Basin-scale salt flow and the evolution of salt structures in rift basins is mainly driven by sub- and supra-salt faulting and sedimentary loading. Crustal extension is often accompanied and followed by thermal subsidence leading to tilting of the graben flanks, which might induce an additional basinward-directed driver for salt tectonics. We designed a new experimental analogue apparatus capable of integrating the processes of sub-salt graben extension and tilting of the flanks, such that the overlapping effects on the deformation of a viscous substratum and the brittle overburden can be simulated. The presented experimental study was performed to demonstrate the main functionality of the experimental procedure and setup, demonstrating the main differences in structural evolution between conditions of pure extension, pure tilting, and extension combined with tilting. Digital image correlation of top-view stereoscopic images was applied to reveal the 3D displacement and strain patterns. The results of these experiments suggest that in salt basins affected by sub-salt extension and flank inclination, the salt flow and downward movement of overburden affects the entire flanks of the basin. Supra-salt extension occurring close to the graben centre is overprinted by the downward movement; i.e. the amount of extension is reduced or extensional faults zones are shortened. At the basin margins, thin-skinned extensional faults developed as a result of gravity gliding. A comparison with natural examples reveals that such fault zones can also be observed at the margins of many salt-bearing rift basins indicating that gravity gliding played a role in these basins.
Salt layers in sedimentary basins play a key role in the structural and sedimentary evolution. As the bulk mechanical behaviour of rock salt on a geological timescale is that of a viscous fluid, it is able to flow in response to an internal pressure gradient and external shear forces
Driving processes of salt flow are well explained for salt-bearing passive margin basins, e.g. the Lower Congo Basin, Western Mediterranean, and the Santos Basin
In salt-bearing rift basins (SBRBs), these driving forces mostly act oppositely (Fig.
Analogue models dedicated to the opposing interplay between gliding and spreading in SBRBs show that salt flows downward if the depocentre above the downthrown block is underfilled and returns to upward-directed flow as soon as syn-kinematic sediments are accumulated in the depocentre
The key motivation of this work is to evaluate the influence of regional-scale gravity gliding on salt flow processes and the tectono-sedimentary evolution of SBRBs. We designed a new experimental apparatus for analogue modelling to integrate the simulation of crustal-scale extension and tilting of the graben flanks. Here, the basic concept of the apparatus and results of preliminary experiments are presented. Based on the techniques of digital image correlation, we compare displacement and strain patterns between different scenarios: (1) only extension, (2) only tilting of the flanks, (3) combined extension and tilting, and (4) extension, tilting, and syn-kinematic sedimentation. In order to demonstrate the effect of a viscous detachment, these scenarios were performed with and without a viscous substratum. The experimental results illustratively demonstrate the functionality of the apparatus and reveal the first main differences between rift basins with tilted flanks and those surrounded by flat flanks.
Literature-based data compilation of worldwide salt-bearing rift basins. The thickness of the salt layer refers to the estimated maximum thickness of the original evaporitic layer. The pre-kinematic layer is defined as the sedimentary layer accumulated after the salt deposition and before the first post-salt rifting phase.
The experimental setup and the procedure are inspired by a generalized natural salt-bearing rift basin (Fig.
Results of the DIC of experiment Eb (without viscous substratum) and Ev (with viscous substratum) in which only basal extension with a rate of 1 mm h
The experimental setup combines two main approaches of previous analogue modelling studies: (1) the simulation of upper crustal extension
The experimental surface is recorded by two stereoscopic CMOS (complementary metal-oxide semiconductor) sensor cameras (“Imager M-lite 12M camera”, 12-bit monochrome, 12 MP resolution) in top view to analyse the evolution of the surface displacements and strains
Results of the DIC of experiment Tb (without viscous substratum) and Tv (with viscous substratum) in which tilting was applied with a rate of 1 mm h
The viscous behaviour of salt
Frictional–plastic behaviour of overburden sediments is modelled by granular mixtures of quartz sand (bulk density
Results of the DIC of experiment ETb (without viscous substratum) and ETv (with viscous substratum). In both experiments extension and titling with rates of 1 mm h
In order to relate geometrical, dynamical, kinematical, and rheological model parameters to the natural prototype, we applied standard scaling procedures
Experiments and key parameters. In an experiment's name, E stands for extension, T for tilting, S for syn-kinematic sedimentation, b for (only) brittle, and v for viscous.
Since all parts of the apparatus can be controlled separately, different modes of basin evolution can be tested, such as pure basal extension, pure tilting of the flanks, and simultaneous basal extension and tilting (Table
Comparison of displacement patterns between experiment Ev (only extension) and ETv (extension and tilting).
In a first series of experiments (“brittle-only experiments”: Eb, Tb, ETb), no silicone layer was included in order to present benchmark experiments for the displacement patterns of the basal parts and to better illustrate the effects of the viscous layer in subsequent experiments. In experiments including a silicone layer (“brittle–viscous experiments”: Ev, Tv, ETv, ETSv), the silicone layer was covered by an even,
For most experiments, no syn-kinematic sand sedimentation was applied to demonstrate the deformation patterns resulting only from the specific mode. Only ETSv involves all relevant processes, which should be investigated with this apparatus (extension, tilting, syn-kinematic sedimentation). At intervals of 5 h, syn-kinematic sand was sieved on the model surface until the graben centre was completely filled up to the top of the sand cover above the footwall sides of the graben. Subsequently, a syn-kinematic sand layer with a uniform thickness of
Comparison of strain patterns between Ev (only extension) and ETv (extension and tilting).
In the following, we compare experiments, which were conducted without a viscous silicone layer (Eb, Tb, ETb) with those containing a viscous substratum (Ev, Tv, ETv). Then, we show differences in displacement and strain patterns between the brittle–viscous experiment with only extension (Ev) and the experiment with extension and tilting (ETv). Finally, the experiment including syn-kinematic sedimentation simultaneous to extension and tilting (ETSv) is presented.
When only basal extension was applied as boundary condition in experiment Eb (without viscous layer), both flanks were pushed apart laterally by the downward-moving central graben block (Fig.
Displacement and strain patterns of ETSv (extension
In experiment Tb (without silicone) in which only tilting of the flanks was applied, minor
In experiments ETb and ETv, graben extension and flank tilting were applied simultaneously (Fig.
Displacement and strain evolution of ETSv (extension
The effects of tilted flanks on the structural development is evaluated by comparing experiments Ev (only extension) and ETv (extension + tilting) (Fig.
In both experiments, the onset of inward movement above the graben flanks began towards the end of the syn-extensional phase and the regions affected by downward movement adjacent to the graben became wider after extension stopped (Fig.
Note that the incremental movement of the graben flanks is not fully continuous and symmetrical during basal extension (Fig.
The strain patterns are analysed by means of the cumulative normal strain in
In ETv, graben-edge fault zones were located closer to the basal faults during the syn-extensional phase (
In summary, syn-extensional flank tilting in ETv has the following effects on the overall deformation patterns: (1) the formation of basin-margin extensional fault zones, which gradually migrate downslope; (2) decreased amounts of extensional strain at the graben-edge fault zones; and (3) the formation of compressional zones (buckle folding) above the hanging wall graben block during the post-extensional phase.
In ETSv, extension and tilting of the basal parts was accompanied by regular intervals of syn-kinematic sand accumulation (Fig.
Graben-edge extension and compressional strain in the graben centre were almost completely suppressed in ETSv (max
Cross sections cut through the centre of ETSv in which syn-kinematic sand accumulation was applied during and after basal extension.
The evolutionary plots in Fig.
Cross sections cut through the final experiment reveal that the structures of basin-margin extensional fault zones in the sand cover are characterized by synthetic rollovers and symmetric grabens (Fig.
The presented experiments on gravity-driven deformation in salt-bearing rift basins (SBRBs) provide first insights into the influence of flank tilting on basin-wide deformation patterns in the brittle–viscous strata. When comparing brittle-only (Eb, Tb, ETb) and brittle–viscous experiments (Ev, Tv and ETv), it can be observed how the basal deformation is significantly overprinted if a viscous layer is present (Fig.
The comparison of the brittle–viscous experiments among each other demonstrates how local deformation structures due to a sub-salt graben are affected by regional-scale flow of the viscous material. When only basal extension is applied (Ev), deformation is restricted to the area in the vicinity of the basal graben (Fig.
When only tilting is applied (Tv), localized extension occurs at the basin margins, while diffuse shortening affects the cover in downslope regions (Fig.
Sketches summarizing the main results of this experimental study.
Due to simultaneous basal extension and flank tilting (ETv), the local (close to the graben) deformation is overprinted by a regional (entire basin scale) downward flow of the viscous material and gliding of the brittle cover (Fig.
During the post-extensional phase in ETv, downslope gliding continued across the entire flanks, which enhanced extension at the basin margins. The basin-margin extensional domain, which has started as 1–2 localized extensional zones, began to expand and migrated downslope (Fig.
Shortening structures above the hanging wall graben centre only occurred in ETv (Fig.
Examples of thin-skinned extensional structures recognized above inclined flanks at the margins of salt-bearing rift basins.
In the experiment with syn-kinematic sedimentation (ETSv), amounts of strain and displacement were considerably less than in the equivalent experiment without sedimentation (ETv) (Fig.
In ETSv, extensional strain at the basin margins and the graben edges occurred at one to two discrete fault zones, whereas it was distributed over several branched fault zones in ETv (without sedimentation) (Fig.
Experiment ETv (without sedimentation) and ETSv (including sedimentation) might represent end-member types of SBRBs of an underfilled basin (low sedimentation rates) and filled up basin (high sedimentation rates). Gravity gliding is favoured in underfilled rift basins or basins experiencing fast flank tilting relative to sedimentation rates. In such a case, the kinematic evolution and structural patterns are similar to passive margin basins in which basin tilting is the main driving factor for salt tectonics
In our experimental setup, basal extension is focused on two parallel faults, which is a simplification of a natural rift basin. Natural rift basins and the process of crustal extension include greater complexities than presented in our experimental setup. Rift basins usually consist of overlapping fault systems with normal faults, fault intersection, ramps, strike-slip segments, and intrabasinal ridges
Natural rift basins often possess asymmetrical half graben shapes characterized by a large-offset normal fault on one side and a smoothly tilted, slightly faulted flank on the other. In such half grabens, decoupled extension, probably caused by gravity gliding, can occur above the slightly tilted flank
A technical issue can be observed in Fig.
Boundary effects were observed along the sidewalls of the silicone basin in all experiments containing a viscous layer. En échelon shear patterns occurred close to the long edges of the silicone basin above the graben flanks (Fig.
In many SBRBs worldwide (Table
One of the most prominent examples of overlapping gravitationally and tectonically driven supra-salt extension can be found at the flanks of the northern Central Graben (central North Sea) (Fig.
The Sole Pit and Silverpit basins (south-western North Sea) are confined by extensional fault zones, which overlie a slightly faulted salt base (Fig.
The south Viking Graben and the Egersund Basin (northern North Sea) are salt-bearing half grabens in which the hanging wall graben centre was rotated during rifting
Extensional fault zones developed above gently dipping flanks of the Polish Basin from the Late Triassic to Jurassic, in particular along the NE flank located above the East European Craton (Fig.
Listric growth faults developed in the post-salt strata on the western flank of the Horn Graben (southern North Sea) (Fig.
These examples imply that there are three different scenarios of the interaction between gravity gliding and crustal extension: (1) gravity gliding took place during the main phase of rifting but no or minor shortening occurred in the basin centre, since here sub-salt extension was sufficient to accommodate and balance upslope, supra-salt extension (probably in the northern Central Graben); (2) gravity gliding is superimposed with minor rifting and caused basin-margin extension balanced by basin-centre shortening (e.g. in the Sole Pit Basin); and (3) no or only minor gravity gliding occurred, and the thin-skinned extensional structures were mainly tectonically driven (e.g. in the Horn Graben). In order to investigate these scenarios and their governing parameters and processes, we intend to perform a comprehensive analogue modelling study, modifying variables such as extension rate, tilting rate, layer thickness, and timing of tilting in relation to extension.
Our new experimental apparatus successfully reproduced the overlapping influence of tectonic extension and gravity gliding on salt flow and salt structure evolution in salt-bearing rift basins. The apparatus is suitable for simulating crustal extension within a symmetric graben structure and thermal subsidence, which is represented by vertical uplift of the graben flanks. Since basal extension and flank tilting can be controlled separately, the effects of both processes can be investigated separately or combined.
The preliminary experimental study shown here reveals primary structural and kinematic differences depending on the presence of a viscous layer in the sedimentary strata and the occurrence of flank tilting in rift basins. Whereas deformation in the brittle-only experiments is strictly coupled to the movement of the basal parts, widespread lateral displacement and distributed faulting affected the cover in brittle–viscous experiments.
In the brittle–viscous experiment in which only basal extension was applied, deformation is concentrated in regions close to the basal graben structure. Extensional fault zones develop in the footwall, while overburden above the basal normal faults slides into the graben. When the graben flanks become tilted simultaneously to graben extension, similar deformation patterns occur close to the graben. However, they are overprinted by gravity gliding inducing widespread basin-wide downward movement of the overburden into the central graben. Consequently, additional extension localizes at the upslope basin margins, which is enhanced after basal extension and tilting have stopped. This extension is accommodated in the graben region by reduced amounts of extension on the graben-edge fault zones and compressional strain above the graben bounding faults. If syn-kinematic sedimentation is included in the experimental procedure, lateral downward-directed displacement is significantly reduced especially during the post-extensional phase. Furthermore, strain in the margin-edge extensional zones is localized on fewer, more discrete faults.
Observed basin-margin extensional structures in the preliminary experiments resemble typical thin-skinned extensional structures occurring at the flanks of many salt-bearing rift basins with inclined sub-salt bases. Such diagnostic structures indicate that gravity gliding might play an important role in the post-salt structural development of such rift basins.
For the analogue model to be representative of the natural prototype, models have to be geometrically, dynamically, kinematically, and rheologically similar to their natural prototype
Geometric scaling ensures that corresponding ratios of lengths and angles are comparable between the model and nature. The choice of the geometrical scaling ratio is determined by the procedure for dynamical scaling
This means that 1 cm in the model represents 1 km in nature, and characteristic geometrical relationships, such as the thickness ratio between overburden and salt or the ratio between extensional displacement and flank uplift, can be scaled accordingly (Table
Fundamental principles of dynamical scaling require that trajectories and ratios of forces acting on the material have to be equal and that rheological behaviours of the involved materials are similar
Because deformation style and localization in the materials is mostly influenced by rheological parameters of the viscous layer and its brittle overburden, a characteristic measure for dynamical scaling is the brittle-to-viscous strength ratio or brittle–viscous coupling (BVC)
The frictional strength of both brittle sediments and experimental granular materials obeys the Mohr–Coulomb criterion, according to which the mean stress
Here,
In an extensional stress field, the maximal principal stress
With the presence of a pore fluid pressure, each pressure has to be corrected by the pore fluid pressure ratio (Hubbert–Rubey coefficient of fluid pressure
Assuming that pore fluid pressure is hydrostatic,
Combining Eqs. (
The force required for extensional failure is then derived by integrating
The large-scale viscous behaviour of a salt layer is characterized by a stress–strain rate relationship. Rock salt (halite), which usually constitutes the main volumetric proportion of a mobilized evaporitic succession
Here,
Tectonic extension induces a Couette shear flow (Turcotte and Schubert, 2014) at the base of the silicone on the flanks. Hence,
For comparison between nature and the model, we inserted a range of possible natural and experimental values (Table
Equation (
The ratio between brittle and viscous material properties also sets the scaling for the strain rates allowing us to relate extension, tilting, and sedimentation rates between the model and nature
Applying average values of viscosity, thickness, and density (Table
Tectonic stresses and gravity are the main driving factors in the geological scenarios modelled here. For scaling purposes we used the Argand number (
The buoyancy force
After integration, Eq. (
When inserting a range of possible values (Table
The average density of a salt layer is nearly constant with depth (
Integrating over the entire thickness of the cover
Because of this density increase, the bulk density of overburden sediments
Compaction of natural sediments cannot be directly simulated in analogue models using the common granular materials. However, it is not the density increase itself, but the increase in the lithostatic pressure at the base of the overburden that is the crucial parameter determining differential loading and the pressure gradient in the viscous layer. In order to simulate an appropriately scaled increase in the lithostatic pressure, we increased the density of the sand mixture (
The increase in density of each new sand layer has to be set in such a manner that the standardized bulk density resembles the exponential increase in the standardized bulk density in nature (Fig.
Physical parameters describing the deformation processes in the experiments. Lateral displacement due to basal extension and downward gliding of the overburden lead to shearing at the base and the top of viscous layer. Vertical displacement of the graben block and the flanks cause gravitational instability in the viscous layer. Syn-kinematic sedimentation, i.e. filling the subsided depocentres, imposes differential loading on top of the viscous layer. Governing parameters describing these processes are as follows:
Diagrams of the
Density configuration in the natural prototype and in the analogue model.
List of parameters and values used for the scaling procedure, i.e. to calculate BVC (Eq.
Data of the experiments presented here are provided in a data publication via the PANGAEA Data Publisher for Earth & Environmental Science
MW invented the concept of this modelling study, designed the experimental apparatus, performed the experiments, and composed the main parts of the paper. Furthermore, data processing and analysis was undertaken by MW. PZ advised the designing and construction of the apparatus and assisted during the experiments. Theoretical input for planing the project, the model design and the concept of the experiments came from FJK. PK supported the comparison of the experiments with the natural case study of the Polish Basin. FJK, PK, and PZ helped with finalizing the paper.
The authors declare that they have no conflict of interest.
Publisher’s note: Copernicus Publications remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
This paper was prepared as part of a bilateral research project between the Institute of Geophysics of the Czech Academy of Sciences and the Institute of Geological Sciences of the Polish Academy of Sciences (project no.: PAN-20-04). We acknowledge Oriol Ferrer, Gaëël Lymer, and Frank Zwaan for providing constructive and valuable reviews, which significantly improved the paper. Graham Hill is thanked for remarks on an early version of the paper. We thank Jiří Semerád for helping with designing and for constructing the experimental apparatus. We also thank GFZ Data Services for providing an openly accessible publication of data related to this study (Warsitzka et al., 2021b).
The research position of Michael Warsitzka is supported by the Czech Academy of Science (AVCR) in the framework of the “Programme to support prospective human resources – post Ph.D. candidates” (no.: L100121901)
This paper was edited by Susanne Buiter and reviewed by Oriol Ferrer, Gaël Lymer, and Frank Zwaan.