The sea cliff Kieler Ufer (KU) is located on the east coast of Jasmund, and it is easily accessible via the stairs at the creek Kieler Bach (54∘33.165′ N, 13∘40.594′ E), which crosses the fault boundary between the imbricates S13 and S14 (Fig. 1a). The abbreviation S used in conjunction with a number stands for “section”, and the section numbers S11 to S16 in the Kieler Ufer cliff profile have been adapted from the Pleistocene-stripe annotations given by Jaekel (1917). The horizontal length of the entire cross section is 1202 m. The highest point of the cliff profile is at the southernmost top of S15 (64 m a.s.l.).

The Kieler Ufer section represents a key area in the contractional fold-and-thrust system of Jasmund's southern structural sub-complex, which is a large imbricate fan with a number of individual thrust sheets and at least three duplex stacks (S01 to S23) (Gehrmann, 2018; Gehrmann and Harding, 2019). The large-scale glacitectonic folds and associated thrusts deforming both Upper Cretaceous (Maastrichtian) chalk and Pleistocene glacial deposits were formed in the Pomeranian W2 phase of the late Weichselian (see Kenzler and Hüneke, 2019). The detailed stratigraphy of the Maastrichtian chalk outlined by the flint-band numbers has been adapted from Steinich (1972). The Pleistocene sediments lie paraconformably on the chalk, parallel to this major lithostratigraphical boundary. The older glacial sediments (older than Pomeranian W2 phase) are subdivided into the tills/diamictons M1 and M2 that are separated and overlain by the units I1 and partly I2 with interbedded gravel, sand, and clay (e.g. Steinich, 1972; Panzig, 1995; Müller and Obst, 2006; Kenzler et al., 2015, 2016). These glacial deposits are cut by thrust faults at the trailing end of each imbricate in the SSW, below the large hanging-wall anticlines of the southerly adjoining thrust sheets. The youngest M3 sediment complex (Pomeranian W2 phase/Mecklenburg W3 phase) lies unconformably upon a major erosion surface, which truncates the underlying glacitectonised sequence (chalk, M1, I1, M2, I2) (e.g. Steinich, 1972; Panzig, 1995; Müller and Obst, 2006; Niedermeyer et al., 2010). At the top of the Kieler Ufer cliff the M3 complex is only exposed in small patches. The tectonic setting at the Kieler Ufer shows an imbricate fan with six major thrust sheets subjected to fault-bend folding (Fig. 1b). The six thrust sheets S11 to S16 strike SSW–NNE and are exposed by the N–S-trending sea cliff at high angle.

Cross-section restoration and balancing is an increasingly applied kinematic-analysis technique in glacial environments. This was well demonstrated for instance by Croot (1987), Pedersen (2005), and Benediktsson et al. (2010). The restoration of the Kieler Ufer section has been performed on the single imbricates of the cliff profile in several experiments using different algorithms and bedding geometries until the best-fit interpretation could be used for shortening calculations and further interpretation (software: Move and the supplementary module 2-D Kinematic Modelling). The tectonic model, which best explains the geometry of the six thrust sheets at the Kieler Ufer is shown in Fig. 1. Regarding the structural evolution of classic fold-and-thrust belts, the restoration worked backwards from the distal to the proximal deformation area (see Pedersen, 2005). Thus, the youngest sheet S16 (closest to the foreland of the southern structural sub-complex) was first restored, and the oldest sheet S11 (closest to the hinterland) was restored in the end. Within the scope of the restoration, the individual thrust sheets (S11 to S16) have been renamed (KU1, KU2, KU3d1, KU3d2, KU3d3, KU4) according to the chronology of re-deformation (see Figs. 1 and 5). The orientation values of beds and faults were adapted from the best-fit model of the Kieler Ufer section, and they have been compared with those given by Steinich (1972).

There are generally three individual thrust sheets: S16, S15, and S11 (KU1, KU2, and KU4). A duplex stack consisting of S14, S13, and S12 (KU3d1, KU3d2, and KU3d3) and the lowermost part of the thrust sheet S15 (KU2) are subject to debate. The highest point of the interpreted cross section is the hanging-wall anticline of S14 (166 m a.s.l.). Three different architectural surfaces occur between S11 and S16 (sensu stricto Pedersen, 2014) (Fig. 1b). The top of the cliff and minor zones, where the base-M3 unconformity occurs (mainly S12 and S16), are first-order surfaces. Another first-order surface is the décollement zone at ca. 120 m b.s.l. The ramps and upper flats between the sections are second-order surfaces. Considering the duplex stack, the intermediate flat at ca. 92 m b.s.l. (KU3d3) also belongs to the second-order surfaces. The bedding outlining the hanging-wall anticlines and the footwall syncline in S13 belong to the third order.

The Kieler Ufer section contains at least five satellite faults, which may have formed during the glacitectonic thrusting (Fig. 1b). Their age (post-, pre-, syn-tectonic) is difficult to determine, as they are truncated by the erosional unconformity at the cliff top. However, the interpretation can be related to the best-fit model using Move and the supplementary module 2-D Kinematic Modelling as well as comparisons with the faults in the surrounding sections. The reverse satellite fault in S16 dips steeply towards the SW, 235/70 (dip direction/dip), and shows an offset of at least 6 m. The satellite fault in S15 (presumably 250/47, 250/56) forms a vertical splay at a branch point about 7 m a.s.l. The offset at the main fault is about 10 m in the lowermost part. Above the branch point, the offset is less strong with ca. 3 to 4 m at each fault. The satellite fault in S13 may be moderately to steeply inclined towards the S (190/61), which is similar to that in S12. The northerly reverse fault in thrust sheet S11 is also assumed to have formed during glacitectonism of the southern structural sub-complex. It dips towards the SW by 72∘. The southern reverse fault (205/80), on contrary, can certainly be interpreted as a pre-Quaternary structure, since it does not displace the chalk top and the Pleistocene beds above (Fig. 1b).

The kinematic analysis of the Kieler Ufer section led to strain quantification and the illustration of stress orientation. It includes strain partitioning and, hence, the separation of the shortening amount into folding and faulting. Cross-section restoration and balancing of the imbricate fans of the southern structural sub-complex revealed that the Kieler Ufer section had an initial length of at least 2482 m before its glacitectonic imbrication. Thus, the entire horizontal shortening of the Kieler Ufer section is 1280 m (51.6 %) at its minimum. This includes both folding and the translation along the flats and ramps. The process of folding constitutes only 20.5 % (263 m) of the entire deformation, while the translation part is even 79.5 % (1017 m). The amount of displacement along the individual thrust faults is 206 m (KU1), 164 m (KU2), 328 m (KU3d1), 521 m (KU3d2), 438 m (KU3d3), and 243 m (KU4). The stress was directed from the SSW to NNE. The orientation fits well into the glacio-dynamic model suggested by Gehrmann and Harding (2018). The glacier with a piedmont-type lobe mainly moved from the SE to NW in the second evolutionary stage and induced local stress in all sides due to the radial propagation in the unconfined ice-marginal zone (see e.g. Ó Cofaigh et al., 2003; Jónsson et al., 2014). This induced a local orientation change of the thrust faults from a dip to the SE in the inland zone of the southern sub-complex to the S/SW at the eastern cliff.

The final cross section of the Kieler Ufer section cannot be completely balanced. There are smaller restoration gaps in the northern parts of the restored sheets KU1, KU2, and KU3d1 (S16, S15, S14) (Fig. 5). A large gap can be seen in the northern end of KU3d2 (S13). Projection of the cliff sections has been considered in the modelling process to reduce such errors. The azimuth of the final large track of the Kieler Ufer section is 210∘, adapted from the main orientation of the thrust faults or individually of the flint bands. The projection has been performed normal to the sections. However, the gaps represent volume loss during deformation rather than construction errors. Tectonic erosion may have occurred at the ramps, when the single sheets were thrust up, leading to small gaps in the frontal edge of the thrust sheets in the restored cross section. In addition, it has to be taken into account that the restored cross section only shows a 2-D interpretation. In particular, the large restoration gap in KU3d2 (S13) may point to a very complicated bedding and deformation history in the area of the Kieler Ufer section, which is highly likely related to interference between the northern and southern structural sub-complex. The specific complexity is confirmed by the high amount of shortening (51.6 %) and the duplex stack in the centre of the Kieler Ufer section.

The Kieler Ufer section represents a structural key location in the thrust-dominated-to-fold–thrust-dominated glacitectonic complex of Jasmund (see Boulton et al., 1999). The specific geometric features of the constructed and restored cross section led to major information on the kinematics and deformation history, including the relationship between the northern and southern structural sub-complex of Jasmund.

The southern sub-complex exhibits a realistic décollement depth at mainly 120 m b.s.l., because the deformation in ancient glacitectonic complexes can extend to a depth of a few tens of metres up to 200 m (e.g. Aber et al., 1989; van der Wateren, 2003; Huuse and Lykke-Andersen, 2000; Vaughan-Hirsch and Phillips, 2017). Modern structurally similar glacitectonic landforms are generally smaller, with décollement depths of about 5 to 30 m (e.g. Boulton et al., 1999; Benediktsson et al., 2010).

Considering the classic separation of glacitectonic complexes into a proximal, central, and distal zone (e.g. Boulton et al., 1999; Pedersen, 2000), the Kieler Ufer section is part of the central zone, which is characterised by a highly complicated structural framework of duplex stacks and steeply dipping thrust faults implying a high-strain area with intensive thrusting and shortening. The Kieler Ufer section has the highest amount of horizontal shortening (51.6 %) compared to the other sections of the southern sub-complex. The highest elevations of the modelled southern structural sub-complex are at the hanging-wall anticlines of the central zone (e.g. S14: 166 m a.s.l.), which may confirm the intensive deformation and compression. Boulton et al. (1999) also described the more intensive folding and thrusting for the centre of the modern glacitectonic complex in front of Holmstrømbreen on Svalbard.

There are also substantial differences between Jasmund and previously described glacitectonic complexes when the entire southern sub-complex is considered. While most glacitectonic complexes indicate a significant strain decrease from the proximal to the distal end, the Jasmund Glacitectonic Complex is characterised by a constantly high strain. Normally, the length of the thrust sheets increases towards the foreland, which implies a greater degree of shortening towards the proposed ice margin (e.g. Pedersen, 2005; Vaughan-Hirsch and Phillips, 2017). However, such a trend cannot be seen in the southern sub-complex of Jasmund. Typically, the angle of the thrust faults and the offsets decrease towards the foreland (Dixon and Liu, 1992; Pedersen, 2005; Vaughan-Hirsch and Phillips, 2017). This is also not the case in the Jasmund Glacitectonic Complex. Even though the steepest thrust sheets are situated in the most proximal zone, there are still steeply to moderately inclined faults in the central and distal zone. This phenomenon can be explained with the complex relationship between the northern and southern structural sub-complex of Jasmund. Since the northern sub-complex was formed first by an earlier glacitectonically effective glacier in the SW Baltic Sea region, there already existed the structural framework of a glacitectonic complex in the north of the southern structural sub-complex. Thus, an obstacle in front of the deformation area has controlled the imbrication during the second stage (see Gehrmann and Harding, 2018). These circumstances intensively affected the deformation process of the southern sub-complex as well as its internal structure. There was no possibility to form gently inclined thrust faults in the distal foreland during the final deformation, as the northern sub-complex acted as a resistance. The stress induced by both the glacier and the northern sub-complex led to the documented highly complex architecture, which represents an exception from classic glacitectonic-complex models (unconfined). The structural configuration of Jasmund can be used as a model characteristic of confined glacitectonic complexes. It is all the more so an example of various and intensive superimpositions in a multi-stage structural evolution (see Pedersen, 2000).