From Oblique Thrust to Strike-Slip Fault: Progressive Stages of an Accretionary Wedge Development Open Access

The observed convergence of tectonic plates is generally not strictly perpendicular to the boundary between the two plates but rather oblique at an angle. This phenomenon has been described in most collisional orogens globally [1-10]. Thrusting processes have been thoroughly studied, and their initiation, geometry, and early-to-middle progression stages are well understood [11-13]. However, the processes associated with the latter stages at the outer part of the wedge remain uncertain. It is presumed that in the final stages, due to stress partitioning, the wedge becomes dissected, and a strike-slip fault forms at the rear part of the wedge [14-17]. During these final stages of oblique thrusting, the external orogenic complexes rotate laterally [18-20], which is associated with changes in the stress field [21, 22]. The existing partitioning models do not include and explain the mechanism of rotation of the external parts of thrusted complexes. Both partitioning and rotation have commonly been perceived and analyzed separately. This raises several questions such as: (1) Is the process of stress-partitioning causable of rotating the external orogenic complexes? (2) Are strike-slip faults associated with stress-partitioning of any significance? (3) If yes, what role does the geometry of such faults play?

An excellent location to study and understand both processes is the western margin of the Outer Western Carpathians (OWC), a typical example of an obliquely thrusted accretionary wedge. Additionally, there are several beneficial aspects, including: (1) a large amount of accessible subsurface data from the previous hydrocarbon exploration, enabling us to reconstruct and study the overall accretionary wedge geometry; (2) the presence of a prominent limestone marker horizon, which reveals the structure preserved at the frontal part of the orogen and provides evidence of brittle deformation from earlier processes; and (3) the relatively young age of these processes (Late Miocene).

We investigated the brittle structures by means of 1 m resolution LiDAR digital terrain models (DTMs), orthophotos, and compass fault-slip data measurements combined with electrical resistivity tomography (ERT) surveying and 2D seismic reflection profiles. The data and the paleostress analysis revealed a complex history and progressive evolution of this region, which can be divided into three main stages: (i) South to SE dipping thrust faults indicate the oldest thrusting of the Carpathian accretion wedge over the Bohemian Massif; (ii) NW-SE-oriented strike-slip faults uncover strain partitioning and lateral escape of the orogen marginal parts to W-NW; and (iii) the youngest N-S trending left-lateral faults evidence the change in compression direction and transition from thrusting to a strike-slip regime.

The studied westernmost margin of the OWC is part of the Alpine-Carpathian range formed during the Alpine-Himalayan orogenesis. This segment of the Carpathians emerged from the collision between the European platform and the ALCAPA microplates [23, 24] in the form of an accretionary prism with the thin-skinned nappe sheet structure (Figure 1). The flexion of the orogenic front of the OWC creates an oblique collision setting. The accretionary wedge is primarily composed of flysch sediments and was thrust over the European platform, while the Carpathian foredeep formed at its front. Concurrently with the oblique thrusting, the tectonic depression of the Vienna Basin formed within the accretionary wedge as a piggy-back basin, later evolving into a pull-apart basin [25].

The autochthonous European basement, represented by the Bohemian Massif underlying the Carpathian Fold-Thrust Belt, consists of Proterozoic to Lower Paleozoic rocks that were metamorphosed, folded, and thrusted mainly during the Variscan orogeny [e.g. 26-28]. These rocks were later intruded by granitoid plutons and subsequently covered by the Permo-Carboniferous molasse basins [e.g. 29-31]. The part of the European basement where the study area is located was affected by Variscan orogeny. After the Upper Triassic, the basement is considered to be consolidated.

In the Lower Jurassic period, the consolidated basement was rifted by the opening of the Piemont Ocean (Figure 1(a)). During this process, extensional normal listric faults affected the basement. This extension was followed by the sedimentation of limestones on the top of the passive margin [32-34]. The extensional regime lasted until the Upper Cretaceous when the tectonic regime shifted from extension to compression.

Due to the interplay of the microplates detached from the African domain, the tectonic regime switched from extensional to compressional during the Upper Cretaceous to Paleogene (Figure 1(b)) [23]. Sediments were incorporated into nappes and a thick accretionary wedge (up to 7 km) developed. An accretionary wedge, also referred to as an accretionary prism [e.g. 35, 36] or a subduction complex [37], is a compressional fold-and-thrust belt composed of oceanic and/or continentally derived trench-floor sediments located between the subducting and overriding plates [38]. The thrust faults exhibit higher-angle dips near the surface and lower-angle dips in the deeper subsurface [11]. The nappes stacked on top of each other within the accretionary wedge.

This accretionary wedge is mainly composed of interbedded claystones and sandstones, representing typical flysch sequences from the Upper Cretaceous to Paleogene. During thrusting, the autochthonous Jurassic limestone bodies were detached from the basement and incorporated as tectonic slices or sheets into the accretionary wedge. These incorporated limestone beds provide excellent information about the flat-ramp-flat geometry of the thrust belt [39].

The boundary between the European platform and the accretionary wedge in the study area was oriented in a SW-NE direction. The wedge generally moved toward the NE, making the direction of thrusting roughly parallel to this orientation. However, some markers indicate that the most external parts of the wedge were thrust toward the N-NW. This is the result of counter-clockwise rotation of orogenic blocks at the peripheral part of the wedge [20-22]. Although this rotation has been previously observed, its mechanism has not yet been explained.

According to several authors [40-42], the thrusting phase in the area of the OWC culminated in the Middle to Upper Miocene, as indicated by the youngest sediments incorporated within the accretionary wedge [22, 43, 44]. However, a detailed analysis of foredeep strata within the accretionary wedge body revealed that wedge propagation was not synchronous along the entire front [43]. Initially, tectonic activity of the wedge began in the south and later shifted toward the N-NE. While thrusting activity moved toward the N-NE, a strike-slip regime dominated in the south (in Austria) [45, 46] at approximately 18 my BP, while in the north (in Poland) it continued until about 15-12 my BP (as suggested by [21]).

The sedimentary depression in front of the accretionary wedge, known as the Carpathian Foredeep, continuously migrated to its present position during the Neogene, with its sedimentary infill gradually incorporating into the wedge [42, 45]. The undeformed sediments in the foreland of the current Carpathian Fold-Thrust Belt represent the modern Carpathian foredeep. Concurrently with the thrusting stage, a new tectonic basin known as the Vienna Basin formed within the accretionary wedge. Originating as a piggy-back basin in the Lower Miocene, it later transformed into a pull-apart basin during subsequent compressional stages due to the increasingly dominant sinistral transtensional regime in the western part of the Carpathian orogene. Therefore, the strike-slip faults responsible for forming this pull-apart basin are characterized by NNE-oriented extensional duplexes with overlapping NE-striking sinistral faults [47, 48].

The structure of the accretionary wedge and its brittle deformation was analyzed through geological field investigations, geomorphological methods, geophysical measurements, archival data, traditional tectonic orientation analysis, and supplemented with paleostress investigations.

The structural analysis focused on identifying brittle deformation features such as fault planes, striae, cleavage, and stylolites. These features were particularly examined in the Jurassic limestone, which preserved and revealed these structures effectively, making it a useful contrasting marker horizon. The data were subsequently visualized using Spheristat software [49] on the lower hemisphere using the Lambert projection. Stylolites were analyzed mainly with a focus on the measurement of the compression axis (Figure 2).

The paleostress analysis of shear faults was conducted using the MARK2010 software [50]. The fault-slip data included the orientation of fault surfaces, striae, and the shear sense recognized through features such as accretionary steps, Riedel shears, P-shears, and slickolites. MARK2010 allows for the analysis of heterogeneous fault-slip data, particularly where striae were generated during different stress phases. The relative temporal relationship of these phases was identified when two or more sets of striae were observed on a single fault surface (i.e. reactivated faults). The applied paleostress analysis is an original method described by [50] using the freeware computer program MARK2010. This method builds upon the stress-space concept introduced by [51], adapted for fault-slip data as the C-line by [52], and further developed into a 9D stress-space by [53]. Homogeneous groups of fault-slip data were segregated using multiple inverse methods [54], allowing for analysis of heterogeneous datasets comprising faults formed under different stress conditions. The resulting stress state in the 9D sigma-space and the directions of principal stresses in geographic coordinates were determined using the generalized Watson density function [55]. The directions of principal stresses—maximum compressional σ₁, intermediate σ₂, and relative extension σ₃—were graphically represented by density maxima projected onto the lower hemisphere. Additionally, distinct points representing the respective principal stresses obtained from the 9D analysis were illustrated.

Large-scale topographic features were investigated using LiDAR-derived DTMs and orthophotos. General overview of LiDAR interpretation in Figure 3 with details in Figure 4. The contrast between competent Jurassic limestone, which preserves brittle fractures and faults, and soft Paleogene sandstone and claystone enabled effective morphostructural mapping of these fragile landforms on LiDAR DTMs over extensive areas [56]. Our data, with a resolution of 1 m, included publicly available airborne LiDAR DTMs provided by the State Administration of Land Surveying and Cadastre, as well as data acquired by TOPGIS. These datasets were visualized and further analyzed in the GIS environment to create slope maps, slope aspect maps, and shaded relief maps. Illumination was applied from the northwest (NW) and west (W), which are perpendicular to the major linear features of interest.

Aerial orthophotos publicly available in [57] from different years and annual seasons were examined. The best performance of the fault-related structures was achieved from the photos of the summer 2022, where the faults were well detectable as they were more water-saturated and contained more soil compared to the dry surrounding rock surface. These stripes then provided a suitable environment for vegetation to grow, and thus, the fault zones could be then easily recognized and highlighted.

The ERT is a near-surface geophysical method, which can determine geological bodies with different electric resistivity resulting from different porosity, water saturation, conductivity of pore fluid, and/or clay content [e.g. 58, 59]. ERT was used for visualization of the shallowest parts of the most distinct faults to understand their geometry and lithological context. We used the Ares equipment in a “Schlumberger” array at usually 0.36–1.5 km long profiles with 3 and 5 m electrode separation. The profiles were conducted perpendicular to the investigated fault systems (for location see Figure 3).

A 2D exploration deep-seismic survey archive was re-interpreted to characterize deep fault architecture and to analyze the relationship between the strike-slip faulting regime and the accretionary wedge geometry. The re-interpretation of 2D deep-seismic profiles was performed using Petrel software, integrating borehole logs obtained from the Czech Geological Survey. These data comprise 25 crosslines and 15 inlines, all of which are migrated time sections with a maximum recording time of 6-second two-way travel time (TWT), providing high-quality information down to approximately 4-second TWT. The 2D grid is complemented by a 3D seismic cube provided by the company MND, a.s. All seismic data were processed using the seismic interpretation package within Schlumberger’s Petrel software. Eight boreholes from the Czech Geological Survey archives (see Figure S1 in the supplementary materials) were incorporated into the survey to identify and map the key stratigraphic horizons.

The LiDAR-derived DTM revealed that the Jurassic limestone beds outcrop as hills and klippes (meaning a tectonic slice of hard rock surrounded by soft rocks usually used in the Carpathian geological terminology) protruding from a relatively flat and smooth topography. These competent limestone formations are surrounded by softer claystones, siltstones, and sandstones of the OWC Flysch and/or Western Carpathian Foreland Basin. The limestone forms a chain of elevations and outcrops along the margin of the Western Carpathian nappes in a NNE-SSW direction (see Figure 3). Three main fault types have shaped the current morphology of these elevations, identifiable on the LiDAR DTM: thrust faults, NW-SE strike-slip faults, and N-S to NNE-SSW strike-slip fault zone (see Figure 3). Thrust faults intersect with NW-SE strike-slip faults, while the N-S to NNE-SSW strike-slip fault zone intersects both of the aforementioned groups—this fault zone we call the Falkenstein-Mikulov fault zone.

The northernmost limestone elevation with identified fault structures is Děvín Hill (Site 1; see Figure 3 and detail in Figure 4(a)). Its western slope is sharply cut almost vertically by a major N-S striking fault, with several subsidiary faults running in NNW-SSE directions. Conversely, Obora Hill (Site 2 Figure 4(a)), located southwest of Site 1, is dissected by the major N-S striking fault on its southeastern flank. Sites 1 and 2 are separated by a 200 m wide depression filled with flysch sediments. Further south, approximately 1200 m from Sites 1 and 2, other fault-related structures and scarps can be observed. The hills of Stolová hora (Site 3a) and Sirotčí hrádek (Site 3b) further south are prominent klippes flanked by faults (Figure 4(b)). The klippe at Site 3a has a rhomboidal-shaped outcrop with corners pointing to the N, S, E, and W. Another group of smaller rhomboidal-shaped limestone tectonic slices located in the central part of the study area are Kočičí skála (Site 4), Turold (Site 5, Figure 4(c)), and outcrops Zámecký vrch and Čertova skála (Site 6, Figure 4(d)). The lenticular geometry results from the oblique intersection of thrusted limestone bodies and a system of the strike-slip fault planes (Figures 4(c) and 4(d)). These klippes are forming a discontinuous chain with a total length of 4200 m. These klippes are separated by weaker Paleogene clayey cataclastic rock. The klippes are oriented in a N-S to NNE-SSW direction. A similar orientation is observed at Svatý kopeček (Site 7), where the hill is segmented into smaller blocks along NNE-SSW striking branches of the major fault, indicating a left-lateral component (Figure 4(d)). The offset of each block increases westward. Site 7 is separated from Site 6 by an approximately 300 m wide depression filled with flysch rocks. The southernmost limestone hill in the Czech Republic is Šibeniční vrch (Site 9), which extends in a NNE-SSW direction and is bordered by faults on both sides. Detailed examination of the LiDAR DTM revealed that the chain of limestone klippes indicating the fault position exhibits a visible curvature. The strike of the 10 km long fault changes by approximately 12° (Figure 3).

Analysis of orthophotos revealed several distinctive markers of brittle deformation in Mesozoic limestone (see Figure 5). At Site 2, vegetation stripes were observed striking in a N-S to NNE-SSW direction (Figure 5(a)). Similar vegetation stripe patterns were also recognized at Site 3. In the southeastern part of Site 1 (Figures 5(b) and 5(c)) and at Site 7, the vegetation stripes appeared oblique to these N-S ruptures, forming angles of 15° and 75° toward the main fault. Similarly, oblique ruptures filled with vegetation were observed on a smaller scale at Site 1 in the form of cleaved limestone, which displayed a similar strike with an axis subtending angles of 15° and 75° (Figure 5(d)). Vegetation stripes are structurally controlled, which was documented in the field (they represent lines on the top of the bedding surface). Another form of brittle deformation in Mesozoic rocks was evident at Site 1, the southern part of Site 2 (Figure 6(a)), and Site 4, where the limestone was noticeably more crushed compared to the massive limestone at other locations.

Structural markers such as stylolites, striae, or Riedel shears were identified on most of the massive limestone outcrops. Stylolites were particularly prominent in the quarry at Site 8 (Figures 6(c) and 6(d)), in Austria (Area 3, for location see Figure 1(d)), and at Site 2. Striae indicating movement direction were measured at most sites, including curved striae at Site 8 (Figure 6(e)), suggesting changes in stress, as well as subhorizontal striae, for instance at Site 1 (Figure 6(b)).

Near-surface geophysical surveying provided insights into the composition of the shallowest parts of the OWC nappes and limestone klippes. The ERT profile DEV3 can be divided into two sections (Figure 7(a)). The eastern part exhibits low-resistivity rocks (claystone, shale), while the western part consists of materials with the high resistivity (carbonates). These two sections are separated by a sharp and steep fault that dips toward the East. The second ERT profile, TUR3, shows high-resistivity rocks in the central part (Figure 7(b)). This mound of high resistivity is surrounded by material with different properties, delineated by low-resistivity areas (interpreted as faults) to the east, west, and beneath. Similarly, the southernmost profile at Site 9 exhibits a pattern similar to profile TUR3 (Figure 7(c)), with a smaller mound of Mesozoic limestone at the central part isolated by surrounding low-resistivity material to the east and west. According to the ERT survey at TUR3, the limestone is intensively karstified as evidenced by extremely high resistivity within its body.

The primary aim of the deep-seismic interpretation was to understand the structures and relationships within the oblique wedge, achieved through the integration of borehole data and reflection characteristics. Three main stratigraphic units were consistently identified across all seismic sections.

The reflections from the autochthonous basement depict mostly parallel and continuous reflectors with medium to high amplitudes, typically with a thickness expressed in TWT of 1–1.5 seconds. These reflections exhibit changing polarity and generally indicate a gentle relief (see Figure 8). They correspond to the basement, which consists of two main units: (1) the Bohemian Massif, primarily composed of granodiorites, and (2) Mesozoic autochthonous sedimentary rocks, predominantly limestone, which discordantly overlie the Bohemian Massif rocks. The reflection packages of the basement are influenced by the low-angle normal listric faults, which create horst-and-graben or half-graben structures. The depth of the basement is constrained by several wells within the investigated area.

The reflections of the Carpathian accretionary wedge were not continuous but rather chaotic with low to medium amplitudes and total thickness of 1–1.5 seconds. The major structural element within the accretionary wedge was the incorporated limestone body with a flat-ramp-flat geometry [39]. This structure is highlighted by high amplitude reflections which are indicating the transition from the Paleogene claystones and sandstones into limestone. The whole accretionary wedge is affected by an extensive system of thrust faults. The basal detachment zone of these faults (main decollement) is located in the depth of approximately 1.5 seconds. Thrust faults are mostly continuous. The only exception is a small wedge of strata pinched in between allochthonous Mesozoic and Carpathian foredeep. In this zone, the thrust faults are offset by another system of faults with vague indication of a flower structure (Figure 8). It has a width of about 200–400 m, including several splay faults. The general geometry of the flower structure is best depicted in the section xline_1370 crossing Sites 1 and 2 (Figure 8). This fault zone is offsetting the strata of the accretionary wedge as well as the oldest strata of the Carpathian foredeep. At the same time, the autochthonous basement is not affected by this fault zone. The intersection of these faults plotted on the LiDAR DTM is in line with the Falkenstein-Mikulov fault zone. Two additional xlines 1310 and 1210 with more detailed zoom into the fault zone are present in supplementary material (Figures S2 and S3).

Paleostress analysis was conducted at nine representative sites in the Czech Republic (locations in Figure 3) and two areas in Austria (locations shown in Figure 1(d)). At these sites, the fault planes and associated striae and/or compressional axes of stylolites were identified and measured using a geological compass. Stylolites were identified at Sites 3, 7, 8, Area 2, and Area 3. Majority of stylolites were measured at Site 8 and in Area 2 and are plotted in Figure 9. Archive data were taken into account while conducting analysis, mostly in areas with a limited number of measured faults [60], but these were not included in the analysis itself. In total, over 400 faults were analyzed. The results are graphically presented in Figure 9, depicting stress phases for each principal stress (σ₁, σ₂, σ₃) projected on the lower hemisphere. The orientations of the measured stylolite axes were plotted onto a stereographic projection (Figure 9), following methods outlined by [61]. The paleostress analysis identified four tectonic phases, labeled as D1–D4. Different tectonic phases were distinguished on those outcrops where the relationship between two sets of striae was assessed (see Figures 6(e) and 2(c)). These phases are numbered according to their sequence in Chapter 5—Discussion.

Northern part of the study area (Sites 1 and 3) is dominated by tectonic Phases D1 and D3, i.e. S to SE -dipping thrust faults and N-S to NNE-SSW strike-slip faults. Contrary to the southern part, which revealed mostly Phases D2 and D4, i.e. NW-SE strike-slip faults and normal faults related to the extension regime. Phases D1 and D2 are mostly in concurrence with previous authors [60]. Slight deviation is visible while comparing Phase D1, where our analysis shows NW as well as NE direction of propagation, whereas archive analysis emphasized mainly NW direction. Phase D3 has not been recognized in the previous studies.

Compressional axes of stylolites were measured in two limestone quarries: one at Site 8 (location shown in Figure 3) and another in area 2 in Austria (location shown in Figure 1(d)). The results revealed three main maximum compressional axes. As stylolites indicate maximal compression directly, we know principal stresses, σ₁: plunging horizontal and trending NNE-SSW, NW-SE, respectively W-E, which is in agreement with D1, D2, and D3, respectively.

The structural pattern obtained from stylolites might be extended and supplemented by fault-slip data. Phase D1 exhibits a horizontal σ₁-oriented NNW-SSE to N-S, with a vertical minimum principal stress σ₃. In Phase D2, σ₁ is horizontal to subhorizontal striking W-E, while σ₃ is horizontal striking N-S. Phase D3 shows a horizontal σ₁ plunging to NW, with a subhorizontal minimum principal stress σ₃. Apart from these well-defined phases, one more phase was recognized, not in concurrence with stylolite compression axes, i.e. D4. This phase is characterized by subvertical sigma 1 and the N-S orientation of sigma 3.

Investigating the brittle structures and geometry of the Jurassic limestone markers has deepened our understanding of the late evolutionary stages of the accretionary wedge. We focused extensively on the faults that played a crucial role in shaping these processes, with the particular attention given to the Falkenstein-Mikulov fault, which stands out as a key structural element.

Based on structural, geomorphological, paleostress, and geophysical data, we identify the Falkenstein-Mikulov fault as a significant fault zone striking primarily in NNE-SSW to N-S direction. This fault zone intersects the boundaries of the majority of Jurassic limestone nappe outliers. The fault direction was primarily determined through LiDAR DTM and orthophoto analyses, facilitated by distinct morphological features including (1) sharp, subvertical terminations of limestone elevations at Sites 1, 3a, and 3b (Figures 4(a) and 4(b)); (2) lenticular NNE-SSW to N-S prolongation of several prominent limestone outcrops (Figures 3, 4(c), and 4(d)); and (3) fault parallel vegetation stripes, rugged relief, and limestone cleavage elongated in NNE-SSW to N-S direction (Figures 5(a) and 5(c)).

The fault zone has a visibly arcuated shape (Figure 3). Previous interpretations of similar arcuate faults in the OWC considered them as sequences of interconnected straight faults kinked at various angles [62], depicted as such on geological maps [63, Figure 1c]. Based on our investigation, we propose that this fault represents rather a single continuous arcuate structure.

Apart from the curved strike of the fault, the direction of limestone ridges in close proximity to the fault also changes at a similar angle (compare Sites 1 and 7, Figure 3). This observation indicates that these limestone bodies rotated as they moved along the fault. Our proposed model suggests that along such arcuate faults, the external parts of accretionary wedges may undergo rotation. This mechanism provides a plausible explanation for the rotational patterns noted by previous authors.

The position of the Falkenstein-Mikulov fault determined on-surface was in alignment with available and newly acquired subsurface geophysical data where the fault was tracked along the aforementioned limestone elevations (1) in ERT profiles (DEV3, TUR3, and SIB1 in Figure 7) where the fault appears as a distinct subvertical boundary between high-resistivity limestone bodies and low-resistivity flysch formations, and (2) in the 2D deep-seismic profile xline_1370, where the fault is part of a broader zone characterized by subvertical fault structures (Figure 8). Both subsurface imaging methods confirmed the subvertical continuation of the fault geometry as initially suggested by features observed on LiDAR DTM.

In terms of kinematics, the Falkenstein-Mikulov fault is identified as a significant left-lateral strike-slip fault. This conclusion is supported by several observations such as (1) displacement of limestone blocks along the fault system in the western part of Site 7 (Figure 4(d)); (2) minor displacements observed along parallel fault branches at the foothill of Site 7, where sinistral kinematics are identified on fault surfaces (Figure 6(b)); and (3) presence of oblique ruptures (Riedel shears) at angles of 15° and 75° toward the main fault zone, observed on the LiDAR DTM and orthophotos at several areas (Sites 1, 5, 7, Figures 3, 5(b), and 5(c)).

Sequence of en-échelon limestone lenses suggests that the Falkenstein-Mikulov fault is not a single fault but rather a series of several parallel faults forming a broader zone. This observation is consistent with subsurface data, particularly at the ERT geophysical profile TUR3 and SIB1 (Figures 7(b) and 7(c)) where the limestone body is sharply and subvertically dissected from both sides by a series of subvertical faults. The existence of this broad zone explains the repetitive pattern visible on the DTM, where two limestone blocks form elevations separated by N-S oriented depressions, each 100—300 m wide and filled with Paleogene flysch complex (between Sites 1 and 2, and between Sites 6 and 7 in Figures 3, 4(a) and 4(d)). This pattern is also evident subsurface in the seismic profile xline_1370, where the width of the fault zone is approximately 300—400 m.

Thrust faults observed in the LiDAR DTM exhibit offsets caused by oblique NW-SE striking faults, particularly noticeable at Sites 2, 3, and 7 (Figure 4). These faults are likely of the same age as the thrusts or younger, indicating their formation during the progressive progradation and locking of thrusts in the later stages of orogenesis [2]. As thrust blocks near the margin begin to lock, stress and deformation are transferred to strike-slip faults until the wedge encounters another locking phase. This younger activity is associated with changes in stress orientation during wedge development, a conclusion supported by paleostress analysis.

The elongated and dissected frontal thrust of the Carpathian accretionary wedge was subsequently intersected by the Falkenstein-Mikulov fault zone. Analysis of the deep-seismic profiles (Figure 8) revealed that the Falkenstein-Mikulov fault influences the reflections of the Carpathian flysch, allochthonous Mesozoic rocks, and even some of the oldest sediments of the Carpathian foredeep incorporated into the accretionary wedge. However, the fundamental strata of the autochthonous Mesozoic rocks and Bohemian Massif appear largely unaffected by the fault in most areas, suggesting that the fault likely formed after the thrusting phase, and it is linked with wedge propagation. Furthermore, the NW-SE strike-slip faults linked to wedge locking and propagation are also cut by the fault (Figures 3 and 4). This observation indicates that the Falkenstein-Mikulov strike-slip fault zone postdates these faults.

Combining geometric and kinematic evidence along with the relative ages of movements and paleostress analysis, we decipher the tectonic evolution of the OWC accretionary wedge. We reconstruct how thrusting transitioned gradually from a predominant reverse component to a dominant strike-slip component (Figure 10). Based on the striation relationships on fault surfaces, we delineated four stages pivotal in the consolidation of the accretionary wedge, as revealed by paleostress analysis:

The first stage involved the thrusting of the Carpathian accretionary wedge over the European platform (tectonic Phase D1, Figure 9). The orientation of maximum and minimum principal stresses suggests low-angle thrusting in the generally N–S direction. During this phase, Mesozoic limestone slabs were incorporated into the accretionary wedge and folded, forming a flat-ramp-flat geometry [39]. Stage 1 can be identified at the outcrops as thrust faults (Figures 3 and 4)

The second stage is characterized by the transition of the N-S compression to strike-slip regime with rather ENE-WSW to E-W σ₁ and the onset of NW-SE strike-slip faults (tectonic Phase D2, Figure 9). It corresponds to the locking of nappes during later thrusting stages and associated stress rotation. Stage 2 is revealed by the presence of NW-SE strike-slip faults on LiDAR DTM (Figures 3 and 4).

Later on, the compressional σ₁ shifted to NW-SE direction while the σ₂ was oriented subvertically during the tectonic Phase D3. It was during this phase that the Falkenstein-Mikulov strike-slip fault zone originated, as evidenced by our data (Figure 9). N-S strike-slip faults developed prominently under this regime, marking the final major phase of the Carpathian collision with the European platform. Stage 3 can be identified on LiDAR as curved N-S strike-slip faults (Figure 3).

Phase D4, characterized by subvertical maximum principal stress and N-S extension, is likely associated with the opening of the Neogene Vienna Basin and/or with the gravitational collapse of thrusted material possibly coinciding with significant landslide events along the slopes.

We can conclude that the tectonic activity initially characterized by thrust faults (as frontal ramps) in Phase D1 transitioned during progressive thrusting to subsequent movement along lateral ramps in approximately perpendicular orientation (Phase D2), which is in concurrence with previous interpretations [39, 60]. Finally, the system evolved into cross-cutting strike-slip faults more-or-less parallel to the direction of oblique thrusting in Phase D3, which is a new observation (see Figure 9). Contrary to previous authors, we also describe faults related to Phase D2 rather as a single continuous arcuate structure than sequences of interconnected straight faults kinked at various angles [62].

As mentioned at the geological settings, the propagation of the accretionary wedge and the compressional regime was gradually shifting towards N-NE direction, i.e. from SSW-SW to NNE [40-42, 45, 46]. The thrusting activity within the wedge translated in the same direction and, as we documented at the study area (Figure 10), was followed by motion along the lateral ramps and strike-slip faults. If we acknowledge such complex evolution of oblique wedge, it becomes apparent that these different tectonic regimes must have been synchronous in various parts of the wedge (Figure 11). This interpretation aligns well with the formation of the Vienna pull-apart basin at OWC, originated by the interplay of strike-slip faults in the SE portion of the wedge, while the thrusting process persisted and continued in the NE.

Such synchronous tectonic activity is performed by movement along faults of different orientations. However, these faults must be continuous as the wedge is constantly propagating. Locally, the thrusting can alter its direction while the orogene complex rotates to a position oblique to almost perpendicular to the course of the main mountain range [18-20]—in the case of sinistral oblique thrusting anticlockwise rotation and vice versa. The precise mechanism of such behavior has not been fully understood yet. We propose that the described fault zone associated with change from transversal segmentation (Phase D2) to strike-slip partitioning (Phase D3) during the final stages of the accretionary wedge development provides a plausible explanation for this phenomenon. The movement along such curved strike-slip faults indisputably leads to the external rotation of the structures as demonstrated in this article.

This behavior described within the accretionary wedge detached from the basement is applicable to the majority of oblique-thrusted accretionary wedges globally. Within these wedges, thrust faults are associated with strike-slip faults parallel to thrusting direction. As these strike-slip faults curve back toward the hinterland, they remain parallel to the boundary yet oblique to the thrusting direction. This process ultimately causes the orogenic complexes to rotate externally.

All data used in the preparation of this contribution are included in the manuscript and supplementary materials. Individual fault measurements are available from the corresponding author upon request.

The authors declare that they have no conflicts of interest.

The research was supported by the international bi-lateral project “Earthquake triggered landslides in recently active and stabilized accretionary wedges,” supported by the Czech Science Foundation (GAČR 22-24206J) and the Taiwanese Ministry of Science and Technology (NTSC 111-2923-M-008-006-MY3), and by the conceptual development project RVO 67985891 at the Institute of Rock Structure & Mechanics, Czech Academy of Sciences.

The authors acknowledge Jiří Svoboda, Jiří Kmet and Jiří Kolařík for pointing out excellent field outcrops, Radomír Grygar for fruitful scientific discussions, several undergraduate students of the geology study program for excellent help during the field geophysical surveys, Kurt Decker and Kateřina Schoepfer for valuable insights regarding seismic interpretation.

Supplementary materials are provided in separate files, and they include: Figure S1—lithology of borehole data used for seismic interpretation; Figure S2—seismic section xline_1310; Figure S3—seismic section xline_1210.