Skip to Main Content
Skip Nav Destination
This is an Open Access article distributed under the terms of the Creative Commons Attribution 3.0 License (http://creativecommons.org/licenses/by/3.0/).

This study summarizes the subsidence history and aspects of the geodynamic evolution of the South Caspian Basin based on the integration of geophysical observations, and subsidence and gravity modelling on selected two-dimensional (2D) profiles. This analysis implies the presence of an attenuated ‘oceanic-type’ crust in the northern portion of the South Caspian Basin, demonstrates characteristics of basin subsidence on variable crustal types and describes sediment-fill evolution in several different parts of the basin. Modelling conducted in this study shows that the observed pattern of subsidence and sedimentation in the South Caspian Basin can be explained by a process of thermal subsidence following Jurassic rifting and further enhanced subsidence that resulted from sediment-induced loading in the Late Tertiary, especially after a large-scale base-level fall after 6 Ma. Variation in crustal type is reflected in differences observed in the degree of subsidence and sediment fill in the overlying stratigraphy. The western part of the South Caspian Basin has subsided differently to the eastern part because of this difference in crustal type. This is also confirmed by gravity modelling, which shows that the South Caspian Basin crustal density is compatible with an oceanic composition in the western part of the South Caspian Basin: the crust in the eastern part of the basin, however, is thicker.

Gold Open Access: This article is published under the terms of the CC-BY 3.0 license.

The South Caspian Basin (SCB) is located between the mountain ranges of the Great Caucasus, Talesh, Alborz and Kopet Dagh (Fig. 1). The SCB is a large intermountain basin situated within the Alpine–Himalayan collision zone (Jackson et al. 2002; Reilinger et al. 2006; Kadirov et al. 2008), the unique characteristics of which are anomalously thick sediment cover, extensive compressional folding, mud volcanism and relatively thin crust. More than 25 km of sediment thickness has accumulated in the basin since its inception, mostly in the Tertiary (Allen et al. 2002; Artyushkov 2007). The northern boundary of the SCB is a submerged line of structural highs that forms the so-called Absheron–Pribalkhan Ridge or, simply, the Absheron Ridge (Fig. 1).
Fig. 1.

Shaded relief map showing the location of the South Caspian Basin and the surrounding mountain ranges. The black dashed-line polygon shows a region covered by the teleseismic receiver function study from Mangino & Priestley (1998); the red solid line shows an east–west crustal cross-section after Mammadov (1992) used for subsidence and gravity modelling by Brunet et al. (2003); the black solid lines show two 2D lithospheric-scale gravity models constructed by Granath et al. (2007); the yellow solid line shows the geological cross-section used in the 2D gravity modelling by Kadirov (2000) and Kadirov & Gadirov (2014), originally sourced from USSR Ministry of Geology (1990).

Fig. 1.

Shaded relief map showing the location of the South Caspian Basin and the surrounding mountain ranges. The black dashed-line polygon shows a region covered by the teleseismic receiver function study from Mangino & Priestley (1998); the red solid line shows an east–west crustal cross-section after Mammadov (1992) used for subsidence and gravity modelling by Brunet et al. (2003); the black solid lines show two 2D lithospheric-scale gravity models constructed by Granath et al. (2007); the yellow solid line shows the geological cross-section used in the 2D gravity modelling by Kadirov (2000) and Kadirov & Gadirov (2014), originally sourced from USSR Ministry of Geology (1990).

According to deep seismic refraction studies conducted in the 1960s (Mangino & Priestley 1998; Piip et al. 2012) and seismic reflection profiling (Glumov et al. 2004; Knapp et al. 2004) the thickness of sediments in the SCB is estimated to be between 25 and 30 km. The results of refraction studies and teleseismic receiver function analysis show that the region around SCB contains a 15–20 km-thick, low-velocity layer (Vp<4.8 km s−1) and a <10–18 km-thick high-velocity (Vp between 6.4 and 7.4 km s−1) ‘basaltic-type’ layer (Jackson et al. 2002), suggesting that the SCB could be underlain by an ‘oceanic-type’ (oceanic affinity) crust. Recent reinterpretation of the seismic refraction data of the SCB shows that the sedimentary layer is underlain by a thinned 10 km-thick crust, that the Moho discontinuity beneath the basin is at depths of between 35 and 40 km, and that velocities below the Moho discontinuity increase from 8.0 to 9.0 km s−1 (Piip et al. 2012). The SCB region with the thin, high-velocity crust is characterized along its northern edge by deep-focus earthquakes (Jackson et al. 2002; Khain 2005; Artyushkov 2007). This crust of oceanic affinity is also believed to be undergoing subduction beneath the Absheron Ridge (Golonka 2007). This contrasts with the western Turkmenistan area, east of the Caspian Sea, where teleseismic data suggest the crust is composed of 15–20 km of ‘granitic’ upper-crustal material and 20 km of ‘basaltic’ lower crust (Mangino & Priestley 1998; Piip et al. 2012).

Previous crustal studies of the SCB also included subsidence and gravity modelling by Brunet et al. (2003), and gravity modelling by Granath et al. (2007) and Kadirov & Gadirov (2014) (Fig. 1). The modelling results in these studies were constrained by a number of Soviet-era deep reflection lines and refraction data. More recent ultra-deep reflection profiles acquired over the last decade were used by Knapp et al. (2004), Mammadov (2008), Egan et al. (2009) and Green et al. (2009). They have provided new insights into basin structure and evolution by means of their subsidence modelling and structural restoration.

Despite numerous studies, many details of the internal crustal structure and evolution of the basin are unclear. Uncertainty over the crustal type and composition of the crust underlying the SCB is discussed in a variety of studies, such as Artyushkov (2007), Knapp et al. (2004), Glumov et al. (2004) and Mammadov (2006, 2008). The majority of authors assume that there is a subduction of South Caspian lithosphere underneath continental lithosphere of the Central Caspian Basin (Khalilov et al. 1987; Granath et al. 2000; Allen et al. 2002; Knapp et al. 2004; Golonka 2004, 2007; Egan et al. 2009; Green et al. 2009). However, insufficiently interpreted earthquake databases, sparse deep seismic refraction and reflection data coverage, as well as poorly constrained subsidence modelling, cannot definitely confirm or deny the presence of oceanic crust or explain the limited extent of such crust in the SCB. Therefore, in this study we describe ‘oceanic-type’ crust, implying crustal thicknesses and properties similar to oceanic crust, without inferring definite crustal type.

This study synthesizes previous studies with geophysical observations available to the Geological Institute of Azerbaijan (GIA) and BP, including a number of deep two-dimensional (2D) seismic reflection profiles, some of which have been published in Knapp et al. (2004). These seismic profiles have record lengths of up to 20 s two-way time (TWT) and reveal deep basin structure. Interpretation of these seismic profiles, and other offshore and onshore seismic and well data, was accumulated into an integrated set of depth and isopach maps across the entirety of the SCB. The three structural ‘geoseismic profiles’ presented in this paper were combined from the various 2D seismic reflection profiles and also include integrated published information on the deep crustal structure of the SCB.

The tectonostratigraphic framework of the SCB is summarized in Figure 2, being a modified stratigraphic column from Abdullayev et al. (2012). Plate tectonic reconstruction for the SCB and palaeo-environment interpretations for key tectonostratigraphic units shown in this stratigraphic column can be found in Jones & Simmons (1996), Geological Institute of Azerbaijan (2003), Golonka (2007), Hudson et al. (2008) and Van Baak (2010).
Fig. 2.

Stratigraphic column for the South Caspian Basin, Azerbaijan with major geological events and local stratigraphic stages (suites) indicated. Modified from Abdullayev et al. (2012) and showing the average sediment thicknesses for each key stratigraphic unit (yellow brackets). Stratigraphic ages are from Popov et al. (2006) and Hudson et al. (2008). Formations: PS, Productive Series; NKG, Post-Kirmaky Shaly Suite; NKP, Post-Kirmaky Sandy Suite; KS, Kirmaky Suite; PK, Under-Kirmaky Sandy Suite; KAS, Kalin (or Qala) Suite; Meot, Meotian; Pont, Pontian; Serr, Serravallian; Burd., Burdigalian; Lang., Langhian; Aq, Aquitanian.

Fig. 2.

Stratigraphic column for the South Caspian Basin, Azerbaijan with major geological events and local stratigraphic stages (suites) indicated. Modified from Abdullayev et al. (2012) and showing the average sediment thicknesses for each key stratigraphic unit (yellow brackets). Stratigraphic ages are from Popov et al. (2006) and Hudson et al. (2008). Formations: PS, Productive Series; NKG, Post-Kirmaky Shaly Suite; NKP, Post-Kirmaky Sandy Suite; KS, Kirmaky Suite; PK, Under-Kirmaky Sandy Suite; KAS, Kalin (or Qala) Suite; Meot, Meotian; Pont, Pontian; Serr, Serravallian; Burd., Burdigalian; Lang., Langhian; Aq, Aquitanian.

A Jurassic-age origin of the SCB has been proposed by a number of researchers (Zonenshain & Le Pichon 1986; Otto 1997; Granath et al. 2000; Brunet et al. 2003). Basin opening is believed to have been caused by a back-arc-rifting episode behind subduction of the Neotethyan Ocean in the Mid-Jurassic–Early Cretaceous (Berberian 1983; Zonenshain & Le Pichon 1986; Golonka 2007). While some authors propose an alternative age of basin opening in the Paleocene–Eocene (Adamia et al. 1977; Allen et al. 2002; Kazmin & Verzhbitskii 2011), or a pull-apart mechanism for the basin formation (Sengör 1990), we assume Jurassic back-arc rifting as a cause of the SCB opening in this paper.

Jurassic-age opening is associated with volcanic sediments penetrated in the Saatly superdeep well (onshore Azerbaijan), where more than 4400 m of Jurassic basalts, andesites, diorites and tuffs were encountered (Shikhalibeyli et al. 1988). No known Jurassic-age volcanic rocks have been penetrated offshore and not much is known of their offshore extent. Knapp et al. (2004) described, from 2D reflection data, a very prominent north-dipping reflector reaching maximum depths of 26–28 km, which they interpreted as a basement–cover boundary of Jurassic age. This reflector has also been identified on 2D seismic profiles that we have used in this study and is onlapped by what is assumed to be a post-Jurassic succession. No obvious down-to-basin faulting of this level can be discerned from 2D seismic profiles.

The stratigraphic column in Figure 2 shows four key post-Jurassic megastratigraphic complexes that can be described in terms of the sediment fill:

  • A Cretaceous–Early Palaeogene sequence (145–33.9 Ma) denotes a shelf to basin transition of carbonate and clastic rocks on the margins of the back-arc rift (Golonka 2007). The Great Caucasus Proto-Caspian back-arc basin was undergoing rifting that finished some time before Santonian, as evidenced by volcanic activity in Georgia (Golonka 2004). Egan et al. (2009) interpreted the Mesozoic rocks of the Great Caucasus as representative of deposition in a passive margin/rift setting, with a transition from the shallow-marine carbonates on the southern edge of Scythian–Turan Platform to deep-water mudstones. The Paleocene–Eocene sediments are exposed in the easternmost Great Caucasus area as the ‘Koun’ Formation. These are predominantly mudstone units deposited in slope to deep-water conditions throughout the Scythian Platform (Golonka 2007). Onshore outcrop and well data suggest thicknesses of 2–5 km for the Cretaceous interval, and up to 2 km of Paleocene and Eocene rocks (Geological Institute of Azerbaijan 2003). For offshore SCB, we observe between 5 and 11 km total thickness for Cretaceous, Paleocene and Eocene combined, with reflective packages being significantly below well control. Unit thickness is greatest in the northern portion of the SCB south of the Absheron Ridge, as shown in isopach maps published by Green et al. (2009). Seismic mapping and depth conversion of the north-dipping basal reflector shows that depth to this sequence varies from 14 km in the southern part of the SCB to between 20 and 26 km in the area south of the Absheron Ridge (Green et al. 2009). Nothing is known about the distribution, provenance and exact age of these rocks offshore, but this interval is thought to contain mostly deep-water clastics and carbonates by analogy with the basin margins.

  • A Oligocene–Late Miocene sequence (33.9–6 Ma) is exposed in the Great Caucasus and consists of largely marine clastics, known to occur within a wide region across the Caspian and Black Sea (Golonka 2007). Oligocene and Miocene rocks of the SCB have been deposited during a compressional episode, related to the collision of the Arabian continent with southern Eurasia (Allen et al. 2002; Hudson et al. 2008). The Oligo-Miocene Maykop, Diatom, Sarmatian, Meotian and Pontian formations (Fig. 2) are of wide geographical extent, and are exposed in western parts of the Absheron Peninsula and in the Great Caucasus. Intervals of locally high total organic carbon (TOC) occur within the Maykop and Diatom formations, and constitute the main source rocks for the hydrocarbon accumulations of the SCB. The age of the boundary between the Maykop Series and the overlying Middle–Late Miocene Tarkhanian and Diatom Series is believed to be located close to the base of Burdigalian age at about 15.9 Ma (Hudson et al. 2008). For onshore Azerbaijan, all of the Oligo-Miocene formations vary in thickness between about 0.5 and 3 km from north to south across the Absheron Peninsula; however, much of the thickness variation could be tectonic in origin as this unit is an important detachment zone in the SCB structures (Devlin et al. 1999). From offshore seismic mapping, we estimate Oligo-Miocene thickness to be between about 1 km in the central parts of the SCB and 3 km in the north, in areas adjacent to the Absheron Ridge. The isopach map for the Maykop, published in Green et al. (2009), shows thicknesses under the Absheron Ridge, the Absheron Peninsula and the Eastern Great Caucasus thrust belt, where there has been tectonic thickening and structural repetition.

  • The Late Miocene–Early Pliocene ‘Productive Series’ (6–2.6 Ma) is an important stratigraphic episode that contains the main clastic reservoirs of the SCB in oil and gas fields such as Azeri-Chirag-Guneshli and Shah-Deniz. The Productive Series (PS) is well known from wells and exposures onshore, offshore wells and seismic data. The PS consists of regionally extensive fluvio/deltaic sandstones, interbedded with regionally extensive lacustrine mudstones acting as seals. Both Green et al. (2009) and Abdullayev et al. (2012) related the formation of PS to large-scale, possibly Messinian-age (6–5.5 Ma) base-level fall that isolated the SCB from the global oceans. The magnitude of this event was estimated to be in the order of 1.5–2 km of vertical base-level fall and more than 900 km of basinwards shift of onlap between the Pontian-age highstand deltaic margin in the Central Caspian and first PS age onlap in the SCB (Abdullayev et al. 2012). This base-level fall integrated the continental river drainage of the Russian Platform and Central Asia into the SCB, thus creating a sediment sink that accumulated about 6 km of sediment thickness over an extremely short timespan (2–3 Ma). The PS across the offshore SCB is relatively unstructured and progressively onlaps pre-existing stratigraphy above the base-level fall unconformity (Abdullayev et al. 2012).

  • Late Pliocene and Pleistocene (2.6 Ma–present). Finally, the PS is capped by an Akchagyl marine transgression episode that reconnects the SCB to the global oceans. Following this transgression, a deeper-water lacustrine environment was established in response to a continued compressional episode related to the Arabia–Eurasia collision (Allen et al. 2002). Pleistocene shelf-margin complexes of the Absheron Formation develop in this lacustrine environment. These shelf-margin complexes prograded into a deep-water lake developing slope and basin-floor turbidites (Abdullayev 2000). The compressional episode responsible for the formation of the major offshore buckle folds and structures is mainly a Late Pliocene–Pleistocene event (Devlin et al. 1999; Egan et al. 2009). Structural growth and rapid progradation of shelf margins during this time also created conditions for submarine slope failures, which were sourced along the basin margin and from adjacent anticlines (Richardson et al. 2011).

In this study we continue and expand upon subsidence modelling conducted by Green et al. (2009), which describes the influence of rapid sediment loading on the total subsidence of Mesozoic-age ‘oceanic-type’ extended crust during the Late Tertiary. The age of rifting in Green et al. (2009) was assigned to a notional Late Jurassic date (145 Ma) and was assumed to be instantaneous or limited in duration. Our modelling assumes the same age for the start of the rifting.

Initial rift subsidence caused by fault deformation and crustal attenuation is followed by long-term gradual subsidence caused by lithospheric cooling and is termed thermal subsidence (Parsons & Sclater 1977; McKenzie 1978; Sclater & Christie 1980). Basin burial curves for the SCB in Brunet et al. (2003) and Egan et al. (2009), using typical compaction trends for the basin, demonstrate exponentially decreasing thermal subsidence after an initial rifting episode. A study by Egan et al. (2009) showed that thermal subsidence in the SCB rapidly decreased between 100 and 90 myr ago, and the remaining total subsidence rate was relatively low up to the Oligocene. However, in the Miocene, the total subsidence process intensified, and the basin bathymetry decreased from between 2 and 3 km to less than 1 km due to a rapid infill by clastic sediments of Pliocene PS (Mammadov 2008). The total subsidence and sedimentation rate during that period was anomalous for basins of comparable history; the SCB is truly unique in this respect. Allen et al. (2002) and Green et al. (2009) showed that total water-loaded tectonic subsidence of the SCB is between 5.5 and 6.5 km on average, and that around 30% (2 km) of this tectonic subsidence has been created since 6 Ma (Allen et al. 2002). It is also known that the SCB is out of isostatic equilibrium (Kadirov & Gadirov 2014) and negative Bouguer anomalies reach up to −120 mGal (Fig. 3a). It is estimated that up to 2 km of tectonic subsidence is uncompensated and, presumably, results from compressive forces acting on the SCB plate since the Pliocene (Zonenshain & Le Pichon 1986).
Fig. 3.

(a) Depth to basement map integrating reflection seismic and published data from Glumov et al. (2004). (b) Bouguer gravity anomaly (mGal) modified from Kadirov & Gadirov (2014). Modelled profiles are shown in red and are numbered.

Fig. 3.

(a) Depth to basement map integrating reflection seismic and published data from Glumov et al. (2004). (b) Bouguer gravity anomaly (mGal) modified from Kadirov & Gadirov (2014). Modelled profiles are shown in red and are numbered.

To understand the subsidence evolution of the SCB and variations in subsidence across the basin rather than along a single profile, integrated and iterative subsidence modelling was conducted on three regional cross-sections. Modelling included five key steps:

  • Flexural backstripping including simple assumptions for crustal stretching and thermal subsidence (Green et al. 2009) performed in ‘FlexDecomp’ software provided by Badley Geoscience Ltd.

  • Whole-crust (and lithosphere) extensional forward modelling with thermal subsidence performed in ‘Stretch’ software provided by Badley Geoscience Ltd.

  • Two-dimensional flexural backstripping using the constraints generated by the forward modelling. This provides a better consistency between the two modelling procedures (Kusznir et al. 1995). A comparison here can be made between the total sediment thickness of the model and the present-day sediment thickness to see whether a match can be achieved.

  • One-dimensional subsidence modelling to constrain the values of 2D crustal stretching models using 1D backstripping and the strain-rate inversion method (White 1994). This has been modelled using ‘backstrip1d’ script, written by Dr Alistair Crosby (BP, formerly from the Cambridge Basin Research Group).

  • Gravity profile modelling conducted using ‘GMsys’ software by Geosoft to constrain the subsidence modelling results.

The modelling workflow has a number of limitations and assumptions. The results of the modelling have been used only qualitatively and in conjunction with mapped observations to interpret crustal structure and basin evolution of the SCB.

Three modelled profiles across the SCB are shown by numbers 1–3 in Figure 3, where they overlie depth to basement and Bouguer anomaly maps, respectively. These profiles were iteratively modelled using the workflow outlined above. The assumptions and parameters used in the modelling are described below.

For the surfaces used in the modelling exercise, we have used nine seismic horizons taken from either an interpreted set of depth-converted seismic horizons available from BP and GIA, or from published maps where the former were not available. Published sources include maps of Top Basement, depth maps and key isopachs by Guliyev et al. (2003), Glumov et al. (2004) and Green et al. (2009), which extend onshore around the SCB. The interpreted seismic horizons (with ages assigned) were: Top Basement (145 Ma); Base Oligocene ‘Maykop’ (33.9 Ma); Lower Miocene ‘Top Maykop’ (15.9 Ma); Base Lower PS (6 Ma); Intra-Lower PS (5.6 Ma); Base Pliocene or Base of Middle PS (5.3 Ma); Base of Upper PS (4 Ma), Top PS (2.6 Ma); and Seabed (0 Ma).

Figure 3a shows the most basal of the maps, interpreted as Top Basement, which was generated by merging basement surface shown by Green et al. (2009), with onshore maps from Guliyev et al. (2003) and Glumov et al. (2004). For the gravity input to the modelling, we used the Bouguer gravity field of the SCB, as shown in Kadirov (2000) and Kadirov & Gadirov (2014), which is a publicly available map set. This Bouguer gravity map is shown in Figure 3b.

Each profile was flexurally backstripped to restore to the post-rift palaeobathymetry, as in the method shown in Roberts et al. (1998). Flexurally backstripping a profile sequentially ‘strips off’ successively decompacted and unloaded sedimentary layers from the total stratigraphy. For each layer in the stratigraphic column, the backstripping process unloads and then decompacts the underlying sediments (restoring palaeo-sediment thickness) and, finally, calculates the equivalent water load. The sections were backstripped with the thermal subsidence incorporated into the modelling using first an assumed stretch factor, (β), value of β=1 implying no crustal deformation and then β=100 to approximate the unlimited extension of ‘oceanic-type’ crust. Constraints of the forward model were then used to vary the β-factor between these two extremes.

Reverse modelling of compaction in the ‘FlexDecomp’ software follows the relationships defined by Sclater & Christie (1980). The program default values for the compaction parameters are set for a 50:50 shale:sand mix. From regional understanding, the lithology for the geoseismic sections was assumed to be largely mudstone with some sandstone input into the PS (up to 50%) from which average compaction parameters were calculated. The modelling presented here uses a principal value for effective elastic thickness (Te) of 3 km that is generally considered typical for a standard rift basin (White 1994; Roberts et al. 1998), which the SCB could have been following its formation. Sensitivities with higher Te of 10 km have also been modelled in the SCB and it was shown (Green et al. 2009) that, while local-scale geometries within the SCB changed, this did not impact the overall subsidence history along the modelled cross-section.

The sea-level curve used for palaeobathymetry corrections until the Pliocene is the eustatic curve of Haq et al. (1987), which is a default in ‘FlexDecomp’ software but can be customized. During the backstripping process at 6 Ma, a sea-level change of −1.5 km was introduced to correct for the substitution of water-filled by air-filled subsidence, as a result of the regional base-level fall at the base of the PS.

Crustal parameters were further refined in the second stage of modelling, which involved forward modelling of the lithosphere, capable of generating sufficient subsidence to accommodate the observed stratigraphic thickness of the SCB. The ‘Stretch’ software we used for modelling utilizes a modification (simplification) of the flexural-cantilever model (Kusznir & Egan 1989; Kusznir et al. 1991; 1995; Green et al. 2009) that models lithosphere extension as pure shear only without reference to specific basement fault geometries. This model differs from the 1D model of McKenzie (1978) in that the flexural response to loading is incorporated. A uniform pure-shear rift subsidence incorporating flexural isostasy was modelled. The forward-modelling workflow was conducted by first assuming 35 km of unstretched crust, instantaneously stretching it using a ductile pure-shear model and then allowing the crust to thermally subside for 145 Ma. Lithospheric thickness in the model was assumed to be 125 km, and the density of the crust and mantle were assumed to be 2800 and 3300 kg m−3 respectively. Modelled basin stratigraphy was then compared to the observed stratigraphic overlay to see whether a match could be achieved. The modelling was then repeatedly optimized to produce a best match, with the output being a 2D profile of the β-factor that is the best match to modern-day stratigraphy.

The ‘Stretch’ software does not model the flexural subsidence related to compression and, for this reason, stretching was overestimated, exceeding β=20 in some areas near the Absheron Ridge. In areas away from the Absheron Ridge modelled β-factors were much lower. The modelling allowed a qualitative analysis of crustal stretching, but needed further calibration to provide realistic basin evolution scenarios.

Following forward modelling, 2D flexural backstripping was re-run using β-factors derived from the forward-modelling process, and post-rift palaeobathymetry was restored as in Green et al. (2009). Despite using modelled β-factors that achieve model match with stratigraphy, some of the excessive ‘stretching’ required to do so can be explained by other factors.

The final step in estimating the tectonic subsidence and benchmark β-factors from forward modelling is a 1D backstripping and strain-rate inversion of a number of pseudo-wells posted along the geoseismic sections and the application of the method described in White (1994). The modelling takes observations of horizon depth, age and palaeo-water depth from a well, and calculates the tectonic (i.e. water-loaded) subsidence history to determine the variation of lithospheric strain rate with time that is required to match the tectonic subsidence observations. Essentially, the 1D inverse method calculates the rifting history that best fits the backstripped well data. A 1D backstripping code also makes predictions of stretching factors and heat flux through time. This method is 1D and also assumes that all subsidence is due to rifting. Therefore, an assumption of no further crustal stretching following Late Jurassic rifting was made in the 1D backstripping, and subsidence post-36 Ma was not fitted to the model. This allowed the determination of pure tectonic subsidence from initial crustal stretching and thermal subsidence, without the influence from other causes, which is not possible in ‘FlexDecomp’ software. The solutions shown are the smoothest fit solutions that fit the data and which produce the most realistic extension and strain rate history, not simply the best attainable fit (White 1994). Limited numbers of data points that can be reliably dated in the early part of basin history is a limitation to this method.

Finally, crustal structure was validated using gravity modelling along the same profiles and cross-checked with ultra-deep seismic reflection profiles where available. The gravity model was applied using forward modelling, including best fit of the initial model to the observed gravity profile, recalculation of the anomaly, and comparison of final observed and modelled anomalies. An update of a geological and layer density model to match observed and modelled gravity profiles was made, enabling a better-informed decision on crustal structure. It is with this fit that a potential for a thin ‘oceanic-type’ crust beneath the SCB, with an increased density of 3000 kg m−3, can be ascertained. Using this view of crustal structure and sediment fill, a model was constructed incorporating both seismic observations and the results of modelling to describe differences in crustal structure and evolution of basin fill inside the SCB.

This profile runs approximately north–south and is built up of available surface sets of various 2D reflection seismic profiles, some of which are ultra-deep 2D TWT profiles. It is broadly similar to the profile modelled in Green et al. (2009) but differs by being constructed to avoid structurally complex parts of the Absheron Ridge. The model consists of nine layers, corresponding to seismically mappable intervals covering the Jurassic to the present (Fig. 4). A portion of the profile, north of the Absheron Ridge, represents continental crust of the Central Caspian Basin, while the southern portion is located over the SCB crust. The profile is characterized by an increase in thickness, from south to north, in the SCB portion of Mesozoic–Lower Palaeogene age. The full profile was subjected to a restoration with ‘FlexDecomp’ software initially with constant β-factors, based on the assumption of changing crustal types across the profile. The northern portion of the profile includes significant crustal shortening under the Absheron Ridge and cannot be confidently restored using this methodology.
Fig. 4.

Geoseismic Profile 1 through the South Caspian Basin and the Absheron Ridge showing the variable thicknesses across the profile. The profile is similar in orientation to the profile used in Green et al. (2009).

Fig. 4.

Geoseismic Profile 1 through the South Caspian Basin and the Absheron Ridge showing the variable thicknesses across the profile. The profile is similar in orientation to the profile used in Green et al. (2009).

The flexural backstripping was complemented by forward lithospheric modelling along the profile. The results of forward modelling are shown in Figure 5a, where the model restores the profile to a maximum sediment thickness of about 25 km, which corresponds to a stretching factor, β, of c. 20 in the thickest portion of the Mesozoic–Palaeogene wedge, decreasing to average values of β=2 in the south (Fig. 5b) that are typical of attenuated continental crust. The northern portion of the profile is located on fairly thick Scythian Plate continental crust and cannot be confidently modelled with high stretch factors. Values above β=20 for oceanic crust are fairly unrealistic and are enhanced by flexural effects of the load over the Absheron Ridge. However, these crustal parameters show a good match between forward modelled and observed stratigraphy across the SCB portion of the profile (Fig. 5c).
Fig. 5.

Results of the forward lithospheric modelling of Profile 1. (a) 2D modelling results with crustal extension to match the stratigraphy, which indicates attenuated crust. (b) Resultant value of the initial β-factor, reaching maximum values of β=20. (c) Modelling results with stratigraphy overlay, with the red dashed line representing the depth to basement and the black solid line representing the approximate Base Tertiary reflector. (d) Restored palaeobathymetry at the Top Jurassic (145 Ma) using the crustal β-factor derived from the modelling, and showing average water depths of between 2.5 and 3 km.

Fig. 5.

Results of the forward lithospheric modelling of Profile 1. (a) 2D modelling results with crustal extension to match the stratigraphy, which indicates attenuated crust. (b) Resultant value of the initial β-factor, reaching maximum values of β=20. (c) Modelling results with stratigraphy overlay, with the red dashed line representing the depth to basement and the black solid line representing the approximate Base Tertiary reflector. (d) Restored palaeobathymetry at the Top Jurassic (145 Ma) using the crustal β-factor derived from the modelling, and showing average water depths of between 2.5 and 3 km.

The resultant laterally variable β-factors were then applied to the profile, restoring Top Basement surface to palaeobathymetry following rifting (Fig. 5d). Results restore bathymetry depths to about 2–2.5 km, similar to values in Green et al. (2009), corresponding to present-day sediment thickness of up 20 km. Although there is a good match over the majority of the area, this is achieved with β-factors that may be too high in some places. This suggests that other subsidence mechanisms are required to keep β-factors below values of 10, which are more acceptable for highly stretched crust.

Along the profile, there is a zone of highly fluctuating reconstructed palaeobathymetry that is probably not realistic (Fig. 5d). This zone is located over implied underthrusting of ‘oceanic-type’ crust under and to the north of the Absheron Ridge. This may have resulted in significant local tectonic thickening in the South Caspian sedimentary sequence immediately to the south of the Absheron Ridge. The observed anomalous depth and thickness can be explained by compressional shortening of the Tertiary stratigraphy. A more realistic modelling of this zone requires a precise geological interpretation of the basin margins.

One-dimensional backstripping using strain-rate inversion flow (White 1994) on the pseudo-well from the central part of the profile just to the south of the Absheron Ridge was undertaken to compare subsidence analysis with 2D forward modelling and strip away potentially dramatic flexural effects (Fig. 6). An attempt has been made to ignore recent subsidence events, which are not extensional in nature, by not fitting data younger than 36 Ma and using an increased smoothing factor. This achieves a β-factor of 7 and water-loaded subsidence of 6–6.5 km, which still implies a highly attenuated crust at this location. However, these values of stretching factor are much smaller than values from 2D forward modelling that best fit the stratigraphy. This suggests that the substantial portion of tectonic subsidence at this location was not generated by extension.
Fig. 6.

(a) Geoseismic Profile 1 showing the location of the pseudo-well where the 1D subsidence model applied (shown by a red dot on the map); (b) water-loaded subsidence history using a constrained model; (c) initial β-factors derived from the constrained model. The black dashed line shows 36 Ma, after which no subsidence fit was applied.

Fig. 6.

(a) Geoseismic Profile 1 showing the location of the pseudo-well where the 1D subsidence model applied (shown by a red dot on the map); (b) water-loaded subsidence history using a constrained model; (c) initial β-factors derived from the constrained model. The black dashed line shows 36 Ma, after which no subsidence fit was applied.

In addition to flexural backstripping and forward lithospheric modelling, gravity modelling of the profile was conducted to support the modelling results as outlined above. Density values and thicknesses have been assigned to layers of the crust and the upper mantle similar to those of Kadirov & Gadirov (2014). These were then modified in order to achieve the best fit between the observed gravity and the calculated gravity: mantle densities were kept at 3300 kg m−3; ‘oceanic-type’ crustal densities were kept at 2950 kg m−3; ‘continental-type’ crustal densities were varied between 2750 and 2800 kg m−3; Mesozoic-age sequence was assigned values of 2600 kg m−3; and the Tertiary-age sequence was averaged at 2400 kg m−3. These modified values of density and layer thickness were kept within bounds provided by the 2D forward modelling and 1D backstripping, and by available knowledge of elevated Moho depths and sediment thicknesses (Glumov et al. 2004). The gravity model matches the observed gravity field of Profile 1 (Fig. 7), which supports the presence of denser, ‘oceanic-type’ crust (5–7 km) underneath the northern portion of the SCB, below and to the north of the Absheron Ridge. This gravity observation is also well supported by reflection seismic and earthquake observations (Kadirov et al. 2008; Kadirov & Gadirov 2014). This area corresponds to large negative Bouguer anomaly (Fig. 3b) and is best modelled as a root of South Caspian crust that has displaced the normal lithospheric mantle, creating a large isostatic anomaly, as shown in Granath et al. (2000). The southern part of the profile has a very small negative Bouguer anomaly, and can be modelled with thicker (+10 km) and less dense crust (we used lower continental crust densities of 2800 kg m−3). The values of the Bouguer gravity field slowly increase towards positive values, south of the SCB.
Fig. 7.

Gravity modelling results of Profile 1 showing the observed (dotted line) and modelled (solid line) Bouguer gravity anomaly. The density model shows: sediment layers in yellow and green; the upper crust in light purple; the lower crust in deeper purple; the oceanic crust in red-brown.

Fig. 7.

Gravity modelling results of Profile 1 showing the observed (dotted line) and modelled (solid line) Bouguer gravity anomaly. The density model shows: sediment layers in yellow and green; the upper crust in light purple; the lower crust in deeper purple; the oceanic crust in red-brown.

A west–east geoseismic profile extracted from available map sets (Fig. 3) was similarly modelled to describe crustal type and subsidence to the south of the Absheron Ridge. Profile 2 (Fig. 8) begins onshore at the Absheron Peninsula near Baku, traverses the western SCB to the SE for 290 km, reaching the modern-day Turkmen Shelf, where it turns and changes direction to the NE and runs for 220 km to the Balkhan Mountains of Western Turkmenistan. The profile is oblique to some of the ultra-deep seismic lines.
Fig. 8.

Geoseismic Profile 2 through the South Caspian Basin showing variable structural styles and sediment thickness differences between the western and eastern parts of the basin, especially in the Mesozoic–Palaeogene section. Productive Series thicknesses are generally consistent around 5–6 km.

Fig. 8.

Geoseismic Profile 2 through the South Caspian Basin showing variable structural styles and sediment thickness differences between the western and eastern parts of the basin, especially in the Mesozoic–Palaeogene section. Productive Series thicknesses are generally consistent around 5–6 km.

The model consists of nine layers corresponding to seismically mappable intervals covering the Jurassic to the present (Fig. 8). The profile is characterized by significant changes in the Mesozoic–Palaeogene thickness along the profile, being 10–5 km thick in the western sector and 4–5 km thick in the eastern sector. Deformation style also changes along the profile, with Tertiary-age compressional folding in the western sector and largely undeformed parallel-lying stratigraphy in the eastern sector.

Similar to Profile 1, after initial flexural backstripping, without thermal input, we conducted lithospheric forward modelling and calculated the amount of crustal stretching required to accommodate the available stratigraphy. Figure 9a shows the results of forward modelling within both syn-rift (yellow) and post-rift (blue) successions. The amount of stretching required along this profile is significant and the calculated total β-factors are quite high. Figure 9b shows the calculated β-factor, where values increase from approximately β=3 in the east to more than β=20 in the west. The western part of the model corresponds to ‘oceanic-type’ highly attenuated crust (5–7 km thick), while the eastern part with its low β-factor values (10–15 km thick) shows the presence of less attenuated crust. The boundary of these two implied crustal types lies on the modern shelf margin of the palaeo-Amudarya Delta in the Turkmenistan sector of the Caspian Basin and represents a significant tectonic step.
Fig. 9.

Results of forward lithospheric modelling of Profile 2. (a) 2D modelling results with crustal extension to match the stratigraphy. (b) Resultant value of the initial β-factor, exceeding values of β=20. (c) Modelling results with stratigraphy overlay with the red dashed line representing the depth to basement, and the black solid line representing the approximate Base Tertiary reflector. (d) Flexural backstripping results with restored palaeobathymetry at the Top Jurassic (145 Ma) using the crustal β-factor derived from the modelling, and showing average water depths of between 2.5 and 3 km.

Fig. 9.

Results of forward lithospheric modelling of Profile 2. (a) 2D modelling results with crustal extension to match the stratigraphy. (b) Resultant value of the initial β-factor, exceeding values of β=20. (c) Modelling results with stratigraphy overlay with the red dashed line representing the depth to basement, and the black solid line representing the approximate Base Tertiary reflector. (d) Flexural backstripping results with restored palaeobathymetry at the Top Jurassic (145 Ma) using the crustal β-factor derived from the modelling, and showing average water depths of between 2.5 and 3 km.

Figure 9c depicts the forward-modelling result with the stratigraphic overlay, and shows a reasonable match between observed and modelled stratigraphy, allowing for differential subsidence between western and eastern parts of the profile. Figure 9d shows the Top Basement surface restored to palaeobathymetry just after rifting, using laterally variable β-factors derived from forward modelling. The section shows a palaeobathymetry of approximately 3 km in the western portion of the profile and approximately 1.0–1.5 km in the eastern portion.

One-dimensional subsidence and strain-rate inversion modelling was conducted on two pseudo-wells (Fig. 10). Location 1 is in the middle of the Mesozoic–Palaeogene depocentre (Fig. 10a), about 20 km south of the Absheron Ridge. Ignoring any subsidence beyond 36 Ma, the 1D model predicts initial β-factor values of approximately 4.5 (Fig. 10c), with water-loaded subsidence of 6 km (Fig. 10b). This is significantly lower than β-factors derived from the forward modelling required to match stratigraphy, which would imply that a large component of the tectonic subsidence is not of extensional origin even this far south. Location 2 (Fig. 10a) used in 1D subsidence and strain-rate inversion modelling is located on the eastern portion of the profile, where the Mesozoic–Palaeogene depocentre thins towards the Turkmenistan coast. One-dimensional modelling for this location predicts initial β values of approximately 2.5 (Fig. 10e), with water-loaded subsidence of 4.5 km (Fig. 10d). The β-factors derived from 2D forward modelling required to match stratigraphy (Fig. 10a) are similar to the values from 1D modelling. This implies that at this location crustal extension and thermal subsidence could explain most of the history of the tectonic subsidence.
Fig. 10.

(a) Geoseismic Profile 2 showing the location of two pseudo-wells where a 1D subsidence model was applied (shown by red dots on the map with pseudo-wells labelled as locations 1 and 2). (b) Tectonic subsidence history using the constrained model for Location 1, where the black dashed line shows the 36 Ma cut-off after which no subsidence fit was applied. (c) Initial β-factors derived from the constrained model for Location 1. (d) Tectonic subsidence history using the constrained model at Location 2. (e) Initial β-factors derived from the constrained model at Location 2.

Fig. 10.

(a) Geoseismic Profile 2 showing the location of two pseudo-wells where a 1D subsidence model was applied (shown by red dots on the map with pseudo-wells labelled as locations 1 and 2). (b) Tectonic subsidence history using the constrained model for Location 1, where the black dashed line shows the 36 Ma cut-off after which no subsidence fit was applied. (c) Initial β-factors derived from the constrained model for Location 1. (d) Tectonic subsidence history using the constrained model at Location 2. (e) Initial β-factors derived from the constrained model at Location 2.

There is an observed difference in gravity field signature and crustal structure between the two pseudo-well locations on the profile. The eastern portion of the profile crosses the so-called ‘Godin Massif’ feature (Mammadov 2008), which is characterized by a rapid change of gravity towards a positive Bouguer anomaly (+10 to +20 mGal), and a thicker crust between 12 and 15 km (Granath et al. 2000, 2007; Glumov et al. 2004; Mammadov 2008; Kadirov & Gadirov 2014). The profile trails the edge of the gravity anomaly with values close to +0 to +10 mGal. The gravity modelling calculation matches the observed gravity field of Profile 2 (Fig. 11), which supports the presence of denser, ‘oceanic-type’ crust (5–6 km thick) in the western portion of the profile and thicker ‘continental-type’ crust (10–15 km thick) to the east, with a change observed across the boundary. The western portion of the profile (represented by Location 1) corresponds to thicker sediment cover and the eastern portion of the profile (represented by Location 2) corresponds to thinner sediment cover, with the change occurring across a gravity-observed, deep-seated tectonic block. Mammadov (2008) interprets this tectonic feature as a fragment of the continental crust or ‘microcontinent’ on the border of the ‘oceanic’ SCB.
Fig. 11.

Gravity modelling results of Profile 2 showing the observed (dotted line) and modelled (solid line) Bouguer gravity anomaly. The density model shows: sediment layers in yellow and green; the upper crust in light purple; the lower crust in deeper purple; the oceanic crust in red-brown.

Fig. 11.

Gravity modelling results of Profile 2 showing the observed (dotted line) and modelled (solid line) Bouguer gravity anomaly. The density model shows: sediment layers in yellow and green; the upper crust in light purple; the lower crust in deeper purple; the oceanic crust in red-brown.

To continue to reconstruct the initial basin shape of the ‘oceanic-type’ crust in the western part of the SCB, modelling along a third profile was conducted using the same workflow as described for the previous two profiles. The profile starts at the Absheron Peninsula near Baku and ends close to the Iranian shore of the Caspian Sea (Fig. 3). This profile was chosen to represent a continuous north–south cross-section across the western portion of basin.

The geological cross-section of this profile (Fig. 12) illustrates the thicker Mesozoic-aged sediments in the northern and middle part of the profile thinning towards the south, with the Tertiary wedge also thinning slightly from north to south. The change in thickness of the Mesozoic wedge in this profile is more gradual and not as dramatic as in Profile 1. The northern portion of the profile is involved in the compression and uplift of the Absheron Peninsula, and is, therefore, not a proper reconstruction. This portion of the profile is involved in compressional deformation and uplift related to collisional tectonics, therefore modelling for this part is inaccurate, as it does not reflect long-wavelength flexural effects.
Fig. 12.

Geoseismic Profile 3 through the northern and southern parts of the South Caspian Basin showing variable structural styles and sediment thicknesses.

Fig. 12.

Geoseismic Profile 3 through the northern and southern parts of the South Caspian Basin showing variable structural styles and sediment thicknesses.

Figure 13a, b demonstrates the results of forward modelling with syn-rift and post-rift sedimentary fill. The middle and northern part of the profile south of the Absheron Ridge shows a resulting crustal thickness of roughly 5–6 km. In the southern part of the profile, the crust thickens rapidly to between 10 and 15 km. The resulting crustal-type variation can be seen in Figure 13a, where the β-factor value varies from around β=3 in the south to approximately β=13 in the north, where the crust is modelled as being thinnest. These values are probably too high to be realistic for oceanic crust as the software is not adapted to the problem, but they are required to accommodate the deposited post-rift stratigraphy during modelling and show a good match with stratigraphy (Fig. 13c). These results indicate that the total thickness of sediments, interpreted along Profile 3, can be accommodated by subsidence of the sediment succession overlying an ‘oceanic-type’ crust on the northern part of the profile, slowly transitioning to an attenuated continental crust in the south, but lacking the sharp transition observed in Profile 2.
Fig. 13.

Results of the forward lithospheric modelling of Profile 3. (a) 2D modelling results with crustal extension to match the stratigraphy. (b) Resultant value of the initial β-factor. (c) Modelling results with stratigraphy overlay, with the red dashed line representing the depth to basement and the black solid line representing the approximate Base Tertiary reflector; data outside the black square is interpolated. (d) Flexural backstripping results with restored palaeobathymetry at the Top Jurassic (145 Ma) using crustal β-factor derived from the modelling, and showing average water depths around 3 km.

Fig. 13.

Results of the forward lithospheric modelling of Profile 3. (a) 2D modelling results with crustal extension to match the stratigraphy. (b) Resultant value of the initial β-factor. (c) Modelling results with stratigraphy overlay, with the red dashed line representing the depth to basement and the black solid line representing the approximate Base Tertiary reflector; data outside the black square is interpolated. (d) Flexural backstripping results with restored palaeobathymetry at the Top Jurassic (145 Ma) using crustal β-factor derived from the modelling, and showing average water depths around 3 km.

One-dimensional subsidence modelling was conducted on the pseudo-well extracted from the middle portion of Profile 3 (Fig. 14) to compare with results of 2D modelling on the same profile, as well as similar 1D modelling results shown on profiles 1 (Fig. 6) and 2 (Fig. 10). Tectonic subsidence in this pseudo-well was calculated to be about 4.5 km and the initial β-factor on this profile was calculated at β=3.5, which is only slightly smaller than the β=5 from the 2D forward modelling at the same location.
Fig. 14.

(a) Geoseismic Profile 3 showing the location of the pseudo-well where the 1D subsidence model was applied (shown by a red dot on the map). (b) Water-loaded subsidence history using the constrained model. (c) Initial β-factors derived from the constrained model. The black dashed line shows 36 myr ago, after which no subsidence fit was applied.

Fig. 14.

(a) Geoseismic Profile 3 showing the location of the pseudo-well where the 1D subsidence model was applied (shown by a red dot on the map). (b) Water-loaded subsidence history using the constrained model. (c) Initial β-factors derived from the constrained model. The black dashed line shows 36 myr ago, after which no subsidence fit was applied.

Gravity modelling along this profile was also carried out (Fig. 15). The best match between the observed and modelled gravity signature can be achieved by the density model using a 15 km-thick ‘oceanic-type’ crust layer in the south of the profile, thinning to between 7 and 8 km in the north of the profile, where it is potentially consumed under the Absheron Ridge. The thickness of the sediment layer in the profile gradually increases from 15 km in the southern part of the profile to 20 km in the northern part.
Fig. 15.

Gravity modelling results of Profile 3 showing the observed (dotted line) and modelled (solid line) Bouguer gravity anomaly. The density model shows: sediment layers in yellow and green; the upper crust in light purple; the lower crust in deeper purple; the oceanic crust in red-brown.

Fig. 15.

Gravity modelling results of Profile 3 showing the observed (dotted line) and modelled (solid line) Bouguer gravity anomaly. The density model shows: sediment layers in yellow and green; the upper crust in light purple; the lower crust in deeper purple; the oceanic crust in red-brown.

The iterative modelling applied to the SCB was used as a tool to confirm a number of key observations about the crustal structure and stratigraphic evolution of the SCB. It also allowed for the synthesis of modelling and observation, subdividing the geodynamic history of the SCB into several tectonostratigraphic stages, from basin formation to present day.

One-dimensional backstripping and strain-rate inversion modelling was undertaken in four locations, shown by red circles in Figure 16, for comparison with 2D models. Shown on the map are values of the β-factor derived from both the 2D ‘Stretch’ forward model and from the 1D subsidence model (the latter shown in brackets), which excludes the last 36 Ma from tectonic subsidence. An excessive β-factor has been calculated for the northern part of the SCB from the 2D modelling, and relatively lower values have been calculated elsewhere for both 1D and 2D modelling, being quite similar for the eastern part of the SCB (Fig. 16). For the 2D forward model, the β-factor maximum exceeds β=20 but, for the constrained model, maximum values are around β=7 (Profile 1).
Fig. 16.

Interpretation of the crustal structure shown over the depth to basement map. This map shows a comparison of β-factors derived from the modelling of 2D lines and 1D strain-rate inversion.

Fig. 16.

Interpretation of the crustal structure shown over the depth to basement map. This map shows a comparison of β-factors derived from the modelling of 2D lines and 1D strain-rate inversion.

Subsidence modelling of the SCB shows that excessive sediment thickness of between 24 and 25 km can be explained by significant crustal extension in the Late Jurassic followed by thermal subsidence, and sediment loading and compaction, with sediment loading and compression increasing dramatically in the Late Tertiary. The amount of crustal stretching is high with values of up to β=7, indicative of what could be called highly attenuated or even ‘oceanic-type’ crust (Fig. 16). Gravity modelling also confirms an ‘oceanic-type’ thin and dense crustal model.

Variations in crustal structure controlled the magnitude of subsidence following rift extension, with the modelling showing that strongly marked thickening of the Mesozoic–Palaeogene-age sediment wedge in the NW portion of the SCB is located in an area of transition from a thicker attenuated crust (15 km) to very thin ‘oceanic-type’ crust. This marked change occurred between the western and eastern portions of the basin through a structural step also observed on gravity data (Fig. 15).

The conceptual model, compatible with these observations, shows that the current SCB geometry could be interpreted as an inherited rift basin fabric. The largest of the modelled β-factors (β>5) can be placed within an elongated zone of the rifted margin, orthogonal to Jurassic-age SW–NE extension (outlined by a green dashed polygon in Fig. 16). The northern boundary of this depocentre is the Absheron Ridge and the western boundary of this depocentre is the West Caspian Fault as described in Kadirov et al. (2008). This proposed rift depocentre is flanked to the east by a deep-seated NE–SW basement feature (shown as a black dashed line on Fig. 16) separating the depocentre from a block of attenuated continental crust named the ‘Godin micro-continent’ by Mammadov (2008). This deep-seated basement feature could be interpreted as a transform fault and underlies a major Pliocene-age structural ridge named ‘Abikh Uplift’ or ‘Alov’ (Guliyev et al. 2003). The southern boundary of the depocentre is not clear but may coincide with a gravity maximum near the Sefidrud Delta. These key crustal boundaries have a major impact on the tectonic framework of the basin and related depositional styles. There are also indications that another potential rift depocentre might have been present on the SE margin of the SCB in the so-called Pre-Alborz (Mazandaran) deep, where the depth to the base of the Pliocene exceeds 8 km (Guliyev et al. 2003; Glumov et al. 2004) and basement features have been identified on seismic data (Babayev & Gadzhiev 1998).

Within the area of the proposed NW rift depocentre, total subsidence following rifting was significantly enhanced through thermal subsidence and sediment loading, which is reflected in the calculated β-factors and observed sediment thicknesses. Outside of this depocentre, Mesozoic–Palaeogene age thickness is relatively limited. In addition, the Mesozoic–Palaeogene-age sequence probably also contains some (probably limited) syn-rift stratigraphy. The original extent and orientation of the rift depocentre and the mechanism for the generation of oceanic crust within this basin are impossible to explain from any type of subsidence modelling attempted, but the highly attenuated nature of the crust over part of the basin hints at complicated rift fabric geometries along the northern margin of the Mesozoic Neotethyan Ocean.

The Late Tertiary evolution of the SCB was a complex interplay of continuing sediment loading and far-field flexural effects related to the uplift of the Caucasus and the potential subduction of the SCB plate under the Absheron Ridge. Late Tertiary compression and subduction has a large, but localized, impact on tectonic (water-loaded) subsidence, especially in the region south of the Absheron Ridge, and accounts for the discrepancy between the 2D forward modelling and the 1D subsidence modelling results outlined above. It can be estimated from the discrepancy that about 1.5–2 km of additional tectonic subsidence has been generated by Late Tertiary compression after about 34 Ma. Alternative causes for the excessive tectonic subsidence in the western portion of SCB could be related to ‘dynamic’ or residual topography in response to convective downward flow (downwelling) within the mantle (Winterbourne et al. 2009). Studies by Crosby & McKenzie (2009) and Winterbourne et al. (2009) demonstrate that positive and negative deviations from the established global age–depth trend for oceanic crust are common and can be as large as ± 1 km. They mostly correlate with long-wavelength gravity anomalies over oceanic crust, such as the large negative anomaly under the SCB. Negative residual topography could also explain the isostatically uncompensated nature of the western part of the SCB.

In addition, simple subsidence modelling cannot account for the marked increase in the rate of deposition observed during the Pliocene PS, not related to specific tectonic compressional events. Total subsidence during the Miocene was significantly enhanced as a result of a 1.5–2 km Messinian-age base-level fall at 6 Ma with a basinward shift of over 1000 km, followed by rapid deposition of terrigenous and lacustrine sediments of the PS. The importance of this large-scale base-level fall event for sedimentation and subsidence is covered in detail in Green et al. (2009) and Abdullayev et al. (2012).

The total subsidence that has produced the current basin shape may result from the interplay of all these factors. Modelling shows that extremely attenuated ‘oceanic-type’ crust, with high calculated β-factors over the NW portion of the basin, is necessary in order to generate the observed sediment thickness. It is not possible to accommodate this sediment thickness with a single average crustal thickness and a single β-factor as an unrealistic initial water depth will result. This, again, implies the complicated nature of the crustal structure across the basin.

Iterative subsidence modelling of the SCB presented in this study shows that the observed character of subsidence and sedimentation in the SCB can be explained by a process of thermal subsidence and sedimentary loading over ‘oceanic-type’ or attenuated continental crust, and is compatible with similar studies by Egan et al. (2009) and Green et al. (2009). Forward modelling of lithospheric extension and gravity modelling confirms the presence of variable crustal types, and infers a tentative Jurassic-age boundary of the rifted margin and its basin: delineated by the Absheron Ridge in the north, the West Caspian Fault in the west and the ‘Godin Massif’ in the east. In addition, gravity modelling shows that negative anomalies north of the Absheron Ridge correspond to dense ‘oceanic-type’ crust, possibly undergoing underthrusting or subduction under the continental crust of the Scythian Plate, and the negative anomaly south of the Absheron Ridge corresponds to an approximately 10 km-thick Mesozoic–Palaeogene wedge overlying same attenuated ‘oceanic-type’ crust. Positive gravity anomalies in eastern and southern portions of the SCB correspond to thinner sediment successions, overlying relatively less attenuated crust.

Late Tertiary compression in the western SCB has had a more localized impact on subsidence, especially in the region south of the Absheron Ridge. Modelling presented in this study indicates that an additional compression mechanism is probably necessary to reduce crustal stretching to more realistic stretching factors, but the amount of that additional subsidence varies between western and eastern parts of the SCB.

Further investigation of the modelling methods is recommended to address underlying deficiencies of extensional models, and more constraints from new reflectivity refraction or gravity data, as well as better depth constraints, are required for the validation of these models.

The largest volume of sediments was accumulated in the basin from the Oligocene onwards. Average sedimentation rates were 500–600 m Ma−1, reaching a maximum during the Lower Pliocene PS. The sediment input and, therefore, total subsidence dramatically accelerated following Messinian-age base-level fall at 6 Ma. More than 6 km of terrestrial and lacustrine sediments were deposited during the PS, which represents a large-scale lowstand systems tract capped by Akchagyl marine deposits. The impact of PS deposition on subsidence was a rapid acceleration of subsidence rate as a result of sediment loading, not a tectonic event (Green et al. 2009). The effect of this sediment loading is most pronounced in the western portion of the SCB, which contains thicker PS.

The process of basin evolution and subsidence in the SCB modelled in the study can be subdivided into five main tectonostratigraphic stages: (1) Late Jurassic-age rifting, assumed to be of limited duration; (2) Mesozoic–Lower Palaeogene thermal subsidence; (3) Oligocene–Lower Miocene-age increase in subsidence rates, probably related to the onset of compression and Caucasus uplift; (4) rapid sedimentation and sediment loading during the PS following a dramatic base-level fall; and (5) renewed Pleistocene-age compression, folding and hydrocarbon entrapment with continuing rapid subsidence. The hydrocarbon systems of the SCB are quite unique because of such rapid sedimentation.

This work mostly reflects doctoral research conducted by the corresponding author at the Geological Institute of Azerbaijan. The authors express their gratitude to academician Professor P. Z. Mammadov (Azerbaijan State Oil Academy) and Dr G. Riley (BP) for their advice and comments, and also to Dr A. Crosby for advice of the 1D subsidence workflow. The authors are also grateful to BP Exploration Ltd for permission to show map data, and to Badley Geoscience Ltd (in particular to Dr A. Roberts) for permission to use the ‘FlexDecomp’ and ‘Stretch’ modelling software, as well as providing helpful comments. The authors would also like to thank Dr M. F. Brunet, S. Egan and L. Eppelbaum for reviewing the manuscript and providing valuable suggestions for its improvement.

1.
Abdullayev
,
N.R.
2000
.
Seismic stratigraphy of the Upper Pliocene and Quaternary deposits in the South Caspian Basin
.
Journal of Petroleum Science and Engineering
 ,
28
,
207
226
.
2.
Abdullayev
,
N.R.
Riley
,
G.W.
Bowman
,
A.
2012
. Regional controls on lacustrine sandstone reservoirs: the Pliocene of the South Caspian Basin. In:
Baganz
,
O.W.
Bartov
,
Y.
Bohacs
,
K.
Nummedal
,
D.
(eds)
Lacustrine Sandstone Reservoirs and Hydrocarbon Systems
 .
American Association of Petroleum Geologists, Memoirs
,
95
,
71
98
.
3.
Adamia
,
S.A.
Buadze
,
V.I.
Shavishvili
,
I.D.
1977
. The great caucasus in the phanerozoic; A geodynamic model. In:
Jankovic
,
S.
(ed.)
Metallogeny and Plate Tectonics in the Northeastern Mediterranean
 .
Faculty of Mining Geology, Belgrade University
,
Belgrade
,
215
229
.
4.
Allen
,
M.B.
Jones
,
S.
Ismail-Zadeh
,
A.
Simmons
,
M.
Anderson
,
L.
2002
.
Onset of subduction as the cause of rapid Pliocene-Quaternary subsidence in the South Caspian Basin
.
Geology
 ,
30
,
775
778
.
5.
Artyushkov
,
E.V.
2007
.
Formation of the superdeep South Caspian basin: subsidence driven by phase change in continental crust
.
Russian Geology and Geophysics
 ,
48
,
1002
1014
.
6.
Babayev
,
D.Kh.
Gadzhiev
,
A.N.
1998
.
About existence of new massif in the south of the South Caspian basin on complex geophysical data
. In:
Reports of the 2nd Azerbaijan International Geophysical Conference
,
Baku, Azerbaijan
,
30 September–2 October 1998
, Azerbaijan National Geophysics Committee,
240
(in Russian).
7.
Berberian
,
M.
1983
.
The southern Caspian: a compressional depression floored by a trapped, modified oceanic crust
.
Canadian Journal of Earth Sciences
 ,
20
,
163
183
.
8.
Brunet
,
M.-F.
Korotaev
,
M.V.
Ershov
,
A.V.
Nikishin
,
A.M.
2003
.
The South Caspian Basin: a review of its evolution from subsidence modelling
.
Sedimentary Geology
 ,
156
,
119
148
.
9.
Crosby
,
A.G.
McKenzie
,
D.
2009
.
An analysis of young ocean depth, gravity and global residual topography
.
Geophysical Journal International
 ,
178
,
1198
1219
.
10.
Devlin
,
W.
Cogswell
,
J.
et al
1999
.
South Caspian Basin: young, cool, and full of promise
.
GSA Today
 ,
9
,
1
9
.
11.
Egan
,
S.
Mosar
,
J.
Brunet
,
M.-F.
Bochud
,
M.
Kangarli
,
T.
2009
. Subsidence and uplift mechanisms within the South Caspian Basin: insights from the onshore and offshore Azerbaijan Region. In:
Brunet
,
M.-F.
Wilmsen
,
M.
Granath
,
J.W.
(eds)
South Caspian to Central Iran Basins
 .
Geological Society, London, Special Publications
,
312
,
219
240
, https://doi.org/10.1144/SP312.11
12.
Geological Institute of Azerbaijan
2003
.
Atlas of Lithological–Paleogeographical Maps of Azerbaijan
 .
Geology Institute of the Azerbaijan (GIA), National Academy of Sciences
,
Baku, Azerbaijan
.
13.
Glumov
,
I.F.
Malovitsky
,
Y.P.
Novikov
,
A.A.
Senin
,
B.V.
2004
.
Regional Geology and Petroleum Potential of the Caspian Sea
 .
Nedra
,
Moscow (in Russian)
.
14.
Golonka
,
J.
2004
.
Plate tectonic evolution of the southern margin of Eurasia in the Mesozoic and Cenozoic
.
Tectonophysics
 ,
381
,
235
273
.
15.
Golonka
,
J.
2007
, Geodynamic evolution of the South Caspian Basin. In:
Yilmaz
,
P.O.
Isaksen
,
G.H.
(eds)
Oil and Gas of the Greater Caspian Area
 .
American Association Petroleum Geolgists, Studies in Geology
,
55
,
17
41
.
16.
Granath
,
J.W.
Soofi
,
K.A.
Baganz
,
O.W.
Bagirov
,
E.
2000
.
Gravity modelling and its implications to the tectonics of the south Caspian basin. Paper presented at the
American Association of Petroleum Geologists Inaugural Regional International Conference, Extended Abstracts
.
Istanbul, Turkey
,
9–12 July 2000
.
17.
Granath
,
J.W.
Soofi
,
K.A.
Baganz
,
O.W.
Bagirov
,
E.
2007
. Gravity modeling and its implications to the tectonics of the South Caspian Basin. In:
Yilmaz
,
P.O.
Isaksen
,
G.H.
(eds)
Oil and gas of the Greater Caspian Area
 .
American Association Petroleum Geologists, Studies in Geology
,
55
,
43
46
.
18.
Green
,
T.
Abdullayev
,
N.
Hossack
,
J.
Riley
,
G.
Roberts
,
A.
2009
. Sedimentation and subsidence in the South Caspian Basin, Azerbaijan. In:
Brunet
,
M.-F.
Wilmsen
,
M.
Granath
,
J.W.
(eds)
South Caspian to Central Iran Basins
 .
Geological Society, London, Special Publications
,
312
,
241
260
, https://doi.org/10.1144/SP312.12
19.
Guliyev
,
I.S.
Levin
,
L.E.
Fyodorov
,
D.L.
2003
.
Hydrocarbons Potential of the Caspian Region: (System Analysis)
 .
Nafta Press
,
Baku, Azerbaijan
.
20.
Haq
,
B.U.
Hardenbol
,
J.
Vail
,
P.R.
1987
.
Chronology of fluctuating sea levels since the Triassic
.
Science
 ,
235
,
1156
1167
.
21.
Hudson
,
S.
Johnson
,
C.L.
Efendiyeva
,
M.
Rowe
,
H.
Feyzullayev
,
A.A.
Aliyev
,
C.S.
2008
.
Stratigraphy and geochemical characterization of the Oligocene–Miocene Maikop series: implications for the paleogeography of Eastern Azerbaijan
.
Tectonophysics
 ,
451
,
40
55
.
22.
Jackson
,
J.
Priestley
,
K.
Allen
,
M.
Berberian
,
M.
2002
.
Active tectonics of the South Caspian Basin
.
Geophysical Journal International
 ,
148
,
214
245
.
23.
Jones
,
R.W.
Simmons
,
M.D.
1996
.
A review of stratigraphy of Eastern Paratethys (Oligocene-Holocene)
.
Bulletin of the Natural History Museum London (Geology)
 ,
52
,
25
49
.
24.
Kadirov
,
F.A.
2000
.
Application of the Hartley transform for interpretation of gravity anomalies in the Shamakhy-Gobustan and Absheron oil- and gas-bearing regions, Azerbaijan
.
Journal of Applied Geophysics
 ,
45
,
49
61
.
25.
Kadirov
,
F.A.
Gadirov
,
A.H.
2014
.
A gravity model of the deep structure of South Caspian Basin along submeridional profile Alborz–Absheron Sill
.
Global and Planetary Change
 ,
114
,
66
74
.
26.
Kadirov
,
F.A.
Mammadov
,
S.A.
Reilinger
,
R.
McClusky
,
S.
2008
.
Some new data on modern tectonic deformation and active faulting in Azerbaijan (according to Global Positioning System Measurements)
.
Proceedings of Azerbaijan National Academy of Sciences, The Sciences of Earth
 ,
1
,
82
88
.
27.
Kazmin
,
V.G.
Verzhbitskii
,
E.V.
2011
.
Age and origin of the South Caspian Basin
.
Oceanology
 ,
51
,
131
140
.
28.
Khain
,
V.E.
2005
.
The problem of origin and age of South Caspian Basin and its probable solutions
.
Geotectonics
 ,
1
,
40
44
(in Russian).
29.
Khalilov
,
E.N.
Mekhtiyev
,
Sh.F.
Khain
,
V.E.
1987
.
About some geophysical data confirming the collision origin of Greater Caucasus
.
Geotectonics
 ,
2
,
54
60
(in Russian).
30.
Knapp
,
C.C.
Knapp
,
J.H.
Connor
,
J.A.
2004
.
Crustal-scale structure of the South Caspian Basin revealed by deep seismic reflection profiling
.
Marine and Petroleum Geology
 ,
21
,
1073
1081
.
31.
Kusznir
,
N.J.
Egan
,
S.S.
1989
. Simple-shear and pure-shear models of extensional sedimentary basin formation: application to the Jeanne d’Arc basin, Grand Banks of Newfoundland.
American Association Petroleum Geologists, Memoirs
,
46
,
305
322
.
32.
Kusznir
,
N.J.
Marsden
,
G.
Egan
,
S.S.
1991
. A flexural-cantilever simple-shear/pure-shear model of continental lithosphere extension: applications to the Jeanne d’Arc Basin, Grand Banks and Viking Graben, North Sea. In:
Roberts
,
A.M.
Yielding
,
G.
Freeman
,
B.
(eds)
The Geometry of Normal Faults
 .
Geological Society, London, Special Publications
,
56
,
41
60
, https://doi.org/10.1144/GSL.SP.1991.056.01.04
33.
Kusznir
,
N.J.
Roberts
,
A.M.
Morley
,
C.K.
1995
. Forward and reverse modelling of rift basin formation. In:
Lambiase
,
J.
(ed.)
Hydrocarbon Habitat in Rift Basins
 .
Geological Society, London, Special Publications
,
80
,
33
56
.
34.
Mammadov
,
P.
1992
.
Seismostratigraphical investigations of geological structure of sedimentary cover of South Caspian superdepression and perspectives of oil–gas productivity
 . Doctoral thesis,
National Academy of Sciences
,
Baku, Azerbaijan (in Russian)
.
35.
Mammadov
,
P.Z.
2006
.
The peculiarities of the South Caspian Depression earth crust on the new geophysical data
.
Azerbaijan National Academy of Sciences, Proceedings of the Sciences of Earth
 ,
3
,
36
49
(in Russian).
36.
Mammadov
,
P.Z.
2008
.
The subsidence evolution of the South Caspian Basin
. In:
Caspian and Black Sea Geosciences Conference 2008. Proceedings of a Meeting held 6–8 October 2008, Baku, Azerbaijan
.
European Association of Geoscientists and Engineers (EAGE)
,
Houten, The Netherlands
,
A11
.
37.
Mangino
,
S.
Priestley
,
K.
1998
.
The crustal structure of the southern Caspian region
.
Geophysical Journal International
 ,
133
,
630
648
.
38.
McKenzie
,
D.P.
1978
.
Some remarks on the development of sedimentary basins
.
Earth Planetary Science Letters
 ,
40
,
25
32
.
39.
Otto
,
S.C.
1997
.
Mesozoic–Cenozoic history of deformation and petroleum systems in sedimentary basins of central Asia: implications of collisions on the Eurasian margin
.
Petroleum Geoscience
 ,
3
,
327
341
, https://doi.org/10.1144/petgeo.3.4.327
40.
Parsons
,
B.
Sclater
,
J.G.
1977
.
An analysis of the variation of ocean floor bathymetry and heat flow with age
.
Journal of Geophysical Research
 ,
82
,
803
827
.
41.
Piip
,
V.B.
Rodnikov
,
A.G.
Buvaev
,
N.A.
2012
.
The Deep Structure of the Lithosphere along the Caucasus–South Caspian Basin–Apsheron Threshold–Middle–Caspian Basin–Turan Plate Seismic Profile
.
Moscow University Geology Bulletin
 ,
67
,
125
132
.
42.
Popov
,
S.V.
Shcherba
,
I.G.
Ilyina
,
L.B.
Nevesskaya
,
L.A.
Paramonova
,
N.P.
Khondkarian
,
S.O.
Magyar
,
I.
2006
.
Late Miocene to Pliocene palaeogeography of the Paratethys and its relation to the Mediterranean
.
Palaeogeography, Palaeoclimatology, Palaeoecology
 ,
238
,
91
106
.
43.
Reilinger
,
R.
McClusky
,
S.
et al
2006
.
GPS constraints on continental deformation in the Africa-Arabia-Eurasia continental collision zone and implications for the dynamics of plate interactions
.
Journal of Geophysical Research
 ,
111
,
1
26
.
44.
Roberts
,
A.M.
Kusznir
,
N.J.
Yielding
,
G.
Styles
,
P.
1998
.
2D flexural backstripping of extensional basins; the need for a sideways glance
.
Petroleum Geoscience
 ,
4
,
327
338
, https://doi.org/10.1144/petgeo.4.4.327
45.
Richardson
,
S.E.J.
Davies
,
R.J.
Allen
,
M.B.
Grant
,
S.F.
2011
.
Structure and evolution of mass transport deposits in the South Caspian Basin, Azerbaijan
.
Basin Research
 ,
23
,
702
719
.
46.
Sclater
,
J.G.
Christie
,
P.A.F.
1980
.
Continental stretching: an explanation of the postmid-Cretaceous subsidence of the central North Sea basin
.
Journal Geophysical Research
 ,
85
,
3711
3739
.
47.
Sengör
,
A.M.C.
1990
. A new model for the late Paleozoic–Mesozoic tectonic evolution of Iran and implications for Oman. In:
Robertson
,
A.H.F.
Searle
,
M.P.
Ries
,
A.C.
(eds)
The Geology and Tectonics of the Oman Region
 .
Geological Society, London, Special Publications
,
49
,
797
831
, https://doi.org/10.1144/GSL.SP.1992.049.01.49
48.
Shikhalibeyli
,
E.Sh.
Abdullayev
,
R.N.
Ali-Zade
,
A.A.
1988
.
Geological results from the Saatly Superdeep Drillhole
.
International Geology Review
 ,
12
,
1277
1277
.
49.
USSR Ministry of Geology
1990
.
Gravity Map of the USSR, Scale 1:2 500 000
 .
USSR Ministry of Geology
,
Moscow
.
50.
Van Baak
,
C.G.C.
2010
.
Glacio-Marine transgressions of the Early and Middle Pleistocene Caspian Basin, Azerbaijan
 . MSc thesis,
Paleomagnetic Laboratory ‘Fort Hoofddijk’, Utrecht University, Faculty of Geosciences
.
51.
White
,
N.
1994
.
An inverse method for determining lithospheric strain rate variation on geological timescales
.
Earth and Planetary Science Letters
 ,
122
,
351
371
.
52.
Winterbourne
,
J.
Crosby
,
A.
White
,
N.
2009
.
Depth, age and dynamic topography of oceanic lithosphere beneath heavily sedimented Atlantic margins
.
Earth and Planetary Science Letters
 ,
287
,
137
151
.
53.
Zonenshain
,
L.P.
Le Pichon
,
X.
1986
.
Deep basins of the Black Sea and Caspian Sea as remnants of Mesozoic back arc basins
.
Tectonophysics
 ,
123
,
181
211
.
Close Modal

or Create an Account

Close Modal
Close Modal