna
formations, the Paleogene Taqiye, Timrat, and Bar Kokhba
formations, the Neogene Kefar Giladi Formation, the
Pleistocene Hazbani Basalt flows and Quaternary gravels
and clay (Figure 2). The conglomerates and lacustrine sediments of the Neogene Kefar Giladi Formation are especially
noteworthy because they accumulated within an extensional
basin adjacent to the Qiryat Shemona master fault and were
subsequently uplifted and folded. In the south, on both sides
of the NE striking Muftalah fault, the Kefar Giladi beds are
vertical and intensely contorted in places. The Metulla fault
to the north is a N-S striking thrust fault dipping 30° to the
east, with the Eocene Timrat Formation overriding lacustrine
beds of the Neogene Kefar Giladi Formation on the western
side of the fault (site L, Figure 2 and Figure 4a). The NE
striking Har Zefiyya fault dips 70° southeastward (site K,
Figure 2), showing normal separation with the Eocene Bar
Kokhba limestone in the hanging wall and the Cenomanian
Karkara beds of the Deir Hanna Formation exposed in the
footwall. However, kinematic indicators (dip-slip slickenside lineations and small-scale pull-apart basins filled with
calcite) suggest a younger reverse motion along this fault.
[10] The subvertical north striking Tel-Hay bounding
fault (site J, Figure 2) dips steeply to the east and normally
displaces the Pleistocene flows of the Hazbani Basalt in the
hanging wall against the Paleocene– lower Eocene Taqiye
Formation in the footwall. However, the uppermost flow of
the Hazbani Basalt at this locality covers the fault plane and
shows no evidence of displacement. Because ages of the
Hazbani Basalt are between 1.5 and 0.8 Ma [Heimann,
1990; Sneh and Weinberger, 2003b; Y. Harlavan, personal
communication, 2003], this observation suggests that at
least in the last 0.8 Ma this segment of the north striking
Tel-Hay fault was inactive.
3.3. Shehumit Hill
3.1. Western Block (Misgav Am block)
[8] The width of the western block reaches a maximum
of about 3 km and decreases northward to less than 1 km.
Farther north in southern Lebanon, its width remains narrow
all the way to the village of Roum (Figure 1). A set of faults
that strike approximately NE-SW divide the western block
into subblocks, each displaying highly deformed beds
which are subvertical in places [Dubertret, 1951]. Strata
exposed in the western block belong to the carbonate rocks
of the Cretaceous Kurnub and Judea Groups, including the
Albian Rama and Kamon formations and the Cenomanian
Deir Hanna Formation (Figure 2). The rocks are folded into
a broad syncline (Figure 3), with moderate dips along the
[11] Shehumit Hill is a north trending ridge about 2200 m
long and 230 m wide, uplifted 85 m above the Hula Valley
floor. It consists of Pleistocene flows of the Hazbani Basalt
(site E, Figure 2). The Tel-Hay fault runs at the foot of the
ridge to the west. At the eastern flank of the ridge, the
Hazbani Basalt is tilted 20– 40° eastward to southeastward
(i.e., columnar joints are dipping 70°– 50° westward). On
the top of the ridge, the Hazbani Basalt is horizontal (i.e.,
columnar joints are vertical), whereas on the western flank it
is covered by large basaltic boulders and urban development. Previous workers considered the ridge as a large-scale
fissure eruption [Picard, 1952], a tilted block [Schulman,
1966] or a push-up swell [Heimann and Ron, 1987]. A
2.8 km high-resolution seismic line that runs perpendicular
3 of 19
TC5005
WEINBERGER ET AL.: DEFORMATION ALONG THE DEAD SEA FAULT
Figure 2
4 of 19
TC5005
WEINBERGER ET AL.: DEFORMATION ALONG THE DEAD SEA FAULT
TC5005
TC5005
Figure 3. E-W cross section (line X-X0 on Figure 2) based on geologic mapping. Note the subdivision
into eastern and western blocks and disharmonic folds in the lacustrine sediments of the Kefar Giladi
Formation. See rock units in Figure 2.
to the ridge to the south along road 9779 shows folding of
the basalts underneath the ridge [Frieslander and Medvedev,
2002; Weinberger and Sneh, 2004], in agreement with the
observed tilting of the exposed Hazbani Basalt. Moreover,
west of Shehumit Hill, the basalt in the subsurface is
deformed into a positive flower structure near the Qiryat
Shemona fault [Frieslander and Medvedev, 2002].
4. Results of Mesostructural Analyses
4.1. Castle Site
[12] This site is located less than 50 m west of the
Margaliyyot fault and consists of Albian-Cenomanian dolostone beds exposed in the moat of the 12th century Chateau
Neff Crusader castle (site A, Figure 2). The site is characterized by a large-scale anticline cut by three prominent
thrust faults that form an imbricate fan structure within a
massive dolostone unit (Figures 5a and 5b, unit 7). Thin
beds of dolostone and marlstone in the underlying unit are
intensely folded and faulted within the core of the anticline
(Figures 5c and 5d, unit 8). We measured orientations of
bedding in units 4 – 8, fold hinges, axial planes and small
back thrusts in unit 8, as well as the larger thrust faults
(F1– F3) cutting these units.
[13] The trend and plunge of the axis of the large-scale
anticline is 190°/06° (lines are reported as trend/plunge)
based on a cylindrical best fit of poles to bedding (i.e., p
diagram) measured along the length of the castle moat
(Figure 6a). The small folds in unit 8 are tight, with typical
wavelengths of several cm to tens of cm (Figure 5d). Of the
19 fold hinges measured, 18 plunge slightly or moderately to
the south and yield a mean of 181°/15° (Figure 6b).
Orientations of axial planes are somewhat scattered, with a
mean of 82°/101° (planes are reported as dip/dip azimuth)
(Figure 6b). The measured fold geometries (hinges, axial
planes, large-scale fold axis) are internally consistent with
overlapping confidence intervals, indicating a roughly subhorizontal axis trending approximately N-S.
[14] Thrust faults at the castle fall into three main
categories: large prominent faults that mark the boundaries
of massive dolostone fault blocks, small faults confined to
individual units, and faults that displace a series of boudins
within an extended bed in unit 8. All three large faults dip
moderately to the west (55° – 65°), thus implying tectonic
transport from west to east. Small thrust faults group into a
Figure 2. Geological map of the study area showing rock units, faults, locations of sites referred to in the text (black
circles) and the location of cross section X-X0 (Figure 3). Inset: Division into structural blocks and the bounding faults.
Rock units are Klhn, Hatira, and Nabi Sa’id formations (sandstone, marl, Neocomian-Barremian,); Klei, Ein el Assad
Formation (limestone, Aptian); Klir, Hidra and Rama formations (limestone, marl, chalk, Aptian-Albian); Klkam, Kamon
Formation (dolostone, Albian-Cenomanian); Kudk, Deir Hanna Formation, Karkara Member (dolostone, Cenomanian);
Kudr, Deir Hanna Formation, Rosh Haniqra Member (chalk, limestone, marl, Cenomanian); Kub, Bina Formation
(limestone, Turonian); Kj, Judea Group (limestone and dolostone, undivided, Aptian-Turonian); Tlt, Taqiye Formation
(marl, Paleocene); Kums, Mount Scopus Group (chalk, undivided, Senonian-Paleocene); Et, Timrat Formation (chalk,
limestone, Eocene); Ebk, Bar Kokhba Formation (limestone, Eocene); Nkg, Kefar Gil’adi Formation (conglomerate,
limestone, clay, Neogene); Peg, Egel Gravel (Plio-Pleistocene); Pbm, Meshki Basalt (Pliocene); Qboh, Hazbani Basalt
(Pleistocene); Qt, Travertine (Pleistocene); Al, alluvium, terrace conglomerates and landslides (Pleistocene-Holocene).
5 of 19
TC5005
WEINBERGER ET AL.: DEFORMATION ALONG THE DEAD SEA FAULT
Figure 4. Photographs of structures observed in the field area. (a) Low-angle Metulla thrust fault at site L
in the eastern block with Eocene Timrat Formation in the hanging wall and Neogene Kefar Giladi
Formation in the footwall. The arrow tips indicate the fault trace. (b) South plunging folds in the Cretaceous
Deir Hanna Formation in the core of the syncline at site C in the western block. (c) West verging chevron
folds in the Deir Hanna Formation at site B in the western block. (d) Syncline in Neogene conglomerates of
the Kefar Giladi Formation adjacent to the Muftalah fault in the eastern block (site G, Figure 2).
(e, f ) Polished vertical faults in the Eocene Bar Kokhba Formation exposed in the quarry (eastern block).
The faults strike N-S with subhorizontal undulations and slickenside lineations. (g) Polished hand sample
taken from a calcite-filled vein within the Bar Kokhba Formation in the quarry (eastern block). Note bands
of calcite parallel to the vein wall (left) and brecciated host rock (right).
6 of 19
TC5005
TC5005
WEINBERGER ET AL.: DEFORMATION ALONG THE DEAD SEA FAULT
TC5005
Figure 5. (a) Photomosaic of folded and thrusted Cenomanian dolostone beds of the Kamon Formation
exposed in the north facing wall of the Chateau Neff Crusader castle (site A, Figure 2). (b) Structural
interpretation of the photomosaic showing stratigraphic units, overall anticlinal geometry and main thrust
faults (F1 –F3). (c) Close-up photo of unit 8 in the core of the anticline showing intense folding of thin
beds. Note the small back thrusts associated with boudins below Fault F3. (d) Small folds in unit 8 (refer
to the rectangle in Figure 5c for exact location of photo).
west dipping synthetic set parallel to the large faults, and
an antithetic set dipping to the east (back thrusts). The
mean orientation for all west dipping faults is 62°/270°,
whereas east dipping thrust faults yield a mean of 56°/107°
(Figure 6c). Bedding cutoff lines (i.e., the intersections of
bedding exposed on a fault surface) were measured where
possible, yielding a mean of 186°/01° (Figure 6d). Although
slickenside lineations are not well preserved on many of the
fault planes, we measured definitive rakes of 70° – 90° on
three thrust fault surfaces, implying a dominantly dip-slip
sense of motion. Structural elements related to the thrust
faults (fault orientations, cutoff lines, slickenside lineations)
combined with results from the fold analysis attest to E-W
finite shortening at the castle locality (Figure 6d).
4.2. Western Block (Misgav Am Block)
[15] Bedding orientations, fold hinges, and axial planes
were measured mainly along the main road that crosses the
block (Figure 2). In addition, we studied two prominent en
echelon folds at the southeastern part of the block (site D,
Figure 2). The stratigraphy and bedding orientations indicate that the block is a large-scale, 1.5 km wide syncline
(Figure 3). The trend and plunge of the syncline axis is
182°/09° based on the p diagram derived from ‘‘regional’’
bedding (i.e., beds unaffected by localized small-scale
folds) measured along the roadcut (Figure 7a). The core
of the syncline (site C, Figure 2) is characterized by subvertical bedding, chevron folds (1 m wavelength) and
steeply plunging folds (2– 5 m wavelength; Figure 4b) at
7 of 19
TC5005
WEINBERGER ET AL.: DEFORMATION ALONG THE DEAD SEA FAULT
TC5005
Figure 6. Structural data collected at Chateau Neff Crusader castle. Small circles represent 95%
confidence intervals; open symbols are means. (a) A p diagram of poles to bedding (n = 26) measured
throughout the castle exposure. The calculated fold axis is 191°/06°. (b) Fold hinges (solid triangles;
mean 181°/15°, n = 18) and poles to axial planes (solid circles, n = 13) measured in unit 8. Great circle
represents the mean axial plane (82°/101°). (c) Poles to east dipping thrust faults (n = 10) and west
dipping thrust faults (n = 7). (d) Summary of main structural elements including the mean of cutoff lines
(open hexagon).
the transition from the stiff limestone and dolostone of the
Karkara Member to the soft chalky rocks of the Rosh Haniqra
Member of the Deir Hanna Formation (Figure 8). Two groups
of folds are distinguished from each other (Figure 7b):
chevron folds in units 2– 5 are characterized by subhorizontal
axial planes and a mean hinge oriented 028°/14°, whereas
folds in units 7 –8 are tight and plunge to the south (mean
hinge orientation of 174°/44°).
[16] Hinges of high-quality chevron folds exposed on the
western limb of the syncline (site B, Figure 2 and Figure 4c)
show considerable scatter in orientation, with a mean of
023°/10° (Figure 7c). Contouring suggests the possibility of
a bimodal distribution of the fold hinges, with a dominant
cluster centered around 030°/10° and a less prominent cluster
at 004°/00° (Figure 7c). Most of the axial planes dip to the
east, with a mean orientation of 49°/103° (Figure 7c),
suggesting west directed tectonic transport and E-W finite
shortening.
[17] Two prominent en echelon anticlines with wavelengths of several hundred meters are exposed at the
southeastern part of the block (site D, Figure 7d). The
exposed cores of the anticlines are composed of the Albian
Rama Formation, whereas the flanks expose the AlbianCenomanian Kamon Formation. On the basis of the geological map [Sneh and Weinberger, 2003b], the fold axes
trend SSW-NNE and terminate against the Qiryat Shemona
8 of 19
TC5005
WEINBERGER ET AL.: DEFORMATION ALONG THE DEAD SEA FAULT
TC5005
Figure 7. Structural data collected in the western block. Small circles represent 95% confidence
intervals; open symbols are means. (a) A p diagram of poles to bedding (n = 22) measured along road
9977 (Figure 2). The calculated fold axis is 182°/09°. (b) Hinges of chevron folds in units 2– 5 (mean
hinge orientation of 028°/14°, n = 5) and tight folds in units 7 – 8 (mean hinge orientation of 174°/44°, n = 8)
from site C in the core of the syncline. See Figure 8 for site C cross section. (c) Contoured fold hinges
(triangles, n = 22) and poles to axial planes (solid circles, n = 12) of chevron folds (see Figure 4c) at site B
(Figure 2). Contour intervals are 2% per 1% area. (d) A p diagram of poles to bedding (n = 14) measured
along a prominent en echelon fold that terminates against the Qiryat Shemona fault (site D, Figure 2).
fault to the east. A fold axis of 207°/41° was calculated for
the southern anticline based on a p diagram derived from
14 bedding measurements (Figure 7d).
4.3. Eastern Block (Metulla Block) General
[18] Bedding orientations and fold hinges were measured
along road 9977 (Figure 2) and in the area to its south. In
addition, data from faults, veins, and clastic dikes were
collected in exceptional exposures of Eocene rocks within
the Kefar Giladi Quarry (site H, Figure 2).
4.3.1. Folds
[19] A p diagram of all bedding orientations measured in
the eastern block yields a calculated fold axis of 193°/11°
(Figure 9a). Ten medium- to large-scale folds, ranging from
several meters up to several hundreds of meters in wavelength, were analyzed separately based on bedding measurements collected across each structure. A large-scale
(>300 m) open syncline in lacustrine sediments adjacent
to the intersection between the Kefar Giladi and Muftalah
faults (site I) yields a calculated fold axis of 210°/18°
(Figures 4d and 8b). A large-scale asymmetric, open anticline in the Kefar Giladi conglomerate adjacent to the NNE
trending Muftalah fault (site G) yields a fold axis of 008°/
01° (Figure 8c). The third large-scale fold exposed in the
quarry is a broad anticline with a calculated fold axis of
169°/16°. Contouring the axes of these 10 prominent folds
9 of 19
TC5005
WEINBERGER ET AL.: DEFORMATION ALONG THE DEAD SEA FAULT
Figure 8. Cross section at the core of the syncline (site B, Figure 2; see location in Figure 3) in the
western block showing numbered mechanical units, all of which are carbonates, and their associated
structures. Horizontal and vertical scales are equal. Measurements of fold hinges in units 2 –5 and 7 –8
are presented in Figure 7b. For a photograph of units 7 – 8 see Figure 4b.
Figure 9. Bedding and fold data collected in the eastern block. Small circles represent 95% confidence
intervals; open symbols are means. (a) A p diagram of poles to bedding (n = 216) measured throughout
the block. (b) A p diagram of poles to bedding (n = 19) for a large-scale open syncline (site I) near the
Kefar Giladi fault. The calculated fold axis is 210°/18°. (c) Fold hinges (n = 10) from site F (mean 016°/17°)
and p diagram of poles to bedding (n = 15) for a large-scale asymmetric anticline near the Muftalah fault
(site G, Figure 4d; fold axis = 008°/18°). (d) Contoured fold axes of ten prominent folds in the eastern
block. Contour intervals are 5% per 1% area.
10 of 19
TC5005
TC5005
WEINBERGER ET AL.: DEFORMATION ALONG THE DEAD SEA FAULT
suggests the possibility of a bimodal distribution, with a
dominant cluster centered around 177°/07° and a less
prominent cluster at 210°/15° (Figure 9d). In addition, 10
hinges of small-scale folds (several centimeters in wavelength) were measured within vertical beds of the Kefar
Giladi lacustrine sediments (site F, Figure 2) and their mean
is 016°/17°.
4.3.2. Quarry Faults
[20] Numerous subsidiary faults within the Eocene Bar
Kokhba limestone are exposed in the Kefar Giladi Quarry
(site H, Figure 2), forming a wide zone of brittle deformation to the east of the Qiryat Shemona master fault. Many of
the fault surfaces are polished with associated breccia and
gouge zones (Figures 4e and 4f). Most of the large faults are
vertical to subvertical, tens of meters in height, and often
form vertical faces on the quarry walls (Figure 4e). Lineations on the fault surface appear as either undulating fault
grooves or fault striae (slickenside lineations) (Figure 4f ).
Occasionally, rhomb-shaped pull-aparts, slickolites, and
wear grooves indicate the sense of motion along the faults.
Data were collected from 119 faults where accurate orientations could be measured, and where the fault striae and
grooves allowed for the measurement of lineations. Eleven
faults display two sets of fault striae. In most cases, the two
sets are shallow plunging striae (i.e., dominantly strike slip);
however, in other cases, one set of striae is shallow plunging
and the other is more steeply plunging. Where timing
relations can be deduced in the field, the steeply plunging
dip-slip striae postdate the shallow plunging strike-slip
striae. Sense of motion was noted where possible, and the
faults were classified according to their size. Out of 119
measured faults, 18 had definitive criteria to assign the sense
of motion with confidence, 76 faults provided a lower-quality
assessment of the sense of motion, and the sense of motion
was ambiguous for the other faults.
[21] The most abundant set of vertical to subvertical faults
range in strike between NNW and NNE (Figure 10a). A less
abundant set of subvertical faults strikes approximately E-W.
There are also some shallowly dipping faults and faults that
do not fall within the two main trends. The overwhelming
majority (80%) of fault striae are subhorizontal (<30°),
indicating a dominant strike-slip motion, mostly on the
approximately N-S striking faults (Figure 9b).
[22] We performed a kinematic analysis of the fault data
using the FaultKin program, which calculates a P and T axis
for each fault based on fault orientation, striae orientation
and the sense of motion (R. W. Allmendinger et al.,
FaultKinWin: A program for analyzing fault slip data for
WindowsTM computers, 2001, available at http://www.geo.
cornell.edu/geology/faculty/RWA/programs.html). The
mean P and T axes provide a first-order approximation
of the infinitesimal shortening (S_ 3) and extension (S_ 1)
directions, respectively, for a population of faults. We first
analyzed the 18 faults with unambiguous senses of motion,
of which 13 have a component of left-lateral motion and 5
are right lateral. The results yield a definitive fault plane
solution with subvertical theoretical fault planes striking N-S
and E-W (Figure 10a), individual P and T axes confined to
separate quadrants, and infinitesimal strain axes character-
TC5005
ized by subhorizontal NW-SE shortening (mean P axis of
308°/08°) and subhorizontal NE-SW extension (mean T axis
of 218°/08°). These infinitesimal strain axes resolve into leftlateral strike-slip motion along the N-S striking fault plane
and right-lateral strike-slip motion along the E-W striking
fault plane.
[23] We then analyzed a combined population of faults
consisting of the 18 faults with definitive senses of motion
and the 76 faults with lower-quality kinematic indicators.
Results show a scatter of P and T axes, but clear concentrations of P axes in NW and SE quadrants near the perimeter
of the stereonet (Figure 11b), consistent with results from
the 18 high-quality faults. The average P axis is 131°/06°
and the average T axis is 221°/02°, conforming to NW-SE
infinitesimal shortening and NE-SW infinitesimal extension.
As with the previous analysis, the results provide a near
perfect strike-slip solution, with two vertical fault planes, one
striking N-S with left-lateral motion, and the other E-W with
right-lateral motion. Similar results are obtained when data
are analyzed separately according to fault size.
[24] FaultKin assigns the fault data into four categories
based on the dip-slip and strike-slip components of motion
(TL, thrust and left components; NL, normal and left
components; TR, thrust and right components; NR, normal
and right components). Most of the data (TL = 41; NL = 35)
are associated with a left-lateral sense of motion. When
analyzed together they yield results similar to the previous
analyses, i.e., a strike-slip solution, N-S and E-W fault
planes, and a P axis trending NW-SE. The abundance of
faults with a left-lateral sense of motion may hide an
infinitesimal strain associated with faults characterized by
either right-lateral or dip-slip motion. Therefore, we analyzed
separately the 18 faults (TR = 11; NR = 7) with right-lateral
sense of motion, as well as the 14 faults with striae plunging
more than 45° (dominantly dip-slip motion). The average
P axis of the faults with right-lateral motion is ESE
(Figure 11c). However, contouring the P axes reveals two
maxima; one approximately E-W and the other NW-SE. The
average P axis derived from faults with striae plunging
more than 45° is ESE, and results in a thrust fault plane
solution (Figure 11d).
4.3.3. Quarry Veins
[25] Two types of opening-mode fractures, calcite-filled
veins and clastic dikes, are exposed in the quarry. The veins
are 2 to 60 cm thick and several meters up to >20 m high.
There is no evidence of shear along vein walls. They are
characterized by vertical bands of calcite, aligned parallel to
vein walls (Figure 4g). Many veins consist of fibrous calcite
crystals aligned perpendicular to the vein walls, with a tabular
geometry and no evidence for internal shear displacement.
Amorphous and botryoidal calcite may also be present,
especially between banded zones. Angular fragments of the
Eocene host rock surrounded by calcite cement are occasionally found within the veins, or between the vein material and
the wall rock, implying a dilational origin (Figure 4g). The
clastic dikes are filled with clay or soil consisting of gravel,
silt and pebbles.
[26] Orientations of veins (n = 38) and clastic dikes (n = 5)
overlap, and hence are analyzed together. The mean plane
11 of 19
TC5005
WEINBERGER ET AL.: DEFORMATION ALONG THE DEAD SEA FAULT
TC5005
Figure 10. Fault and vein data collected from the Eocene Bar Kokhba Formation in the quarry (site H,
eastern block). Contour intervals are 2% per 1% area. (a) Contoured poles to fault planes (n = 119)
showing that most faults are vertical and strike between NNE and NNW directions. (b) Contoured fault
striae (n = 119). Note that most fault striae are subhorizontal. (c) Poles to opening-mode veins (n = 38)
and clastic dikes (n = 5). The mean plane is 87°/014°.
of the combined population of veins and clastic dikes is
87°/014° (Figure 10c). Contouring the data suggests that
two sets exist; a dominant E-W striking set (mean of 86°/
008°) and a second possible set that strikes NW-SE
(Figure 10c). The dominant set of veins and clastic dikes
thus reflect approximately N-S extension.
[27] Where crosscutting relations are observed, the veins
consistently postdate the fault planes. In some cases the veins
appear to be ‘‘utilizing’’ a preexisting fault (i.e., plane of
weakness), which may be responsible for some scatter in
their orientations and the occasional NW-SE striking veins.
Three laminae from banded vein material were dated by
the U-Th method in the Geological Survey of Israel laboratories. The lamina closest to the host rock is in equilibrium
and is older than 400,000 years B.P. The age of the middle
lamina is 206,500 years B.P. ± 14,000 (2s), and the lamina
closest to the center of the vein has an age of 191,000 years
B.P. ± 14,000 (2s) [Weinberger et al., 2008].
5. Discussion
5.1. Synthesis of Structural Data: Evidence for Two
Directions of Regional Shortening
[28] Our field observations and measurements attest to
the dominantly contractional nature of deformation in the
Qiryat Shemona region as previously noted by Ron et al.
[1997] and Weinberger and Sneh [2004]. The structural
elements measured at a variety of scales in the study area
group into two distinct categories based on their inferred
directions for regional shortening. The first category represents approximately NW-SE finite shortening and consists
of (1) two large en echelon anticlines exposing Cretaceous
12 of 19
TC5005
WEINBERGER ET AL.: DEFORMATION ALONG THE DEAD SEA FAULT
TC5005
Figure 11. Results of kinematic analysis of the fault data collected at the quarry (site H) in the eastern
block. (a) S_ 3 (P) and S_ 1 (T) axes and fault plane solution for the 18 faults with unambiguous sense of
motion (highest quality). (b) S_ 3 and S_ 1 axes and fault plane solution for 94 faults (18 highest-quality
faults and 76 faults of lower quality). (c) Contoured S_ 3 axes and fault plane solution for 18 of the 94 faults
with a component of right-lateral motion. Contour intervals are 2% per 1% area. (d) S_ 3 and S_ 1 axes and
fault plane solution for 14 of the 94 faults with striae plunging more than 45°.
rocks in the western block and the large-scale Kefar Giladi
syncline in the eastern block whose axes are oriented NNESSW (Figure 2) and (2) a population of chevron folds in the
core and western limb of the broad syncline with hinges
trending NNE-SSW within the western block. The second
category represents approximately E-W finite shortening
and includes (1) approximately N-S trending axes of largescale folds such as the castle anticline in the western block
(Figure 5), the broad syncline of the western block
(Figure 3), the Kidmat Tel Hay syncline, quarry anticline
and other prominent folds in the eastern block (Figure 9)
and the Shehumit anticline east of the Tel Hay fault;
(2) small folds with approximately N-S hinges in the castle
and within the core and western limb of the syncline of the
western block (Figures 5 and 6); and (3) approximately N-S
striking thrust faults at the castle and the approximately N-S
striking segment of the Metulla thrust fault (Figure 2); In
addition, the approximately E-W striking opening-mode
fractures, including calcite-filled veins and clastic dikes in
the quarry and Har Zefiyya (Figure 2, site K, eastern block),
indicate approximately N-S extension, which is compatible
with approximately E-W finite shortening.
[29] The main structural elements and kinematic results
for the entire study area are combined in summary stereoplots in order to facilitate the interpretation of regional
deformation (Figure 12). A summary plot of 19 fold axes
derived from (1) averages of fold hinge populations at
individual localities and (2) p diagrams of large folds is
presented in Figure 12a. Contours of the fold axes reveal
two subhorizontal maxima, one trending approximately N-S
13 of 19
TC5005
WEINBERGER ET AL.: DEFORMATION ALONG THE DEAD SEA FAULT
TC5005
Figure 12. Summary stereoplots of major structural elements and kinematics for the entire study area.
(a) Fold axes (solid circles, n = 19) with contour intervals of 2% per 1% area. Axes are derived from
averages of fold hinge populations at specific localities, as well as p diagrams of large folds. Note two
maxima along the perimeter (shallow plunging folds) indicating two fold populations based on trend of
axes. (b) Interpretation of structural elements, including mean and 95% confidence intervals of the two
main fold populations (open triangles), P (S_ 3) and T (S_ 1) axes from the kinematic analysis of quarry faults
(Figure 11b), mean fault plane of large thrusts at the castle, and mean orientation of the dominant vein set
from the quarry. Large solid arrows represent finite shortening and extension directions derived from the
various structural elements. Dashed black lines are mean trends of fold axes. Gray dashed line is trend of
fold axes expected during pure strike-slip deformation.
and the other approximately NNE-SSW, as well as a third
cluster of more steeply dipping axes plunging to the south.
The steep fold plunges likely developed as a consequence of
fold tightening, and thus are interpreted to belong to fold
populations with similar trends. On the basis of the contours, the fold axes were divided into two separate populations with statistically distinct means at the 95% confidence
intervals, one with a mean of 176°/16° and the other 205°/
05° (Figure 12b). The two directions of regional shortening
reflect two main phases of deformation in the Qiryat
Shemona area associated with the development of the
DSF; an early phase of NW-SE regional shortening applied
to the boundary of the transform followed by a later phase
of E-W regional shortening associated with N-S regional
extension (Figure 12b). The early phase was associated with
NW-SE infinitesimal shortening as indicated by the mean P
and T axes derived from the kinematic analysis of undifferentiated faults in the eastern block (Figures 11a and 11b);
the later phase was associated with approximately E-W
infinitesimal shortening as indicated by the mean P and T
axes derived from the kinematic analysis of quarry faults
with a component of right-lateral motion (Figure 11c).
[30] The E-W regional shortening postdates the NW-SE
regional shortening for the following reasons. First, the
youngest rocks exposed in the area, the 1.5 – 0.8 Ma
Hazbani Basalt, are affected only by E-W shortening as
manifested by the N-S trending Shehumit anticline. Second,
the E-W striking calcite-filled veins are opening-mode
fractures that reflect N-S extension, parallel to the DSF.
Where crosscutting relations are observed, the veins consistently postdate the N-S striking faults and utilize them as
preexisting planes of weakness. Further, these veins are
dated as middle to late Pleistocene in accordance with a
later phase of E-W shortening. Third, the steeply plunging
fault striae associated with dip-slip motion during E-W
infinitesimal shortening postdate the low plunging striae,
which are associated with strike-slip motion as a response to
NW-SE infinitesimal shortening. Fourth, the most recent
movement across the steeply dipping NE trending Har
Zefiyya fault is reverse slip despite its normal stratigraphic
separation, which is compatible with E-W regional short-
14 of 19
WEINBERGER ET AL.: DEFORMATION ALONG THE DEAD SEA FAULT
TC5005
TC5005
Table 1. Location of Euler Poles and Angular Velocities of Arabian Plate Relative to Sinai Plate Determined in Previous Studiesa
Pole
1
2
3
4
5
6
7
Pole Location
32.8°N,
32.76°N,
28.78°N,
28.71°N,
32.8°N,
14.2°N,
31.1°N,
22.6°E
20.69°E
28.61°E
30.00°E
28.4°E
21.9°E
26.7°E
Angular
Velocity (w)
(deg/Ma)
Speed
(mm/a)
Azimuth
(CW From North)
350° Velocity
(mm/a)
080° Velocity
(mm/a)
Data
Sourceb
0.283
0.1774
0.2158
0.2114 ± 0.1584
0.370 ± 0.027
0.230 ± 0.044
0.434 ± 0.012
5.94
4.26
3.07
2.65 ± 1.98
4.32 ± 0.31
9.83 ± 1.87
6.52 ± 0.18
2.10°
2.63°
326.09°
319.05°
359.35°
306.59°
347.67°
5.81
4.16
2.81
2.27 ± 1.70
4.25 ± 0.30
7.14 ± 1.36
6.51 ± 0.17
1.25
0.93
1.24
1.36 ± 1.02
0.70 ± 0.05
6.75 ± 1.29
0.26 ± 0.01
Joffe and Garfunkel [1987, Table 4]
Wdowinski et al. [2004, Table 8]c
Wdowinski et al. [2004, Table 8]c
Wdowinski et al. [2004, Table 8]c
Reilinger et al. [2006, Table 1]d
Gomez et al. [2007a, p. 1026
Seyrek et al. [2007, p. 255]
Predicted motion of each pole is given for Qiryat Shemona (33°150N; 35°340E).
The Euler pole of Joffe and Garfunkel [1987] is based on an instantaneous kinematic model for the last 5 Ma. The Euler pole of Seyrek et al. [2007] is
based on geological observations from Turkey, Lebanon and Israel. The Euler poles of Wdowinski et al. [2004], Reilinger et al. [2006] and Gomez et al.
[2007a] are based on current GPS measurements.
c
Wdowinski et al. [2004] calculated three Euler poles based on different assumptions/models.
d
Reilinger et al. [2006] calculated the relative motion between Arabian and Sinai plates based on kinematic block modeling and showed convergence
along the DSF from the Dead Sea northward.
a
b
ening. Nevertheless, the NE trending faults (e.g., Har
Zefiyya, Muftalah; Figure 2) are the least exposed structures
in the study area and hence are less studied and understood.
Because they terminate against the north trending bounding
faults and displace Neogene rocks, they likely formed
during strike-slip motion along the DSF. On the basis of
their azimuths and angular relations with respect to the
major north trending faults (e.g., Qiryat Shemona, Tel-Hay)
they may represent P shear strands.
[31] Our structural analysis provides evidence for an
early phase (pre-Pleistocene) of almost pure strike-slip
motion across the DSF (i.e., NW-SE finite shortening) that
formed the NNE-SSW trending en echelon folds (the 205°
trend in Figure 12b), large N-S striking vertical faults and
probably the NE-SW striking faults that may be considered
as P shear strands. The early fold axes may have rotated up
to 16° counterclockwise due to progressive simple shear
along the transform, as indicated by the angle between the
current trend of fold axes (205°) and the expected trend
(221°) based on the infinitesimal strain axes associated with
pure strike-slip deformation (Figure 12b). The late phase of
deformation (post-lower Pliocene), characterized by a significant component of E-W regional shortening normal to
the DSF (i.e., oblique convergent strike slip), formed the
N-S trending folds, N-S striking thrust faults, and reactivated select NE striking faults in reverse motion. In
addition, this phase is associated with N-S extension that
formed the youngest features, i.e., E-W striking veins and
clastic dikes (Figure 12b). The scatter in orientations and
intermediate shortening trends between the two NW-SE
and E-W maxima may (e.g., fold hinges from site F,
Figure 9c) indicate a gradual transition between the two
prominent phases of deformation.
5.2. Tectonic Explanation for Evolving Contraction
Across the DSF in Northern Israel
[32] The Qiryat Shemona fault strikes 350° in northern
Israel before it splays into the NNW striking Roum fault
and NE striking Yammunneh fault in southern Lebanon
(Figure 1). The strike of the Qiryat Shemona fault suggests
that during the first phase of deformation (i.e., N-S directed
strike-slip motion), the Qiryat Shemona fault served as a
releasing bend. Consequently, an associated depression
developed along its margin, providing space for the accumulation of the 400 m thick Kefar Giladi conglomerate
and lacustrine sediments during the upper Miocene – lower
Pliocene. During the second (and current) phase of deformation, these sediments have been uplifted and folded,
indicating a transition from almost pure strike-slip to
convergent strike-slip motion along this strand of the
DSF. On the basis of the assigned age of the Kefar Giladi
Formation, this transition (i.e., tectonic inversion) took
place post-lower Pliocene, which might be attributed to
changes in the relative motion between the Sinai and
Arabian plates.
[33] We calculated the relative motion between the Sinai
and Arabian plates using published Euler poles and rates of
rotation in order (1) to determine the local velocity vector
along the Qiryat Shemona plate boundary fault and (2) to
decompose the velocity vector into fault-parallel (strike-slip;
350° azimuth) and fault-normal (shortening or extension;
080° azimuth) components. The calculations are presented
in Table 1 for a fixed Sinai plate. The plate kinematic model
of Joffe and Garfunkel [1987] for the last 5 Ma predicts
northward motion of Arabia relative to Sinai at Qiryat
Shemona (Table 1, Euler pole 1), suggesting strike-slip
motion with minor divergence along the Qiryat Shemona
fault. A similar result is obtained using Euler poles 2 and 5
(Table 1). It is noteworthy that Reilinger et al. [2006]
calculated the relative motion between the Sinai and Arabian plates based on Euler pole 5 and a kinematic block
model that includes the effect of elastic strain accumulation
on block boundaries. They found that the strand of the DSF
between the Dead Sea and southern Turkey is currently
undergoing convergence. The other four Euler poles predict
northwestward motion of Arabia relative to Sinai (Table 1,
Euler poles 3, 4, 6, and 7), resulting in a westward
component of velocity normal to the plate boundary. This
prediction implies convergent strike-slip motion along the
Qiryat Shemona fault. We also calculated the uncertainties
15 of 19
TC5005
WEINBERGER ET AL.: DEFORMATION ALONG THE DEAD SEA FAULT
TC5005
Figure 13. Schematic of the two phases proposed for deformation across the DSF in northern Israel.
(a) Early phase of pure strike-slip motion. (b) Late phase of convergent strike-slip motion (transpression)
characterized by strain partitioning along the DSF. Gray features were formed in the early phase. S_ 1 and
S_ 3 axes refer to infinitesimal strain, whereas the ellipses represent finite strain.
associated with the velocity vectors and their components to
verify whether the convergent strike-slip motion along the
Qiryat Shemona fault is statistically significant. The uncertainties arise due to uncertainties in the pole position and
angular velocity; the former are minor whereas the latter are
generally significant and are calculated here (Table 1; Euler
poles 4 – 6).
[34] The results indicate that the convergent-type motion
is statistically significant. Because even a minor component
of convergence (a < 5°, where a is the angle between the
strike of the plate boundary and the plate motion velocity
vector) resulting from changes in relative plate motion can
lead to a high degree of strain partitioning [Teyssier et al.,
1995], the inferred present-day convergence of a > 10°
across the Qiryat Shemona segment of the DSF provides a
viable kinematic explanation for the strain partitioning
documented in this study.
5.3. Strain Partitioning
[35] We characterize the second phase of deformation as
strain partitioning resulting from convergent strike slip, with
strike-slip (simple shear) motion along discrete N-S fault
zones and coaxial shortening roughly normal to these zones
in the intervening blocks. The degree of strain partitioning
is nearly complete [e.g., Teyssier et al., 1995, Figure 3]
considering that a > 10° and the 70– 90° angle between the
plate boundary and the directions of infinitesimal shortening
(as manifested by fault kinematics) and finite shortening (as
manifested by young folds, thrust faults and veins), thus
resembling deformation adjacent to the SAF in portions of
central California [e.g., Mount and Suppe, 1987, 1992;
Zoback et al., 1987; Tavarnelli, 1998; Bawden et al.,
2001; Fuis et al., 2003]. Other examples of strain partitioning along the DSF have been observed in the convergent
Lebanese restraining bend to the north [Griffiths et al.,
2000; Gomez et al., 2007b] and in the divergent Dead Sea
basin and Gulf of Elat – Aqaba to the south [Ben-Avraham
and Zoback, 1992; Sagy et al., 2003].
[36] What makes this case study in northern Israel
unique, and potentially valuable for understanding transform margin deformation globally, is the evidence for the
progression through time from almost pure strike-slip (nonpartitioned) to partitioned transpressive deformation. We
envision an early phase of almost pure strike slip distributed
across the transform (Figure 13a). At this point the plate
boundary consisted of a limited number of poorly
connected fault segments and the upper crust was strong.
With time, slip accumulated along the plate boundary and
a change in relative plate motion occurred. The former led
16 of 19
TC5005
WEINBERGER ET AL.: DEFORMATION ALONG THE DEAD SEA FAULT
TC5005
Figure 14. Proposed model for the current phase of strain partitioning across the DSF in northern Israel.
Strike-slip component of plate motion is localized to discrete N-S fault zones whereas the intervening
blocks are subjected to E-W shortening, resulting in the development of a fold-thrust belt along the
transform margins. Potential seismicity (stars and the associated focal plane solutions) may arise from slip
along the weak N-S strike-slip faults, reactivation of faults within the blocks, or actively growing blind
thrust faults. Fault segments (vertical lines drawn next to the edges of the blocks) form fault zones
adjacent to the N-S bounding faults. Gray lines represent locked faults.
to the development of weak, throughgoing, and wellconnected fault zones in the upper crust, whereas the latter
led to transpression along these zones of the transform
(Figure 13b). As a consequence, the late (and current)
phase of deformation as depicted in Figure 14 is partitioned into strike-slip motion along the weak N-S striking
major vertical fault zones and the development of a
‘‘minifold-thrust belt’’ due to E-W shortening adjacent to
the transform.
5.4. Increasing Plio-Pleistocene Convergence Along
Other Parts of the DSF
[37] Evidence for a Pleistocene tectonic transition is
found in a series of locations along the DSF (see summary
by Schattner and Weinberger [2008]). The most striking
example is the tectonic transition in the Hula basin, which
formed as a pull-apart basin 4 Ma, but entered a new
geodynamic phase during the Pleistocene. At that time a
left-lateral, throughgoing, diagonal strike-slip fault developed within the basin due to increasing northward convergence across the central and northern sectors of the DSF
[Schattner and Weinberger, 2008]. Gomez et al. [2007b]
showed that the Lebanese restraining bend also experienced
a two-stage tectonic evolution. The early phase involved
wrenching and rotation, and the later (current) phase is
characterized by strain partitioning. The transition between
these phases was attributed to changes in plate motion, with
northward relative motion in the early phase, and increasing
convergence of the plates during the later phase. At the
north end of the DSF in the Karasu Valley, Seyrek et al.
[2007] showed crustal thickening and folding due to transpression which they attribute to changes in relative plate
motion during the Pliocene.
5.5. Implications for Seismicity
[38] The change in direction of regional shortening documented by structural analysis will have an effect on the
population of faults that are seismically active. The early
phase of NW-SE regional shortening is optimally oriented
for slip along vertical, N-S striking fault planes such as the
abundant mesoscale faults observed in the quarry as well as
the larger Qiryat Shemona and Tel Hay faults (Figure 2).
17 of 19
TC5005
WEINBERGER ET AL.: DEFORMATION ALONG THE DEAD SEA FAULT
However, when subjected to E-W regional shortening
accompanied by N-S extension, many of these faults will
become locked as a consequence of the large component of
normal stress acting across the fault surfaces, reducing their
likelihood for frictional sliding. Evidence for reduced strikeslip activity along N-S vertical faults is provided by the
undeformed 1.5 – 0.8 Ma basalt flow overlying the Tel-Hay
fault and a late phase of dip-slip motion that overprints early
strike-slip movement along N-S striking faults in the quarry.
[39] Among preexisting faults, those striking NE-SW are
aligned for reactivation during the phase of E-W shortening.
This explains the late component of reverse motion across
the NE striking Har Zefiyya fault (Figure 2) and the
kinematic results for quarry faults with a component of
right-lateral motion (Figure 11c). N-S striking thrust faults
that formed during the late phase of transform-normal
shortening (e.g., the castle faults, Metulla fault) are strong
candidates for slip (Figure 11d). Further, although focal
plane solutions of thrust faults have not been reported in this
region [Salamon et al., 2003], the actively developing foldthrust belt adjacent to the DSF in northern Israel raises the
possibility of earthquakes originating from blind thrust
faults, analogous to the 1983 Coalinga, 1987 Whittier
Narrows, and 1994 Northridge earthquakes along the San
Andreas Fault in California [Stein and King, 1984; Davis et
al., 1989; Shaw and Suppe, 1996; Dolan et al., 2003].
Finally, major N-S faults (e.g., Qiryat Shemona master
fault) may remain active during this later phase if they are
weak (i.e., frictionless [e.g., Zoback et al., 1987]), resulting
in the partitioning of seismic activity characterized by thrust
motion within contracting blocks and pure left-lateral strikeslip motion along the weak zones bounding the blocks [e.g.,
Stein and Hanks, 1998; Fuis et al., 2003] (Figure 14).
6. Summary
[40] We documented the style and sequence of deformation associated with an actively convergent section of the
plate boundary DSF in northern Israel. The study area is
divided into distinct structural blocks by a series of distributed faults that comprise this section of the DSF. Two of
TC5005
these blocks, which are adjacent to the approximately north
striking Qiryat Shemona master fault, are broadly folded,
with more intense localized contraction manifested by tight
folds and thrust faults. Kinematic analyses of these folds
and faults at a variety of scales as well as E-W striking veins
provide evidence for two directions of regional shortening:
NW-SE shortening responsible for the formation of NE-SW
trending fold axes and left-lateral strike-slip motion along
N-S trending faults, and E-W shortening as indicated by NS trending fold axes, N-S striking thrust faults and extensional veins of mid-late Pleistocene age that strike E-W. The
present analysis provides evidence for the transition from an
early (pre-Pleistocene) phase of almost pure strike-slip
motion to a late (post-lower Pliocene) phase of ‘‘partitioned’’ transpression, with the latter characterized by discrete left-lateral strike-slip motion across weak N-S faults
and the development of a fold-thrust belt in response to
transform-normal shortening. Calculations of the relative
motion between the Sinai and Arabia plates using published
GPS-driven Euler poles and rates of rotation show significant plate convergence in northern Israel and provide a
viable kinematic explanation for the strain partitioning
along this section of the DSF. The present observations
and analysis raise the possibility that blind thrust faults
adjacent to the DSF in the Qiryat Shemona area pose a
seismic risk to populations in northern Israel and southern
Lebanon. Major N-S faults (e.g., Qiryat Shemona master
fault) may remain active if they are weak, resulting in the
partitioning of seismic activity characterized by thrust
motion within contracting blocks and strike-slip motion
along the block-bounding faults.
[41] Acknowledgments. This study was supported by grant 2004232
from the United States – Israel Binational Science Foundation (BSF). We
are grateful to Shimon Wdowinski for supportive discussions and many
helpful suggestions during the course of this study. We are also grateful to
the Associate Editor, Keith Klepeis, Atilla Aydin, and Dyanna Czeck for
thorough and helpful reviews. We thank Mira Bar-Matthews for U-Th age
determinations of calcite-filled veins and Yehudit Harlavan for K-Ar age
determinations of basalts from the circum of Hula Valley. Discussions with
Uri Frieslander, Uri Schattner, and Tsafrir Levi are greatly appreciated. The
structural data were plotted using Rick Allmendinger’s programs Stereonet
and FaultKin.
References
Aydin, A., and A. Nur (1982), Evolution of pull-apart
basins and their scale independence, Tectonics, 1,
1303 – 1317, doi:10.1029/TC001i001p00091.
Aydin, A., and B. M. Page (1984), Diverse PlioceneQuaternary tectonics in a transform environment,
San Francisco Bay region, California, Geol. Soc.
Am. Bull., 95, 91 – 105, doi:10.1130/00167606(1984)95<1303:DPTIAT>2.0.CO;2.
Bawden, G. W., W. Thatcher, R. S. Stein, C. Wicks,
K. Hudnut, and G. Peltzer (2001), Tectonic contraction
across Los Angeles after removal of groundwater
p u m p i n g e ff e c t s , N a t u re , 4 1 2 , 8 1 2 – 8 1 5 ,
doi:10.1038/35090558.
Ben-Avraham, Z., and M. D. Zoback (1992), Transform-normal extension and asymmetric basins: An
alternative to pull-apart models, Geology, 20, 423 –
426, doi:10.1130/0091-7613(1992)020< 0423:
TNEAAB>2.3.CO;2.
Beydoun, Z. R. (1977), Petroleum prospects of Lebanon, Am. Assoc. Petrol. Geol. Bull., 61, 43 – 64.
Butler, R. W. H., H. M. Spencer, and H. M. Griffiths
(1998), The structural response to evolving plate
kinematics during transpression: Evolution of the
Lebanese restraining bend of the Dead Sea Transform, in Continental Transpressional and Transtensional Tectonics, edited by R. E. Holdsworth et al.,
Geol. Soc. Spec. Publ., 135, 81 – 106, doi:10.1144/
GSL.SP.1998.135.01.06.
Christie-Blick, N., and K. T. Biddle (1985), Deformation and basin formation along strike-slip fault, in
Strike-Slip Deformation, Basin Formation, and
Sedimentation, edited by K. T. Biddle and N. ChristieBlick, Spec. Publ. Soc. Econ. Paleontol. Mineral., 37,
1 – 34.
Daeron, M., L. Benedetti, P. Tapponnier, A. Sursock,
and R. C. Finkel (2004), Constraints on the post
18 of 19
25-ka slip rate of the Yammouneh fault (Lebanon)
using in situ cosmogenic 36Cl dating of offset limestone-clast fans, Earth Planet. Sci. Lett., 227, 105 –
119, doi:10.1016/j.epsl.2004.07.014.
Davis, T. L., J. Namson, and R. F. Yerkes (1989), A
cross section of the Los Angeles area: Seismically
active fold and thrust belt, the 1987 Whittier Narrows earthquake, and earthquake hazard, J. Geophys. Res., 94, 9644 – 9664, doi:10.1029/
JB094iB07p09644.
Dewey, J. F., R. E. Holdsworth, and R. A. Strachan
(1998), Tranpression and transtension zones, in
Continental Transpressional and Transtensional
Tectonics, edited by R. E. Holdsworth et al., Geol.
Soc. Spec. Publ., 135, 1 – 14, doi:10.1144/GSL.SP.
1998.135.01.01.
Dolan, J. F., S. A. Christofferson, and J. H. Shaw
(2003), Recognition of paleoearthquakes on the
TC5005
WEINBERGER ET AL.: DEFORMATION ALONG THE DEAD SEA FAULT
Puente Hills blind thrust fault, California, Science,
300, 115 – 118, doi:10.1126/science.1080593.
Dubertret, L. (1951), Carte geologique au 1:50000
feuille de Mardjayoun, report Libanaise, notice explicative (1952), 64 pp., Min. des Travaux, Beyrouth, France.
Freund, R., Z. Garfunkel, I. Zak, M. Goldberg, T. Weissbrod,
and B. Derin (1970), The shear along the Dead Sea
Rift, Philos. Trans. R. Soc. London, Ser. A, 267,
107 – 130, doi:10.1098/rsta.1970.0027.
Frieslander, U., and B. Medvedev (2002), High resolution seismic reflection survey at the Qiryat Shemona region (in Hebrew), Geophys. Inst. Rep. 398/
246/02, 9 pp., Isr. Geophys. Inst., Lod.
Fuis, G. S., et al. (2003), Fault systems of the 1971 San
Fernando and 1994 Northridge earthquakes, southern California: Relocated aftershocks and seismic
images from LARSE II, Geology, 31, 171 – 174,
d o i : 1 0 . 11 3 0 / 0 0 9 1 - 7 6 1 3 ( 2 0 0 3 ) 0 3 1 < 0 1 7 1 :
FSOTSF>2.0.CO;2.
Garfunkel, Z. (1981), Internal structure of the Dead Sea
leaky transform (rift) in relation to plate kinematics,
Tectonophysics, 80, 81 – 108, doi:10.1016/00401951(81)90143-8.
Ginat, H., Y. Enzel, and Y. Avni (1998), Translocated
Plio-Pleistocene drainage systems along the Arava
fault of the Dead Sea Transform, Tectonophysics, 284,
151 – 160, doi:10.1016/S0040-1951(97)00165-0.
Glikson, Y. A. (1966), The lacustrine Neogene in the
Kefar Gil’adi area, northern Jordan Valley, Isr.
J. Earth Sci., 15, 85 – 100.
Gomez, F., G. Karam, M. Khawlie, S. McClusky,
P. Vernant, R. Reilinger, R. Jaafar, C. Tabet,
K. Khair, and M. Barazangi (2007a), Global
Positioning System measurements of strain
accumulation and slip transfer through the restraining bend along the Dead Sea Fault system in Lebanon, Geophys. J. Int., 168, 1021 – 1028, doi:10.1111/
j.1365-246X.2006.03328.x.
Gomez, F., T. Nemer, C. Tabet, M. Khawlie, M. Meghraoui,
and M. Barazangi (2007b), Strain partitioning of
active transpression within the Lebanese restraining
bend of the Dead Sea Fault (Lebanon and SW Syria),
in Tectonics of Strike-Slip Restraining and Releasing
Bends, edited by W. D. Cunningham and P. Mann,
Geol. Soc. Spec. Publ., 290, 285 – 303, doi:10.1144/
290.10.
Griffiths, H. M., R. A. Clark, K. M. Thorp, and S. Spencer
(2000), Strain accommodation at the lateral margin
of an active transpressive zone: Geological and seismological evidence from the Lebanese restraining
bend, J. Geol. Soc., 157, 289 – 302.
Heimann, A. (1990), The development of the Dead
Sea Rift and its margins in northern Israel during
the Pliocene and the Pleistocene (in Hebrew with
English summary), Geol. Surv. Rep. GSI/28/90,
83 pp., Isr. Geol. Surv., Jerusalem.
Heimann, A., and H. Ron (1987), Young faults in the
Hula pull apart basin, central Dead Sea Transform,
Tectonophysics, 141, 117 – 141, doi:10.1016/00401951(87)90179-X.
Joffe, S., and Z. Garfunkel (1987), Plate kinematics of
the circum Red Sea—A re-evaluation, Tectonophysics, 141, 5 – 22, doi:10.1016/0040-1951(87)90171-5.
Meghraoui, M., et al. (2003), Evidence for 830 years of
seismic quiescence from palaeoseismology, archaeoseismology and historical seismicity along the
Dead Sea Fault in Syria, Earth Planet. Sci. Lett.,
210, 35 – 52, doi:10.1016/S0012-821X(03)00144-4.
Mount, V. S., and J. Suppe (1987), State of stress near
the San Andreas Fault: Implications for wrench tectonics, Geology, 15, 1143 – 1146, doi:10.1130/
0091-7613(1987)15<1143:SOSNTS>2.0.CO;2.
Mount, V. S., and J. Suppe (1992), Present-day stress
orientations adjacent to active strike-slip faults: California and Sumatra, J. Geophys. Res., 97(B8),
11,995 – 12,013, doi:10.1029/92JB00130.
Niemi, T. M., H. Zhang, A. Mohammad, and H. J.
Bruce (2001), Late Pleistocene and Holocene slip
rate of the Northern Wadi Araba fault, Dead Sea
Transform, Jordan, J. Seismol., 5, 449 – 474,
doi:10.1023/A:1011487912054.
Norris, R., P. O. Koons, and A. F. Cooper (1990), The
obliquely convergent plate boundary in the South
Island of New Zealand, J. Struct. Geol., 12, 715 –
725, doi:10.1016/0191-8141(90)90084-C.
Picard, L. (1952), The geology of Kefar-Giladi (in Hebrew), Bull. Israel Explor. Soc., ‘‘Lif’’ memorial
vol. B, 73 – 77.
Quennell, A. M. (1959), Tectonics of the Dead Sea Rift,
paper presented at XX International Geological
Congress, Asoc. de Serv. Geol. Afr., Mexico City.
Reilinger, R., et al. (2006), GPS constraints on continental deformation in the Africa-Arabia-Eurasia
continental collision zone and implications for the
dynamics of plate interactions, J. Geophys. Res.,
111, B05411, doi:10.1029/2005JB004051.
Ron, H., G. Shamir, and Y. Eyal (1997), Deformation of
Margaliot block between Roum and Margaliot
faults, paper presented at Israel Geological Society
Annual Meeting, Kefar Gil’adi, Israel.
Sagy, A., Z. Reches, and A. Agnon (2003), Hierarchic
three-dimensional structure and slip partitioning in
the western Dead Sea pull-apart, Tectonics, 22(1),
1004, doi:10.1029/2001TC001323.
Salamon, A., A. Hofstetter, Z. Garfunkel, and H. Ron
(2003), Seismotectonics of the Sinai subplate—The
eastern Mediterranean region, Geophys. J. Int., 155,
149 – 173, doi:10.1046/j.1365-246X.2003.02017.x.
Schattner, U., and R. Weinberger (2008), A midPleistocene deformation transition in the Hula basin, northern Israel: Implications for the tectonic
evolution of the Dead Sea Fault, Geochem. Geop hy s . G e os ys t . , 9 , Q 07 00 9, d oi: 10 .1 02 9/
2007GC001937.
Schulman, N. (1966), The Qiryat Shemona (northern
Jordan Valley) basalt ridge: A tilted fault block,
Isr. J. Earth Sci., 15, 161 – 164.
Seyrek, A., T. Demir, M. S. Pringle, S. Yurtmen, R. W.
C. Westaway, A. Beck, and G. Rowbotham (2007),
Kinematics of the Amanos Fault, southern Turkey,
from Ar/Ar dating of offset Pleistocene basalt
flows: Transpression between the African and Arabian plates, in Tectonics of Strike-Slip Restraining
and Releasing Bends, edited by W. D. Cunningham
and P. Mann, Geol. Soc. Spec. Publ., 290, 255 –
284, doi:10.1144/SP290.9.
Shaw, J. H., and J. Suppe (1996), Earthquake hazards of
active blind-thrust faults under the central Los Angeles basin, California, J. Geophys. Res., 101,
8623 – 8642, doi:10.1029/95JB03453.
Sneh, A., and R. Weinberger (2003a), Geology of the
Metulla quadrangle, northern Israel: Implications
for the offset along the Dead Sea Rift, Isr. J. Earth
19 of 19
TC5005
Sci., 52, 123 – 138, doi:10.1560/1G3J-NX0HKBL3-RUY9.
Sneh, A., and R. Weinberger (2003b), Geological map
of Israel, sheet 2-II Metulla, scale 1:50,000, Isr.
Geol. Surv., Jerusalem.
Stein, R. S., and T. C. Hanks (1998), M>=6 earthquakes
in southern California during the twentieth century:
No evidence for a seismicity or moment deficit,
Bull. Seismol. Soc. Am., 88, 635 – 652.
Stein, R. S., and G. C. P. King (1984), Seismic potential
revealed by surface folding: 1983 Coalinga, California earthquake, Science, 224, 869 – 872,
doi:10.1126/science.224.4651.869.
Sylvester, A. G. (1988), Strike-slip faults, Geol. Soc.
Am. Bull., 100, 1666 – 1703, doi:10.1130/00167606(1988)100<1666:SSF>2.3.CO;2.
Tavarnelli, E. (1998), Tectonic evolution of the Northern Salinian Block, California, USA: Paleogene to
Recent shortening in a transform fault-bounded
continental fragment, in Continental Transpressional and Transtensional Tectonics, edited by
R. E. Holdsworth et al., Geol. Soc. Spec. Publ.,
135, 107 – 118, doi:10.1144/GSL.SP.1998.135.01.07.
Teyssier, C., B. Tikoff, and M. Markley (1995), Oblique
plate motion and continental tectonics, Geology, 23,
4 4 7 – 4 5 0 , d o i : 1 0 . 11 3 0 / 0 0 9 1 - 7 6 1 3 ( 1 9 9 5 )
023<0447:OPMACT>2.3.CO;2.
Tikoff, B., and C. B. Teyssier (1994), Strain modeling
of displacement-field partitioning in transpressional
orogens, J. Struct. Geol., 16, 1575 – 1588,
doi:10.1016/0191-8141(94)90034-5.
Wdowinski, S., Y. Bock, G. Baer, L. Prawirodirdjo,
N. Bechor, S. Naaman, R. Knafo, Y. Forrai, and
Y. Melzer (2004), GPS measurements of current
crustal movements along the Dead Sea Fault, J.
Geophys. Res., 109, B05403, doi:10.1029/
2003JB002640.
Weinberger, R., and A. Sneh (2004), The geology of
Qiryat Shemona region, paper presented at Israel
Geological Society Annual Meeting, Hagoshrim,
Israel.
Weinberger, R., M. Bar-Matthews, and M. R. Gross
(2008), Preliminary U-Th ages of calcite-filled
veins from the Kefar Giladi Quarry, Rep. TR-GSI/
04/08, 8 pp., Isr. Geol. Surv., Jerusalem.
Woodcock, N. H., and M. Fischer (1986), Strike-slip
duplexes, J. Struct. Geol., 8, 725 – 735,
doi:10.1016/0191-8141(86)90021-0.
Woodcock, N. H., and B. Rickards (2003), Transpressive duplex and flower structure: Dent Fault System, NW England, J. Struct. Geol., 25, 1981 – 1992,
doi:10.1016/S0191-8141(03)00057-9.
Woodcock, N. H., and C. Schubert (1994), Continental
strike-slip tectonics, in Continental Tectonics, edited by P. L. Hancock, pp. 251 – 263, Pergamon,
Oxford, U. K.
Zoback, M. D., et al. (1987), New evidence on the state
of stress of the San Andreas Fault system, Science,
238, 1105 – 1111, doi:10.1126/science.238.4830.
1105.
M.
R.
Gross, Department of Earth Sciences, Florida
International University, University Park, P.O. Box 344,
11200 SW 8th Street, Miami, FL 33199, USA.
A. Sneh and R. Weinberger, Geological Survey of
Israel, 30 Malkhe Israel Street, Jerusalem 95501, Israel.
([email protected])