Surface cracks—geomorphological indicators for late Quaternary halotectonic movements in Northern Germany

Loading and unloading effects of the Scandinavian Ice Sheet triggered halotectonic movements in Northern Germany. We present newly detected geomorphological features—termed surface cracks—which indicate a relation between ice sheet‐induced salt movement and surface processes. As a part of the Central European Basin System, numerous Zechstein salt structures are abundant in the North German Basin. On the basis of high‐resolution digital terrain data, more than 160 surface cracks were mapped in Northern Germany, which were grouped into 30 clusters. Almost all of the surface cracks occur above the top regions of Zechstein salt structures. The surface cracks can be several kilometres long, up to more than 20 m deep and more than 100 m wide. The comparison of the shape of the salt structures and the orientation of the cracks reveals a geometric dependency, indicating that the cracks preferably occur near the crest margins of the salt structures. Furthermore, 3D seismic data from two sites show that subsurface faults originating from salt movement exist beneath the surface cracks. We interpret the cracks as surface ruptures due to ice sheet‐induced halotectonic movements. The cracks occur in a variety of Quaternary sediments and landforms. This indicates that widespread halokinetic movements occurred in the region after the last (Weichselian) deglaciation and likely before the thawing of the permafrost, possibly in a time frame from c. 30–20 ka until c. 15 ka.


| INTRODUCTION
Large parts of Northern Germany and the adjoining areas were shaped by the advances of the Scandinavian Ice Sheet (SIS) during the Pleistocene (Böse et al., 2012). Throughout the last decade, significant progress has been achieved in the understanding of the timing and dynamics of the ice advances (Hughes et al., 2016;Lauer & Weiss, 2018;Lüthgens et al., 2020;Roskosch et al., 2015;Stroeven et al., 2016), as well as in deciphering the related surface processes.
For the latter, the growing availability of high-resolution digital terrain models is of substantial importance, as these models allow precise mapping of landforms, leading to new insights in ice-related geomorphic processes, such as drumlin formation (Clark et al., 2009), erosion by subglacial meltwater (Lesemann et al., 2010), formation of icemarginal fans (Hardt et al., 2015) and ice-stream patterns (Szuman et al., 2021), as well as glacitectonic deformations (Gehrmann & Harding, 2018).
Glacitectonic deformations of sediment or bedrock, such as folds or thrusts, are the direct result of the ice sheet overriding the surface (Phillips, 2018). Apart from these deformations, the load of the SIS pressed down the Earth's crust, significantly affecting several spheres of the Earth system (Spada, 2017). A number of complex adjustment processes are still active, responding to the pressure relief (unloading) caused by the ice decay. These processes are summarized under the term glacial isostatic adjustment (GIA; Lambeck et al., 2014). One aspect of GIA, which is still ongoing, is the postglacial rebound: Fennoscandia is in an upward movement, whilst areas south of the Baltic Sea are subsiding (Spada, 2017). Among other effects, this rebound results in the reactivation of tectonic faults. As an example, it has been shown that major rivers of Northern Germany follow deep-seated tectonic faults in response to GIA (Reicherter et al., 2005;Sirocko et al., 2008). Additionally, the occurrence of historic earthquakes along Late Cretaceous fault zones in Germany was attributed to GIA . However, GIA-related tectonic processes are difficult to differentiate from long-term tectonic processes related to plate movements, such as the Alpine convergence (Liszkowski, 1993;Reicherter et al., 2005;Sirocko et al., 2008;Stewart et al., 2000).
As part of the Central European Basin System (CEBS; Warren et al., 2008), numerous salt structures exist in the subsurface of Northern Germany and adjacent areas at top depths ranging from less than a few hundred to a few thousand metres. The response of salt structures to ice load and GIA adds an additional degree of complexity to the understanding of subsurface and surface processes in the CEBS. The influence of subsurface salt structures on the present-day geomorphology has been investigated previously: relations with the distribution of terminal moraines (Gripp, 1952;Schirrmeister, 1998), subglacial tunnel valleys (Kristensen et al., 2007;Wenau & Alves, 2020) and modern river valleys (Reicherter et al., 2005;Sirocko et al., 2008) have been shown. These correlations may result from halotectonic movements which were triggered by oscillations of the SIS in the Pleistocene (Lang et al., 2014). However, distinct landforms in the Quaternary landscape, which can be directly related to ice sheetinduced salt movements, have so far not been recognized.
In this paper, we present more than 160 negative linear landforms (here termed surface cracks) in Northern Germany, which are up to several tens of metres wide, up to several kilometres long and cut up to several tens of metres into the surface. These landforms were detected on the basis of high-resolution LiDAR terrain data. The cracks cannot be categorized as any previously described landform of the glacial landscape. Through a deeper investigation into the morphology and spatial occurrence of these newly detected geomorphic features, it has become apparent that beneath almost all of them, salt structures are present in the deeper subsurface or close to the surface.
We describe the geomorphological characteristics of the cracks and their spatial distribution in the glacial landscapes of Northern Germany, as well as their spatial correlation to salt structures. For three specific localities, we present a more detailed description. At two of them, we additionally analyse 3D seismic subsurface data to investigate a possible correlation with structures of the deeper subsurface and the observed surface cracks. We present a theory on the possible development of the observed cracks which is based on the loading and unloading effects that ice sheets have on salt structures (Lang et al., 2014). We propose that the cracks are the geomorphic imprint of surface expansion that are due to ice sheet-induced salt rise, and we establish a time frame for their possible formation.

| Geological background
The CEBS (Warren et al., 2008) ranges from the North Sea in the west to Poland in the east (Figure 1, inset map). The basin development was initiated in the Permo-Carboniferous time period and is today filled with sedimentary deposits of more than 7 km in thickness (Hoth et al., 1993). During the Zechstein (c. 258-252 Ma;STD, 2016), several marine transgressions flooded this vast basin. The seawater evaporated under arid tropical climate conditions, resulting in a characteristic depositional sequence of evaporitic rocks, such as rock salt (halite), anhydrite, gypsum and carbonate. Seven main evaporitic cycles can be differentiated (STD, 2016).
These evaporitic rocks, which consist of a large volume of rock salt (halite), were subsequently buried by Mesozoic and Cenozoic sediments. Due to the increasing sedimentary load and the relatively low density of rock salt compared with other sedimentary rock types, the salt can start to flow and move upward in certain areas, causing a deformation of the overburden (Nettleton, 1987;Warren, 2016). The resulting salt-related structures can develop into different geometries and are known, for example, as salt pillows, salt domes and salt diapirs.
The respective structures form under the influence of gravity and tectonic extension or compaction (Brandes et al., 2012;Scheck et al., 2003b;Trusheim, 1987). The tectonic processes related to the salt movement are termed halotectonics. Salt movement and the formation of salt structures started in the mid-to late Triassic and continued in the late Cretaceous to the earliest Cenozoic until the Neogene (Scheck et al., 2003a;Scheck-Wenderoth et al., 2008). Halotectonic processes not only deform the overburden, but also affect the sedimentation in the CEBS . For example, rim synclines, which developed in the surroundings of rising salt structures due to salt migration, were filled with thick Cenozoic sediments (Brandes et al., 2012;Scheck-Wenderoth et al., 2008). Apart from the tectonic events, the loading and unloading of ice sheets during the Pleistocene may additionally have influenced salt movements (Lang et al., 2014;Strozyk et al., 2017).
Salt structures in the subsurface can be mapped by using indirect (geophysical) and direct methods (e.g. by measurements of the gravity field, seismic exploration and deep drilling). The occurrence of salt structures in the CEBS is well known, as they provide an important economic resource for the storage of renewable energies or for geothermal energy Pollok et al., 2015), as well as for the possible storage of radioactive waste (BGE, 2020;Warren, 2016). Depending on the evolution and stage of a salt structure (pillow, diapir), different fracture styles or graben structures may develop in the overburden in reaction to the respective salt activity (Yin & Groshong, 2006). However, the details of the deformations and structural layering around a salt structure are often less resolved and require a thorough in-depth interpretation of all available subsurface data.
In the Cenozoic, large parts of the CEBS were flooded by the paleo North Sea, resulting in about 80 m-thick marine clay deposits of the Oligocene ('Rupelton'; Knox et al., 2010). The Rupelton acts in many areas as a hydraulic barrier between the deeper saline aquifer system and the upper utilized drinking water (Yordkayhun et al., 2009a) that is most often related to Quaternary sediments.

| Brief quaternary landscape history
The SIS oscillated several times into Northern Germany during the Pleistocene (Ehlers et al., 2011). While the Elsterian and Saalian ice masses reached as far as the low mountain ranges, the ice advances of the last glacial cycle (Weichselian) were less extensive in Germany (Böse et al., 2012; Figure 1). As a result, glacial landscapes of different ages are present: the old morainic (last glaciated during the Elsterian or Saalian) and the young morainic landscapes (last glaciated during the Weichselian). During the Weichselian, at least two ice advances reached Northern Germany, with the older advance (W1, late MIS 3 to early MIS 2, c. 30 ka) reaching farther south than the younger (W2, MIS 2, c. 20 ka). Thus, the Weichselian glacial landscapes can be differentiated in an older (W1 advance) and a younger (W2 advance) young morainic area (Hardt et al., 2016;Lüthgens et al., 2011Lüthgens et al., , 2020. After the ice decayed, and also in the non-glaciated areas, periglacial climate conditions prevailed for several thousands of years in Northern Germany (Vandenberghe et al., 2014) (Sirocko et al., 2008). This can be explained by a relatively high thermal conductivity of rock salt, leading to an increased heat flow towards the surface.
The formation and transformation of dunes and aeolian cover sands was favoured in different phases of the Late Glacial and Holocene. A first sign of strong aeolian activity occurred in the study area from 15 ka onwards in geochronological records (Kappler et al., 2019). A stability phase around 12.7-11.5 ka, as indicated by the Usselow and Finow soil formation, was followed by another aeolian activity phase during the Younger Dryas (Kaiser et al., 2009;Kappler et al., 2019).  Liedtke (1981). Ice marginal position of W1 advance according to Lüthgens et al. (2020). Salt structures by InSpEE (2015), land surface by Shuttle Radar Topography Mission (Jarvis et al., 2008), bathymetry by IOWTOPO (Seifert et al., 2001). The blue area in the inset map shows the extent of the Zechstein basin (Słowakiewicz et al., 2018). Inset map based on data provided by NaturalEarth (free vector and raster map data available at naturalearthdata.com) [Color figure can be viewed at wileyonlinelibrary.com] 2.3 | The influence of ice sheet-induced tectonics and halotectonics on the modern surface The morainic landscape of Northern Germany shows an underlying pattern of the three main tectonic strike directions NW-SE, NNE-SSW and NE-SW, as seen in the direction of river valleys and coastlines (Sirocko et al., 2002(Sirocko et al., , 2008. Likewise, there is a correlation between deep-seated faults and the location of subglacial tunnel valleys (Dobracki & Krzyszkowski, 1997;Stackebrandt, 2009). Additionally, ice-marginal valleys and ice-marginal positions have been observed to run parallel to major tectonic lineaments or boundaries (Reicherter et al., 2005).
As the GIA is still in progress, related processes may reactivate pre-Quaternary tectonic fault zones and thus influence Quaternary surface processes (Sirocko et al., 2008). According to Brandes et al. (2015), the ice decay after the Weichselian glaciation triggered the reactivation of Late Cretaceous reverse faults in Northern Central Europe, as seen in the spatial distribution of historic earthquake epicentres. The Late Cretaceous faults of Northern Central Europe strike in a WNW-ESE direction, which is generally parallel to the extent of the Weichselian ice margin in Northern Germany. Thus, the ice-induced stress field matches the paleostress field, leading to a relatively high reactivation potential of these faults, in contrast to faults with a different orientation to ice marginal positions Stewart et al., 2000). Similarly, Ihde et al. (1987) determined recent vertical crustal movements on the basis of relevellings in Northeast Germany and found that especially the NW-SE and NNE-SSW striking faults have been recently mobile.
In addition to ice sheet-induced tectonics, it has been recognized that halotectonic processes have an influence on Quaternary landscape development. For example, Weichselian (Schirrmeister, 1998) and Saalian (Lehné & Sirocko, 2007) ice marginal positions correlate with the distribution of subsurface salt structures ( Figure 1). This phenomenon can be explained by a loading effect of the advancing ice sheet which triggers salt rise and the subsequent formation of a topographic obstacle that the ice sheet cannot pass (Lang et al., 2014;Lehné & Sirocko, 2007). Alternatively, the elevated terrain above a salt structure may form a sediment trap between the advancing ice sheet and the salt structure. The sediments accumulated in this trap would be compressed by the advancing ice, and the formation of push moraines in the vicinity of salt structures would be favoured (Sirocko et al., 2008). Wenau and Alves (2020) report that the formation of tunnel valleys in the North Sea basin are coupled to subsurface faults originating from salt walls.
The loading and unloading effects of the advancing and downwasting ice sheets have had a significant effect on the activity of the salt. The basic mechanisms behind these processes were previously described by Liszkowski (1993) and modified by Sirocko et al. (2008).
Their concepts were later tested in a numerical model by Lang et al. (2014). The model results show that the load of a 300-1000 mthick ice sheet which transgresses the structure will force the salt downwards, leading to a lateral extension of the upper sections of the diapir. After the ice vanished (unloading), the diapir will start to rise again and deform the overburden. The load of an ice sheet also influences diapirs outside the actual ice extent. The surface pressure that the ice applies on thinner salt layers forces them to flow towards diapirs not covered by the ice sheet (Lang et al., 2014).

| METHODOLOGY AND METHODS
During the investigation of high-resolution terrain model data of Northern Germany, the repeated occurrence of linear negative landforms that were, for the most part, previously unrecognized was discovered. A first description of these features, only for the Schorfheide region, was given by Krambach et al. (2016) but without going into detail on the formation process. The Schorfheide region served as reference area, and accordingly we concentrated on digitizing elongated, negative linear landforms (surface cracks) with a minimal depth of at least 50 cm, with a clearly distinguishable slope, which may be straight or curved, but which do not meander or display any other characteristic of fluvial landforms.
In order to further investigate the characteristics of these landforms (i.e. their basin-wide distribution and whether they represent a rare or a more common geomorphologic feature) and the possible processes involved in their formation, the following methods were applied.

| Terrain data, geological data and mapping
In Northern Germany, the federal states of Brandenburg, Mecklenburg-Western Pomerania and Schleswig-Holstein were systematically analysed in a search for geomorphic landforms that might resemble cracks. The analysis was explicitly not restricted to the vicinities of salt structures but was carried out on the whole territory of the aforementioned states. However, the available quality of terrain data differed between the federal states.
The LiDAR digital terrain model (DTM) of Brandenburg has a horizontal resolution of 1 m/pixel and is freely available as a download in .
XYZ format (GeoBasis-DE/LGB, 2020). The DTM resembles the bare earth surface; the vegetation was removed by the supplier using the last pulses of the laser scan (Tarolli, 2014). The subsets were merged and resampled into geoTIFF datasets using the GDAL warp utility with a cubic resampling to 2 or 5 m. The resampled 5 m version was used for initial scanning of the whole area. If an area with suspicious landforms was found, the 2 m version of the DTM was used for detailed analysis and mapping. Outside the state of Brandenburg, preprocessed hillshade terrain data were accessed via the web map service (WMS) interfaces provided by the geoportals of Mecklenburg-Western Pomerania (https://www.geoportal-mv.de/portal/) and Schleswig-Holstein (https://www.gdi-sh.de/DE/GDISH/Geoportal/ geoportal_node.html).
In order to reduce mapping bias that results from certain hillshade options (Smith & Clark, 2005) and other subjective properties, the results were always compared to aerial imagery. If applicable, profile tools for the validation of cross-sections and contrast enhancement on the source DTM were applied. Furthermore, the mapping was independently performed by three people. Areas of interest were mutually evaluated, and they were discarded or combined where necessary.
The distribution and properties of the salt structures (InSpEE, 2015;Pollok et al., 2015) were freely available from the geo- The clusters were named after nearby villages (e.g. Bockenberg) or landscapes (e.g. Schorfheide) and received an abbreviation (e.g. BB for Bockenberg or SH for Schorfheide). Nearby surface cracks were subsequently grouped into clusters. The cracks received the abbreviation of the cluster followed by a sequential number (Table 1) The database containing all mapped surface cracks is available as a download in geopackage (*.gpkg) and KML format in the online Supporting Information for this paper (Data S1).

| Geophysical data
At two sites, we could consider 3D seismic surveys for studying the deeper subsurface. They were initially conducted for geothermal exploration and CO 2 sequestration, respectively. The first example is the Groß Schönebeck structure for the exploitation of deep geothermal energy (Bauer et al., 2020;Krawczyk et al., 2019); the second refers to the CO 2 storage pilot site Ketzin (Lüth et al., 2020).
In 2017, a 3D seismic survey was carried out in the Groß Schönebeck area (21 in Figure 1   Periglacial-fluvial landforms oriented towards Lake Werbellinsee and Lake Meelake are disrupted by cracks ( Figure 5B-d), indicating  Figure 6A). The main faults are oriented NW-SE and NE-SW ( Figure 6). Due to the lower resolution of the seismic data in the shallower subsurface (which was designed to decipher structural information for the target horizon in more than 4 km depth), the detailed behaviour of subsurface faults above the Zechstein is less distinct. Although the overall seismic reflectivity in that area is reduced, the seismic horizons seem not to be strongly affected by faults ( Figure 6C and 6D) and the observed ruptures may be caused by internal Zechstein stringer deformations primarily. Notably, the orientation of the surface cracks shows a remarkable accordance with the positioning of the salt structure doming ( Figure 6B).

| Ketzin cluster
The According to the study of Yordkayhun et al. (2009a), the faults could be traced at least into the lowermost Rupelian clay.

| DISCUSSION
Can we differentiate between tectonics and halotectonics?
The complex geological history of Northern Central Europe results in geomorphic features responding to major tectonic lineaments. As previously shown, ice-induced stress led to reactivation of some tectonic faults in the region . Local, probably ice sheet-induced, halotectonic movements are therefore challenging to differentiate from large-scale tectonic processes. Brandes et al. (2013) showed that at the Elm salt pillow, the salt-induced stress vectors are dominated by local effects such as the shape of the Zechstein salt structure. They form radial vectors around the salt structure and deviate up to 90 from the tectonic paleostress fields. As WNW-ESE is a main tectonic strike direction in Northern Germany with recently mobile faults Ihde et al., 1987;Sirocko et al., 2002Sirocko et al., , 2008,  Table 1). As the shape of the salt structures is also controlled by tectonic lineaments (Scheck-Wenderoth et al., 2008), a WNW-ESE to W-E direction is tectonically inherited, as previously shown (Kossow et al., 2000). It should be noted that many of the cracks are curved; hence, a precise determination of a strike direction is impeded. Notwithstanding this, the shape and orientation of the cracks appear to be primarily controlled by the shape of the underlying salt structure ( Figure 10).
In the case of the Schorfheide cluster, where the surface of the Groß Schönebeck salt structure is displayed in 3D seismic data, the surface cracks appear mostly in the area to the north/northeast of the peak of the salt structure. The curved cracks show a tendency to adhere to the shape of the salt pillow, although it is at a depth of more than À2000 m. Similar observations were made at several other sites.
In the example of the Flieth salt pillow, the Bockenberg and Weiler clusters seem to follow the shape of the salt structure ( Figure 10-1

| Suggested formation process
According to their geomorphological characteristics, the surface cracks cannot be of a fluvial or other erosional origin (Krambach et al., 2016). If the landforms were frost cracks due to permafrost thawing, which might be favoured above salt structures due to the increased heat flow through the salt (Sirocko et al., 2008), we would expect the cracks to be more homogeneously distributed above other salt structures in the region. Additionally, we would expect to see a stronger correlation with the depth of the salt structures, as the thickness of the overburden attenuates the heat flow effect (Sirocko et al., 2008). If the surface cracks were fractures caused by differential compaction due to ice load, a process suggested for the development of the Hugin Fracture in the North Sea (Landschulze & Landschulze, 2021), we would also expect them to be more widespread in the region. Most of the cracks were detected in areas where salt rise due to ice load was proposed earlier, especially along the W2 ice marginal position (Figure 1; Lang et al., 2014;Schirrmeister, 1998) or at the salt diapir of Sperenberg (W1 ice marginal position; Figure 1; Stackebrandt, 2005).  Figure 5). Sandy or diamictic substrates should react differently to surface expansion owing to their different cohesion strengths (Regmi et al., 2015). However, both substrates might react similarly if they are frozen. Nonetheless, loose sediment can also react rigidly in non-permafrost environments, as illustrated (e.g. by earth fissures as reported from presently tectonically active regions; Zang et al., 2021).
These earth fissures can develop at the surface in unconsolidated sediment under the influence of tectonism, faulting and human intervention (e.g. subsidence due to groundwater extraction; Asfaw, 1998;Peng et al., 2020). Although these fissures are smaller in width and depth than the surface cracks and are not related to halotectonics, they may serve as a modern example illustrating the interplay between subsurface processes and surface ruptures.
The loading and unloading effects of advancing and decaying ice sheets on subsurface salt structures can cause uplift in front of the ice sheet (Sirocko et al., 2008) and intrasalinar flow (Lang et al., 2014).
Both effects help to explain the formation of surface cracks The rise of this diapir during the Quaternary is attributed to loading effects of the SIS (Stackebrandt, 2005). The existence of surface cracks in Sperenberg, where salt rise during the Quaternary has been documented, supports the suggested formation process.
According to their model results, which returned a maximum uplift of 4 m, Lang et al. (2014) claim that the estimated vertical displacement of the aforementioned studies is too high. However, our results may indicate that the high uplift rates, as stated by Sirocko et al. (2002) and Stackebrandt (2005), are not necessarily overestimated.

| Timing
The cracks must have formed after the respective areas were last glaciated by the Weichselian ice sheet because any pre-existing cracks must have been deformed by the advancing ice or filled with sediment in the glacial or paraglacial depositional environments. Central and parts of southern Brandenburg were last glaciated during the late MIS 3/early MIS 2 (W1 advance) and were ice-free after 30 to 25 ka onwards (Hardt et al., 2016;Lüthgens et al., 2020). Only Northern Brandenburg and regions to the north were glaciated again during the MIS 2 (W2 advance) and ice-free from c. 19-18 ka onwards (Lüthgens et al., 2011(Lüthgens et al., , 2020. Surface cracks were found in the older young morainic area of the W1 advance (e.g. Sperenberg, Ketzin) as well as in the younger young morainic area of the W2 advance. Thus, the cracks of Southern Brandenburg could be more than 10 ka older than those to the north of Brandenburg or in the areas of Mecklenburg-Western Pomerania and Schleswig-Holstein.
The minimum age we consider is related to the thawing of the permafrost. Most of the permafrost soils in the region had probably vanished due to the climate amelioration of the Bølling-Allerød interstadial, starting at c. 14.6 ka (Köhler et al., 2014), although patches of discontinuous permafrost may have prevailed until the Younger Dryas (Isarin, 1997) and Preboreal (early Holocene), as shown in different palaeoenvironmental studies in the wider region (Błaszkiewicz, 2011;Dietze et al., 2016;Kaiser et al., 2009;Krüger et al., 2020).
Dunes were occasionally accumulated inside surface cracks (SH-1, SH-2; Figure 5). These cracks must be older than the dunes or more or less of the same age.

| Relation to Weichselian ice dynamics
From the spatial distribution of the surface cracks in Northern Germany ( Figure 1) and the statistics presented in Figure 4A, it becomes apparent that the majority of these landforms exist in the landscapes of the W2 advance (c. 60%), whereas significantly fewer cracks were detected in the landscapes of the W1 advance (c. 40%).
In addition, the clusters with a larger number of cracks tend to be This phenomenon may relate to different ice dynamics of both ice advances. The W1 advance of the SIS in the study region was guided by topographic obstacles such as the horst of Bornholm or the depression of the Oder basin and had a distinct lobate form (Hardt et al., 2016;Lüthgens et al., 2020). As such, the ice thickness and the corresponding pressure on the subsurface were comparably smaller, presumably resulting in the formation of fewer cracks (e.g. Netzeband, Sperenberg). In contrast, the W2 advance occurred during the Last Glacial Maximum (LGM) when the ice sheet was presumably thicker and the ice front more uniform. Thus, the ice applied comparably more pressure on the ground and caused more intense salt movement, as indicated by the higher number of cracks occurring along the W2 terminal moraine-regardless of the depth of the salt structures.
The relationship between ice thickness and intensity of salt movement is in agreement with the model results of Lang et al. (2014). Our study is another indication that during geologically recent time spans, significant halokinetic movements of salt structures in Northern Germany occurred. Further investigations will reveal whether the described processes also occurred in other areas of the CEBS and in other comparable settings. Beyond the geomorphological imprint, and in the context of the ongoing search for radioactive waste disposals, the described processes suggest that further research is necessary to assess the long-term stability of salt structures.
• Laser scan data of Federal State Brandenburg: https://data.