Accelerated subsidence and sedimentation in the Levant Basin during the Late Tertiary and concurrent uplift of the Arabian platform: Tectonic versus counteracting sedimentary loading effects


Corresponding author: Oded Bar. Present address: Geological Survey of Israel, 30 Malkhe Israel Street, Jerusalem, 95501, Israel. (


[1] Since the Middle Eocene, the northwest Arabian Platform has been emerging from the water and rising above sea level, whereas the adjacent Levant Basin has been subsiding and accumulating a thick sedimentary section. This study investigates these opposing vertical motions and the relative roles of tectonic-driven versus isostatic adjustment forces involved. Such a distinction is strongly dependent on a reliable estimation of the paleobathymetry, which in pelagic environments may vary substantially. We use two different methods to estimate the paleo-water-depth in the deep Levant Basin in the Tertiary time. Isostatic balancing calculations with the adjacent inland region are employed to estimate the paleo-water-depth in the deep Levant Basin in the Middle Eocene, just before the commencement of the Late Tertiary tectonic phase. For later periods, we use morpho-structural elements such as abrasion surfaces and incised canyons to build laterally changing topo-bathymetry profiles. Our results indicate that in the Mid-Eocene water depth in the deep Levant Basin was 2–3.5 km, which gradually decreased to 1.5 km today. Based on these results, our analysis shows that the enhanced subsidence of the Levant Basin reflects an isostatic response to sedimentary filling of a pre-existing deep-water basin with no involvement of a downward tectonic force. On the contrary, we suggest that the regional tectonic force was upward, counteracting a sedimentary loading effect. Further regional implication of this understanding is that the cause for uplifting and exposure of the NW Arabian Platform in the Late Tertiary extended far westward beyond the inland region.

1 Introduction

1.1 Late Tertiary Reactivation of the Levant Passive Continental Margin

[2] Following the Middle Eocene, at ca. 37 Ma, the paleogeography of the Middle East region (Figure 1B) underwent significant changes. Huge areas in northwest Arabia, previously covered by seawater, emerged above sea level, were eroded, and shorelines retreated hundreds of kilometers towards the present-day Mediterranean coasts [e.g., Ziegler, 2001]. Contemporaneously, the Levant continental margin accumulated a thick succession of fine clastic sediments, reaching a thickness of more than 2 km along the Israeli continental shelf [e.g., Gvirtzman, 1970]. Further west in the deeper Levant Basin, more than 5 km of fine clastic sediments accumulated [Gvirtzman et al., 2008; Steinberg et al., 2011] (Figure 2). This study discusses the vertical displacement involved in the young concurrent opposite movements, inland uplift, and basin subsidence, analyzing the relationships between tectonic forces, sedimentation, and erosion.

Figure 1.

A. Map of the study area combining the present topography in the east with a subsurface structural map of the base Saqiye Group in the west. We used this structural map, which describes the nearly Base Oligocene surface, as a proxy for the Early Oligocene relief showing morpho-structural steps and incised canyons (discussed in text). The structural map is from Steinberg et al. [2011] with minor modifications in its eastern part from Fleischer and Gafsou [2003]. Outlines of the El-Arish Canyon are modified after Gvirtzman et al. [2008]. Solid dots - location of wells studied (Figures 2, 47, 9). Black lines - faults of the Continental Margin Fault Zone, from Gvirtzman et al. [2008]. Backstripped wells - marked in red; other wells, in orange. Numbers refer to wells' names as follows: 1= Til 1; 2= Kefar-Darom 1; 3= Nezarim 1; 4= Pleshet 1; 5= Yam-West 1; 6= Yam 2; 7= Yam 1; 8= Joshua 1; 9= Echo 1; 10= Hof-Ashdod 1; 11= Ashdod 1; 12= Bene-Darom 1; 13= Gan-Yavne 5; 14= Mivtah 1; 15= Gedera 1; 16= Hulda 1; 17= Motza 1; 18= Yam-Yafo 1; 19= Jaffa 1; 20= Petah-Tiqva 2; 21= David 1; 22= Ramallah 1. B. Map of the regional tectonic setting of the Middle East and Eastern Mediterranean. Gray rectangle outlines the study area. Straight black lines outline the course of the geological cross sections of Figure 2.

Figure 2.

A. Schematic regional geological cross section from Jordan to the Eratosthenes seamount [modified after Garfunkel, 1998] demonstrating the extended crustal structure of the Levant continental margin. B. Geological cross section modified after Gardosh et al. [2010], emphasizing late Paleozoic and early Mesozoic extensional structures later deformed by Late Cretaceous to Early Tertiary shortening. C. Geological section from the Judea Mountains plateau to the southeast Levant Basin (this study) emphasizing mainly post Middle Eocene rocks above the thick brown line, which represent the Early Oligocene structure (later tilted and buried). The stepped structure of the continental margin includes four blocks (Mt. backbone, foothills and east coastal plain, continental shelf, and deep Levant Basin) separated by three morpho-structural steps (west mountains front, coastal plain step, and continental margin fault zone). Information west of the Yam-West-1 well is based on seismic data from Gvirtzman et al. [2008] and Steinberg et al. [2011]. D. Interpreted seismic reflection profile across the western part of the continental margin and the eastern part of the deep Levant Basin [modified after Gvirtzman and Steinberg, 2012]. This section illustrates the disturbed Continental Margin Fault Zone between the relatively undeformed Levant Basin and Continental Margin blocks. Location of all sections in Figure 1.

[3] Among the crucial parameters for understanding geodynamic mechanisms controlling the vertical motions that shaped the southeastern Mediterranean and its inland margin are the relative roles of tectonically driven motion versus motion due to isostatic response to sediment loading and unloading. To distinguish between the two mechanisms, we used the backstripping method [Watts and Ryan, 1976; Steckler and Watts, 1978], in which the sedimentary column is gradually unroofed backward in time and the crustal motion due to isostatic response and tectonic force is calculated at each stage.

[4] Reconstruction of the vertical displacement and distinction of the tectonic component from the total basin subsidence requires knowledge of three factors at each point in time: sediment load (thickness and density of the sedimentary column), paleo-water-depth, and paleo-sea-level. Sediment thickness and density at each stage were deduced from well data corrected for compaction at later burial stages. Paleo-sea-level was taken from global eustatic curves such as Haq et al. [1987] and Miller et al. [2005]. A first estimation of paleo-water-depth is commonly deduced from sedimentological and paleontological data, particularly in shallow marine environments, where the possible range of water depths is relatively small. However, in pelagic environments, where water depth may vary from hundreds to thousands of meters, there is a large uncertainty, and different environmental estimations may lead to completely different tectonic interpretations.

[5] Foraminifera assemblages and benthic to planktonic ratios from three oil wells located along the continental shelf [e.g., Buchbinder et al., 2000; Almogi-Labin et al., 2001] indicate that the region, which is only a few tens of meters deep today, was 500–1000 m deep in the Early Pliocene. However, similar studies for the deeper section of the continental shelf are not available and paleo-water-depth indicators for the deeper Levant Basin are absent for the entire sedimentary section. To overcome this absence, we developed a nonroutine geologic approach for paleo-water-depth estimation.

[6] Our approach for reconstructing the paleo-water-depth is based on two independent methods. For the Middle Eocene period, just before the renewal of tectonism, we assumed thermal equilibrium. Accordingly, we assumed that isostatic compensation was achieved at a depth equivalent to the base of the inland crust (~35 km) and that deeper lateral variations can be ignored. This assumption allows us to calculate the water depth in the deep Levant Basin by balancing its crustal column, including the unknown thickness of the water column, with the crustal column of the inland region where the paleo-water-depth is known.

[7] For later periods, we reconstructed a paleobathymetry and paleotopography profile extending from the deep Levant Basin to the Judea Mountains plateau. This reconstruction is based on geological data, such as height differences across morpho-structural steps and relicts of ancient abrasive and erosive surfaces, which were originally close to sea level.

[8] Using this approach for reconstructing paleobathymetry and paleotopography, we can distinguish between the effects of sediment loading and deep-seated tectonic processes that controlled vertical movements of the NW Arabian Platform and the Levant continental margin during the time in which Arabia rifted from Africa and collided with Eurasia.

1.2 Geological Setting

[9] The southern Levant continental margin was formed through a series of rifting phases that extended over a few tens of millions of years from the Permian to the Middle Jurassic. The northward shift of the rifted continental blocks resulted in the formation of the Levant Basin, which is a part of the eastern Mediterranean Sea, a relict of the ancient Neo-Tethys ocean [e.g., Robertson and Dixon, 1984; Garfunkel, 1998, 2004].

[10] The faulted basement across the continental margin with its tilted-block structure is schematically shown in Figure 2A [from Garfunkel, 1998]; in this plot, the faults do not extend above Jurassic strata. A later study by Gardosh et al. [2010] shows mostly reverse faults that reach Cretaceous and Early Tertiary levels (Figure 2B). This description follows Freund et al. [1975] and Druckman et al. [1995b] who postulated that the Syrian Arc deformation (see below) reactivated normal Early Mesozoic faults. Thus, the current structure does not show a simple horst-and-graben structure (Figure 2B) as was schematically envisioned by Garfunkel [1998].

[11] In any case, data from central and southern Israel show that after the Early Mesozoic rifting passive margin conditions were established and the continental margin was accompanied by prolonged thermal relaxation [Garfunkel and Derin, 1984]. Along with the post rift thermal subsidence, a series of transgressions and regressions affected the rate and nature of deposition. Sea invasion and inland sedimentation reached a climax in the Senonian-Middle Eocene period when pelagic sedimentation reached as far as Sinai, Jordan, and parts of Saudi Arabia [Beydoun, 1991; Ziegler, 2001].

[12] Contemporaneously, the beginning of the closure of the Neo-Tethys in the Late Cretaceous [e.g., Robertson and Dixon, 1984; Garfunkel, 1998, 2004; Robertson et al., 2006] resulted in a horizontal compression and formation of the Syrian Arc folds belt along the NW margin of the Arabian Platform [e.g., Krenkel, 1924; Walley, 1998]. Syn-folding deposition associated with typical thickness variations continued about 50 m.y., during which most of the fold crests were submerged under deep waters most of the time [Buchbinder et al., 1988].

[13] In the Late Eocene (ca. 37 Ma), these pelagic conditions and the slow rates of deposition were interrupted by inland emergence, enhanced basin sedimentation, and strong structural deformation. A short pulse of shortening formed three morpho-structural steps that reshaped the previously moderate, gradually westward descending seafloor. The modified structure of the ancient seafloor is illustrated on the map of Figure 1A, which, as a result of combining a near base Oligocene structural map in the west with the present topography in the east, resembles the Early Oligocene relief (Figure 2C).

[14] The eastern step, which accentuated an old Syrian Arc structure [Bar, 2009] forms the present-day Judea Western Mountains Front Step (WMFS). The central step, now buried under Neogene sediments below the present-day coastal plain of Israel (CPS), accentuated the Cretaceous continental slope [Gvirtzman, 1970; Gvirtzman et al., 2008; Bar, 2009]. The western step, bounding the deep Levant Basin block on the east approximately below the present-day continental slope, vertically accentuated a Senonian-Eocene, 10 km-wide fault zone. This step was termed by Gvirtzman et al. [2008] as the Levant Continental Margin Fault Zone (CMFZ).

[15] Further details about the formation and the burial of these steps are discussed by Gvirtzman et al. [2008], Bar [2009], and Steinberg et al. [2011]. Here, we briefly emphasize details that are important for the paleo-topography-bathymetry reconstruction. Most importantly, we note the rapid incision of submarine canyons across the steps (Ashdod, Afiq, El-Arish, and Hanna canyons, Figure 1A), which are each 1200–2000 m deep [Druckman et al., 1995a; Buchbinder et al., 2005; Bar, 2009]. The observed depth of each of these canyons approximately indicates the heights of the morpho-structural steps (discussed further below).

[16] In the following Oligocene-Miocene times, massive sedimentation in the deeper Levant Basin diminished the morphological expression of the CMFZ [Steinberg et al., 2011], while the CPS continued to dominate the seafloor morphology much longer [Bar, 2009]. During the Messinian Salinity Crisis [e.g., Hsü et al., 1973], the sea level of the Mediterranean Sea drastically dropped, and deep canyons were incised again across the continental margin. However, the Messinian incision was not as deep as the former Late Eocene-Early Oligocene incision and did not reach the bottom of the earlier canyons, as evident by the thick Oligo-Miocene sections preserved in the canyons [e.g., Druckman et al., 1995a]. Finally, during the Pliocene Nile derived sediments buried the coastal plain step. These sediments extended the continental shelf westwards and built the present-day continental slope.

[17] The present-day terraced structure of the eastern Mediterranean continental margin is composed of four relatively nondeformed blocks separated by the three morpho-structural steps described above (Figures 1A, 2C, and 2D). The four blocks are: ((1)) The Judea Mountains block, ((2)) the foothills and eastern coastal plain block, ((3)) the continental shelf block, and (4) the deep Levant Basin block. Two later tectonic phases, at the late Early Miocene and during the Middle to Late Pliocene, caused a longer wavelength regional uplift without a reactivation of the morpho-structural steps and, therefore, the terraced structure of the continental margin were maintained [Bar, 2009].

[18] The transition from the truncated, nearly 1000 m high, Judea Mountains plateau to the 1500 m deep Levant Basin, with its >5 km thick Late Tertiary sediments (Figure 2) raises questions such as what caused the contemporaneous inland emergence and basin subsidence and what were the contributions of tectonic mechanisms, sedimentation, and erosion to this process. To answer these questions, we used a backstripping analysis. This analysis depended greatly on constraining the paleo-water-depth from the Middle Eocene, prior to the period of tectonic deformation, and during that period, between the Early Oligocene to Middle Miocene.

2 Methods

2.1 Backstripping

[19] Thirteen columnar sections based on data from wells (some composed of two adjacent wells) were backstripped (for locations see Figure 1). One well from the Judea Mountains block (Ramallah-1 well), three wells along the coastal plain block (David-1 and Petah Tiqva-2, Hulda-1, Pleshet-1), three wells along the present-day coastline at the foot of the CPS (Jaffa-1, Hof Ashdod-1, Nezarim-1-Kefar Darom-1), and four wells in the continental shelf block (Yam Yafo-1, Yam-2, Yam West-1, Til-1). Two more columnar sections were based on seismic data from pseudowells in the Levant Basin as interpreted by Ben-Avraham et al. [2002], Netzeband et al. [2006], Gardosh et al. [2008b], and Steinberg et al. [2011].

[20] For the post Middle Eocene section, which is the focus of this study, unit subdivision was improved much beyond the lithostratigraphic division, based on the biostratigraphic revision of Bar [2009]. Information missing from wells that did not penetrate the base of the sedimentary section was completed from the nearest wells that penetrated the basement (Besor 1, David 1, Gevim 1) or from seismic data.

[21] A decompaction correction (thickness correction for compaction effect) followed the procedure described by Sclater and Christie [1980], using their parameters and those of Royden and Keen [1980]. Gradual unloading of the sedimentary column, considering paleo-water-depth and paleo-sea-level corrections, was applied according to equation ((1)) [Steckler and Watts, 1978].

display math(1)

where Y is the depth to the basement of a water-loaded basin (relative to the present-day sea level and positive downward), S* is the thickness of the decompacted sedimentary column, Wd is the paleo-water-depth, ΔSL is the paleo-sea-level (relative to the present sea level and positive upward), and ρm, ρs, and ρw are the densities of the mantle, sedimentary column, and water, respectively.

[22] For periods of exposure, the topography was considered instead of the bathymetry, as is seen in equation ((2)).

display math(2)

where H is the paleo-topographic elevation (positive upward).

[23] Paleo-water-depth, which is especially important in this study, was reconstructed utilizing geological data as explained below. Paleo-sea-level fluctuations [e.g., Miller et al., 2005], which are one to two orders of magnitude smaller than the paleo-water-depth correction in this study, were not considered.

2.2 Isostatic Calculation of Paleobathymetry

[24] The Middle Eocene period, just before the reactivation of the Levant continental margin, is a special time in the history of the Levant region, reflecting relative tectonic quiescence. Approximately 120 m.y. after the Early Mesozoic rifting and approximately 80 m.y. after the Early Cretaceous magmatism, we may assume that the Levant passive margin was approaching thermal equilibrium [Segev et al., 2011]. This implies that by then, the isostatic effect of lateral density variations in the mantle lithosphere were quite small (e.g., according to Leroy et al. [2008] the vertical motions resulting from thermal conduction from the period of 100 myr after stretching and onwards are in the order of 100 m). Therefore, we can practically assume that isostatic compensation was achieved at the depth of the deepest crustal root. This enabled us to balance the crustal column of the Levant Basin, with its unknown overlying water depth, with the crustal column of southern Israel, where the water depth could be estimated from sedimentological data:

display math(3)

where C and ρc are the thickness and the density of the crust, respectively, and the subscripts LB and SI refer to Levant Basin and Southern Israel, respectively.

[25] Using appropriate values of crustal and sediment thicknesses and densities, the relations in equation ((3)) are used to constrain the water depth of the Levant Basin in the Middle Eocene. Thicknesses of the crystalline crust are inferred from seismic refraction studies. Densities of the crystalline crust are calculated from seismic velocities using the empiric relations of Christensen and Mooney [1995]. Thicknesses and densities of the Middle Eocene sedimentary column, corrected for decompaction, are calculated using the procedure described above for backstripping. The water depth in southern Israel is estimated from sedimentological data.

[26] In order to evaluate the dependence of the backstripping tectonic analysis on the paleo-water-depth correction, for the model parameters, we selected two sets of values that provide constraints on minimum and maximum paleo-water-depths.

2.3 Morphological Reconstruction of Paleobathymetry and Paleotopography

[27] For the period postdating, the renewed tectonism, isostatatic balancing cannot be used for paleobathymetric estimations. On the other hand, a wealth of other geological data enables reconstructing a relatively detailed bathymetric-topographic profile from the Judea Mountains plateau all the way to the Levant Basin. First, we used relicts of flat morphological surfaces formed by abrasive and erosive processes as proxies for approximate paleo-sea-levels (zero elevation) during their formation. Next, we used known height differences across morpho-structural steps to reconstruct either paleotopography or paleobathymetry landward and seaward to the zero-elevation surfaces, respectively (Figure 3A). For example, a buried canyon provides a minimum height estimation for the paleo step through which it incised (Figure 3C). In a case of block tilting, the reconstruction took into consideration the possibility of some vertical overlap between steps (Figure 3B). These principles are applied below to reconstruct the paleo-profiles from the Judea Mountains plateau to the Levant Basin for the Early Oligocene and Middle Miocene.

Figure 3.

Principles for the reconstruction of paleobathymetry-topography profiles based on height differences across morpho-structural steps and on relicts of ancient abrasive and erosion surfaces, which were originally close to sea level. A. Estimation of paleo-relief assuming a subhorizontal state of structural blocks. In this case, water depth in the deep part of the basin is the sum of all submarine steps and similarly topographic elevation accumulates landwards. The depth and/or elevation are calculated with respect to a certain surface inferred to have been at sea level at a certain time and are representative for that time. B. Estimation of paleo-relief considering the possibility of tilted blocks with a partial overlap of vertical differences across the steps. C. Schematic figure illustrating how a canyon cross cutting a morpho-structural step is used to constrain the step's height.

[28] There is some controversy surrounding the magnitude of sea-level drop during the Messinian Salinity Crisis. Estimates range between hardly any sea level lowering [e.g., Lu, 2006] to a major drop of more than 1000 m, and almost complete desiccation [e.g., Hsü et al., 1973; Rouchy and Caruso, 2006]. Considering the isostatic response of the lithosphere to water unloading, if the Mediterranean basin did dry out, the paleogeography of its rims must have changed significantly at that time [e.g., Govers et al., 2009]. However, since this event was very short (possibly, within one precession cycle of ~ 21 kyr [Meijer and krijgsman, 2005]) and was caused by far field processes (closure of Atlantic-Mediterranean gateways) not related to local tectonism, we simply ignored it. We assume that this transient effect did not exert a major influence on the long-term geodynamic processes, which are the scope of this study.

3 Results

3.1 Paleobathymetry and Paleotopography Reconstruction

[29] Figure 4 presents paleobathymetric-topographic profiles for the Middle Eocene (A), Early Oligocene (B), and Middle Miocene (C), reconstructed by the method described above. To define the range of uncertainty each profile presents minimal and maximal estimations. For comparison, the present-day profile is also presented (D). In addition, all profiles are underlain by an independently prepared cross section of the decompacted sedimentary column (based on Figure 2), showing how unit thicknesses decreased with time until reaching their present-day thickness. Note, however, that these composite geological cross sections are entirely based on geological data of paleo-sediment thickness and paleo-water-depth, rather than calculated by sedimentary unloading (for one-dimensional backstripping see section 3.2 below).

Figure 4.

Reconstructions of variation with time of the paleobathymetry-topography along a simplified geological cross section running from the Judea Mt. plateau to the Levant Basin (location in Figure 1): A. Middle Eocene, B. Early Oligocene, and C. Middle Miocene. Figure 4D is a simplification of Figure 2, which represents the present-day situation. For each period, minimal and maximal relief estimations are presented on the left and right hand sides, respectively. Geological constraints are marked on the figures and explained in the text (values in meters). Unit thicknesses are decompacted as explained in the text. Note that the difference between Figures 4A and 4B is only related to tectonic deformation with no sedimentation, i.e., the thin and noncontinuous Late Eocene section is ignored.

3.1.1 Middle Eocene

[30] Sedimentological analysis of pelagic Early to Middle Eocene rocks in Israel indicates that a carbonate ramp gradually descended at that time from the region that eventually became inland Israel westward into the Levant Basin [Buchbinder et al., 1988]. In the Judea Mountains plateau, Eocene rocks are generally missing, but one single Early Eocene outcrop preserved near the city of Jerusalem [Lewy et al., 1995] indicates that, at least in the Early Eocene, seawater covered that region. The existence of Middle Eocene submarine slumps west of the Judea Mountains [Buchbinder et al., 1988] further indicates marine conditions throughout the Judea region. The angle of the unconformity between Senonian and underlying strata west of Jerusalem indicates that during the main phase of the Syrian Arc folding, only one third of the observed WMFS existed [Bar, 2009]. This means that the seafloor relief in the Middle Eocene was mild (Figure 4A). The exact water depth in the inland region is hard to estimate. However, given the mild seafloor relief and given that the Ramon anticline crest in southern Israel was within the photic zone, as indicated by rich nummulitic fauna [Benjamini, 1979; Buchbinder et al., 1988; Sneh, 1988], we estimate that water depth in nearby deeper (free-nummulitic) basins was between 100 and 500 m (Figure 4A).

[31] To constrain the possible range of water depths in the Levant Basin at that time utilizing equation ((3)), we used crustal and sediment thicknesses and densities as listed in Table 1. The two datasets in Table 1 present two possible end-member values for the required parameters that were combined selectively to constrain minimum and maximum paleo-water-depths of the Levant Basin. The results are 2 km and 3.5 km, respectively (Figure 4A).

Table 1. Parameters for Calculating Paleo-water-depth in the Southeastern Levant Basin (LB) During the Middle Eocene by Balancing its Crustal Column with That of Elat, Southern Israel (SI)a
 Southern Israel (SI)Levant Basin (LB)
Minimal Estimation of Bathymetry
  1. aThis calculation is based on the assumption of thermal equilibrium in the Middle Eocene as explained in the text (equation ((3))). The two datasets in the table represent two possible endmember values for the required parameters that were combined selectively to constrain minimum and maximum paleo-water-depths of the Levant Basin. The results of ~2 km and ~3.5 km, respectively, are marked in bold. The water depth for Elat is estimated from sedimentological data as explained in text. Mean densities of the crystalline crust are calculated from seismic velocities following the empiric relations by Christensen and Mooney [1995]. Asthenosphere density is taken as 3250 kg/m3.
WaterWater depth (m)Wd500This study2090Calculated (equation ((3)))
 Density (kg/m3)ρw1030 1030 
SedimentThickness (m)S*1100Gvirtzman [1997]7890Ben-Avraham et al. [2002]
 Density (kg/m3)ρS1950Gvirtzman [1997]2400This work
CrustThickness (m)C32,900Hofstetter and Bock [2004]9000Netzeband et al. [2006, Profile 1]
    Al-Damegh et al. [2005]  
 Density (kg/m3)ρC2870Seismic velocity - Weber et al. [2004]2840Seismic velocity - Netzeband et al. [2006, Profile 1]
 Maximal Estimation of Bathymetry
WaterWater depth (m)Wd100This study3570Calculated (equation ((3)))
Density (kg/m3)ρw1030 1030 
SedimentThickness (m)S*1100Gvirtzman [1997]5630Netzeband et al. [2006]
Density (kg/m3)ρS1950Gvirtzman [1997]2400This work
CrustThickness (m)C34,900Hofstetter and Bock [2004]9300Ben-Avraham et al. [2002]
Al-Damegh et al. [2005]
Density (kg/m3)ρC2840Seismic velocity - El-Isa et al. [1987]2900Seismic velocity - Ben-Avraham et al. [2002]

3.1.2 Early Oligocene

[32] The Early Oligocene reconstruction of Figure 4B is based on the following observations, which are described from east to west:

  1. [33] The Judea Mountains Plateau [Bar, 2009] is part of an extensive erosion surface previously known from Jordan and southern Israel [e.g. Picard, 1943; Quennell, 1958; Garfunkel and Horowitz, 1966; Zilberman, 1992]. During the Early Oligocene, this erosion surface had shaped the Judea Mountains Plateau and leveled the entire region to nearly sea level.

  2. [34] The subhorizontal unfolded geometry of the Early Oligocene erosion surface, which truncates the steeply folded layers of the WMFS, indicates that folding ended before the Oligocene. This constrains the duration of the main (two-thirds) folding phase between Late Eocene and Early Oligocene, so that by then the WMFS already reached its present height.

  3. [35] Pelagic Oligocene sedimentation in the western foothills of the Judea Mountains [Buchbinder et al., 2005] indicates that a deep-water environment existed at the foot of the WMFS, contemporaneous with truncation on the Judea Mountains Plateau.

  4. [36] Information about the CPS is obtained from the size of the submarine canyons that had crossed it and from the thickness of the younger filling sediments. By the Early Oligocene, a 1100 m deep canyon was incised by the Afiq channel [Druckman et al., 1995a] and up to 2000 m was incised by the Ashdod channel [Buchbinder et al., 2005]. Thus, the CPS at that time was ≥1100–2000 m high.

  5. [37] Thickness variations in Senonian to Middle Eocene sections across the CMFZ indicate that faulting along this zone started alongside with the Syrian Arc deformation, prior to the time interval discussed in this study [Gvirtzman et al., 2008]. However, canyon incision across a 1500–2000 m high step occurred only in the Oligocene, as indicated by the canyon fill penetrated in the Hanna-1 oil well and from seismic mapping [Gardosh et al., 2008a]. We therefore suggest that differential vertical motions and the formation of a bathymetric step occurred during the Oligocene.

[38] Altogether, these observations indicate that in the Early Oligocene, the Judea Mountains plateau was nearly at sea level and the WMFS had nearly reached its present morphological height of 500 m. From this, we inferred that the foothills and the eastern coastal plain block must have been 500 m below sea level. The base of the CPS was 2500–1600 m below sea level, as evident from the size of the ancient canyons. However, unlike the Judea Mountains and the foothills blocks that were subhorizontal during their formation, the ancient dip of the continental shelf block is poorly constrained, so the precise depth of the eastern edge of that block at the top of the CMFZ is uncertain. Therefore, the height of the CMFZ step inferred from the canyon that crossed it is not enough to constrain the absolute depth of the Deep Levant Basin block below it.

3.1.3 Middle Miocene

[39] The Middle Miocene reconstruction of Figure 4C is based on the following observations, once more described from east to west:

  1. [40] Shallow water, Middle Miocene carbonates capping flat abrasive surfaces west of the Judea Mountains plateau indicate that by the Middle Miocene the foothills block was raised to nearly sea level [Buchbinder et al., 1993].

  2. [41] Lack of deformation along the Afiq canyon axis indicates that structurally the CPS was not reactivated after the Early Oligocene. However, by Middle Miocene, times deposition of 300–400 m at the foot of the CPS [Bar, 2009] decreased the bathymetric step to a depth of 1600–800 m.

  3. [42] Cessation of faulting along the CMFZ is harder to constrain. The exact age of the uppermost strata, which are faulted, is not clear. They could be from the Oligocene or maybe even Early Miocene [Gvirtzman et al., 2008]. In any case, by the Middle Miocene, the CMFZ step had already reached most of its present-day structural height. Therefore, for the Middle Miocene reconstruction, we took the present-day height of the CMFZ step (~4 km) as marked in Figure 4C and filled it with the sedimentary section that existed in the Middle Miocene [Gardosh et al., 2008b]. This reduces the bathymetric expression of the CMFZ step to a depth of ~1200 m.

[43] Altogether, these observations indicate that when the western foothills region rose to nearly sea level, the Judea Mountains plateau must have reached an elevation of 500 m above sea level. The continental shelf and deep Levant Basin blocks were 700–1550 m and 1900–2750 m below sea level, respectively.

3.2 Backstripping

[44] Figure 5 presents four curves of basement subsidence through time for each of the 13 backstripped sections:

  1. [45] Basement subsidence curve, calculated from the cumulative thickness of the sedimentary column as measured in wells, with no decompaction correction (this curve is labeled “uncorrected”).

  2. [46] Basement subsidence curve as in curve 1, corrected for compaction during burial (labeled “decompacted”).

  3. [47] Basement subsidence curve as in curve 2, after gradual sedimentary layers unloading, with no water depth correction (labeled “sediment unloaded”).

  4. [48] Tectonically driven basement subsidence curve as in curve 3, with a correction for paleo-water-depth (uncertainty is shown by error bars). This final curve shows what would have been the subsidence of the basin, if only tectonic forces operated, with no sedimentation and erosion and the basin had been filled only with seawater. Note that for wells in the Judea Mountains and coastal plain blocks (Ramalla-1, David-1 and Petah-Tiqva-2, Hulda-1, and Pleshet-1), where shallow water sediments were continuously deposited from the Permian until the Turonian (270–89 Ma), the tectonic curves approximately coincide with the unloaded sediment curves. For the rest of the wells, where a pelagic environment prevailed most of the time and water depth is unknown, a broken line annotated by question marks schematically shows the tectonic curves.

Figure 5.

Subsidence curves reconstructed for13 oil wells across the study area (location in Figure 1). For each well, four curves are presented. (1) Basement subsidence curve obtained by the cumulative thickness of the sedimentary column as measured in wells with no decompaction correction (labeled “uncorrected”). (2) Basement subsidence curve as in curve 1, corrected for compaction during burial (labeled “decompacted”). ((3)) Basement subsidence curve of a water-filled basin as in curve 2, after gradual sedimentary layers unloading, with no water depth correction (labeled “sediment unloaded”). (4) Tectonically driven basement subsidence curve as in curve 3, with a correction for paleo-water-depth (labeled “tectonic”). Uncertainties in water depths are shown by error bars. The postulated paleo-water depth is marked by pale blue. During periods of shallow marine deposits the tectonic curve approximately coincides with the sediment unloaded curve (dashed black line).

[49] Gathering the decompacted curves of all wells in one figure emphasizes the main regional events (Figure 6A). The two rapid pulses of subsidence during the Permian-Triassic (270–220 Ma) and the Jurassic (200–150 Ma) are the result of two rifting phases that stretched the continental crust and formed the Levant Basin and its continental margin [e.g., Garfunkel and Derin, 1984]. These pulses were followed by thermal subsidence (postrifting lithospheric cooling) that was temporarily interrupted in the Early Cretaceous while a pulse of magmatism affected the inland region [Gvirtzman and Garfunkel, 1998; Gvirtzman et al., 1998]. Between 100 and 90 Ma, a noticeable decrease in subsidence rates occurred in all regions. At ca. 37 Ma, a series of tectonic and sedimentary processes began, as further discussed below.

Figure 6.

Composite presentation of the backstripping results. A. Decompacted curves of all wells demonstrating major regional processes seen in most wells. B. Four curves selected from Figure 6A to represent the four structural blocks: mountain backbone block, the foothills-east coastal plain block, the continental shelf block, and Levant Basin block. C. Four representative tectonic curves (same locations as in Figure 6B). These selected curves demonstrate the main tectono-sedimentary processes of the continental margin since their formation: Permo-Triassic (270–220 Ma) and Jurassic (200–150 Ma) rifting, subsequent prolonged thermal relaxation, and Late Tertiary renewal of tectonism (since ~37 Ma). Note: 1) the significance of the water depth correction indicated by comparing Figures 6B and 6C, and 2) the decrease in sedimentation rates at ~90 Ma (B) do not reflect a tectonic change (C) and that the opposite direction of vertical motion in the Late Tertiary (B) does not reflect opposite directions of tectonic forces (C). Details in text.

[50] A comparison of decompacted subsidence curves for four wells representing the four main regional blocks (Judea Mountains, coastal plain, continental shelf, and Levant Basin) (Figure 6B) with their tectonic subsidence curves (Figure 6C) emphasizes two important points:

  1. [51] The prominent decrease in the rate of total subsidence as seen at 100–90 Ma in land wells (Ramalla-1 and David-1 and Petah Tiqva-2 wells) does not reflect tectonic change, but rather, water deepening. The subsidence curves of these wells suggest that up until the Turonian rapid sedimentation of predominantly limestone and dolomite kept up with subsidence rates and maintained a shallow-water continental shelf. Then, at the beginning of the Senonian, deposition changed to pelagic chalks, sedimentation rates decreased, and the excess subsidence resulted in increasing water depth.

  2. [52] The total subsidence curves after 37 Ma show different trends (Figure 6B): uplift in the Judea Mountains block (Ramalla-1 well), no change in the coastal plain block (David-1 and Petah Tiqva-2 wells), and subsidence in the continental shelf (Yam West-1 well) and Deep Levant Basin blocks (pseudowell). However, following paleo-water-depth corrections, the tectonic curves (Figure 6C) show that since 37 Ma, all four regions had been subjected to an upward tectonic force. This does not mean that the entire region was actually uplifted. The relatively shallow landward region was uplifted, exposed, and truncated (Judea Mountains block), but the continental margin and especially the deep Levant Basin, subsided due to an excessive load of terrigenous sediment influx to the basin (Figure 6B). The coastal plain block in the transition zone between those two domains remained relatively stable (see further discussion below) (Figure 6B).

[53] To examine these results in detail, Figure 7 presents the backstripping results of all wells zeroed to a common base line that shows the motion relative to 37 Ma. Most of the total subsidence curves (Figure 7A) show no significant changes until the Pliocene. However, curves representing the tectonic component of the motion, set for the minimal (Figure 7B), maximal (Figure 7C) and inferred (Figure 7D) paleo-water-depth reconstructions, reveal three phases of activity. The first phase in the Late Eocene to Early Oligocene (~37–32 Ma) consists of differential motions that accompanied the formation of the three morpho-structural steps described above. Regions at the foot (west) of the steps subsided (Hof Ashdod-1, Til-1, Jaffa-1, Nezarim-1, and Kefar darom-1), whereas regions at the elevated (east) side of the steps uplifted (e.g. Ramalla-1). The two later tectonic phases, in the Early Miocene and in the Pliocene, consist of a regional uplift seen in almost all wells.

Figure 7.

Subsidence curves for the past 40 m.y. with a common base line zeroed at the Middle Eocene. A. Decompacted curves of all wells indicating subsidence and deposition of most regions. B–D: Tectonic curves of a water-loaded basin, considering (B) minimal, (C) maximal, and (D) inferred paleo-depth reconstructions. Note the differential tectonic movements during the period extending from the Late Eocene to the Early Oligocene (38–32 Ma), when some regions rose while others subsided (see text), and the regional uplift phases during the Early to Middle Miocene (21.5–15 Ma) and Late Pliocene (5.3–1.8 Ma), when the entire study area was subject to an upward tectonic force.

4 Discussion

4.1 Interpretation of the Backstripping Results

[54] One of the conclusions drawn from the backstripping analysis is that the enhanced subsidence of the Levant Basin and its continental margin since 37 Ma (Figures 6A, 6B, and 7A) was mainly due to sedimentary loading rather than to deep tectonic forces. On the contrary, the deep tectonic force in most of the study area was upward, at least since the Early Miocene (Figures 6C and 7B–D). As a result, the relatively shallow region of the Judea Mountains was uplifted, exposed, and truncated, whereas the relatively deep-water filled Levant Basin became an accommodation space for an extensive influx of terrigenous material which was accelerated by the erosion at the emerging inland margin and subsided under the sedimentary load in spite of the upward tectonic force.

[55] The differentiation of the relative effects of isostatic subsidence in response to sedimentary loading and tectonic uplift became possible due to careful paleo-water-depth reconstructions. Figure 8 further demonstrates the significance of the paleo-water-depth correction by showing how a given sedimentary section could result in three different tectonic interpretations due to differences in paleo-water-depth estimation. In this example, a basin composed of four sedimentary units is analyzed to examine what the tectonic motion was during the deposition of the uppermost unit (Unit d).This example resembles the question studied here regarding the Late Tertiary deposition in the continental margin. Interpretation A in Figure 8 assumes no accommodation space before the accumulation of Unit d and thus requires tectonic subsidence to account for the water-loaded basin geometry obtained by the unloading of Unit d. This resembles the interpretation of Tibor et al. [1992], who suggested that the southeast Mediterranean continental shelf subsided in the Pliocene due to the downflexing of the lithosphere under the load of the Nile delta deposits offshore Egypt. In contrast, Gvirtzman et al. [2008] preferred a B type interpretation (Figure 8), in which a preexisting, relatively deep-water basin filled by sediments can explain the rapid sedimentation in the Pliocene with no tectonic involvement. Nonetheless, the results of this study indicate a third, type C, interpretation (Figure 8). The paleo-water-depth reconstruction shows that the pre-Pliocene Levant Basin offshore Israel was much deeper than that obtained by sediment unloading alone (B in Figure 8). This requires an upward tectonic force (C in Figure 8 and Figure 6C) that partially restrained the isostatic subsidence response to the sedimentary load to explain the present-day situation.

Figure 8.

A schematic illustration showing the significance of paleo-water-depth reconstructions for tectonic interpretation. Note how the same sedimentary section can be interpreted in three different ways depending on the paleo-water-depth reconstruction.

[56] Figure 9 presents a simplified E-W cross section in four time stages, constructed to show the spatial and temporal variations in vertical motions of the crust and calculated tectonic forces across the different major subterrains (i.e., deep Levant Basin, continental shelf and slopes, coastal plain and mountain backbone). These pairs of cross sections are displayed three times for minimal (Figure 9A), maximal (Figure 9B), and inferred (Figure 9C) paleo-water-depth reconstructions.

Figure 9.

Reconstruction profiles of the top of the basement along an E-W cross section (location in Figure 1) in four time stages (Middle Eocene, Oligocene, Middle Miocene, and the present), considering (A) minimal, (B) maximal, and (C) inferred paleo-water-depth reconstructions. The four profiles at the lower part of each of the three pairs of graphs represent the top of the basement covered by the decompacted sediment and the water columns. The four profiles at the upper part of each of the three pairs of graphs represent the tectonically driven component of the crustal vertical motion calculated for a water-loaded basin. The temporal development of the tectonic profiles indicates that if tectonic forces operated alone without an external sedimentary loading most of the region would have been mainly uplifted. Note, however, that this is a two-dimensional presentation of a one-dimensional analysis, and that a flexural calculation would presumably provide different values, as discussed in the text.

[57] The four profiles at the lower part of each of the three pairs of graph sets (Figures 9A, 9B, and 9C) represent the top of the basement covered by the decompacted sediment and the water columns in the Middle Eocene, Oligocene, Middle Miocene, and the present. These basement profiles for all three paleo-water-depth statuses show a prominent rotation of the continental margin around a horizontal axis that produces a prolonged eastward uplift and westward subsidence.

[58] The four profiles at the upper part of each of the three pairs of graph sets represent the tectonically driven component of the crustal vertical motion calculated for a water-loaded basin. The temporal development of these profiles indicates that, if tectonic forces operated alone with no external sedimentary loading, most of the region would have been mainly uplifted. For the minimal paleo-water-depth reconstructions (Figure 9A), this tectonic upward force decreases from nearly 2 km in the inland region to nearly zero in the deep Levant Basin, where the only force that existed is sedimentary loading. On the other hand, for the maximal paleo-water-depth reconstruction (Figure 9B), the tectonic force in the Middle Eocene in the deep Levant Basin was downward (i.e., tectonic subsidence) whereas from the Oligocene to present, a similar upward tectonic component is obtained for the entire area, including the deep Levant Basin. Between these two end-member possibilities, our inferred paleo-water-depth reconstruction implies that the regional upward tectonic force decreased to nearly zero in the deep Levant Basin, but was still quite significant in the continental shelf-slope region (Figure 9C). This implies that without the inferred upward tectonic force, the shelf-slope region (and possibly the deeper basin) would have been today deeper than observed.

[59] Considering these results and their associated uncertainty, we suggest that the wavelength of the deep tectonic processes that uplifted and exposed the Arabian Platform in the Late Tertiary extended far westward across the continental margin and possibly reached the deep Levant Basin. However, this conclusion should be considered with caution because of the limitations of a 1D local isostasy calculation, which does not consider lithospheric strength and flexure. The ~150 km distance between the deep Levant basin and the Judea Mountains indicates that the flexural influence of these regions on one another is minor. Yet, the exact extent of the regional uplift force westwards still depends on 2D effects and cannot be determined accurately.

5 Conclusions

[60] Since the Late Eocene the NW Arabian Plateau began uplifting, whereas the Levant Basin began to subside and to accumulate a thick (> 5 km) sedimentary section. The present elevation of the rising inland region is about 1 km whereas the present water depth of the deep Levant Basin is about 1.5 km.

[61] A paleo-geographic study of the transition zone between the NW Arabian Platform and the deep Levant Basin during the Late Tertiary enabled a detailed reconstruction of paleo-water-depth from Israel's mountain backbone to the offshore. At the end of the Middle Eocene, ca. 37 m.y. ago, seawater covered the entire region. At that time, the water depth increased from a few hundred meters in the region that eventually became the Judea Mountains to 2–3.5 km in the deep Levant Basin. Since then, these regions moved in opposing vertical directions around a hinge line located approximately along the present-day coastline.

[62] Backstripping analysis of 13 columnar stratigraphic sections based on oil wells in central and offshore Israel indicates that both the inland and the Levant Basin were influenced by a regional upward tectonic force. This conclusion, which seems to be obvious in the elevated inland region, may appear enigmatic for the subsided Levant Basin. Our analysis indicates that the extensive flux of clastic sediments into the basin resulted in a downward isostatic response to the growing sedimentary load that was greater than the upward tectonic force and thus caused a net subsidence. However, the subsidence of the basin is more moderate compared to what could be expected if the isostatic response to the sedimentary loading alone would have been the only controlling factor.

[63] The magnitude of the upward tectonic force estimated varies from 1.5 to 2 km in the Judea Mountains plateau (inland) to 0–2 km in the Levant Basin, depending, respectively, whether minimum or maximum paleo-water-depth is assumed.

[64] The continental margin represents a transition zone between the inland range where isostatic response to erosion corroborated the upward tectonic force and the Levant Basin where isostatic response to the sedimentary load counteracted and overpowered the upward tectonic force.


[65] We thank R. Calvo for his help in computer programming, R. Siman-Tov for her help in biostratigraphic analysis, and B. Buchbinder for many stimulating discussions. We also acknowledge the assistance of the Associate Editor C. Faccenna, and the two constructive reviews by D. Cosentino and R. Govers. This research was supported by the Israeli Science Foundation (grant 768/07) and by the Israeli petroleum commissioner.