Elevation and thickness of the 11–10 Kyr old ‘Sinkholes Salt’ layer in the Dead Sea: clues to past limnology, paleo-bathymetry and lake levels

The sinkholes along the Dead Sea (DS) shores form by dissolution of an 11–10 kyr old subsurface salt layer (hereafter named the ‘Sinkholes Salt’) that precipitated on the lake’s floor during periods of negative water balance, water level decline and salinity increase. We analyze the variations in absolute elevation and thickness of this layer in 40 boreholes along the western shores of the DS, reconstruct water-body stratification, past lake levels, and paleo-bathymetry during salt deposition, and comment on the role of the salt-layer elevation in future sinkhole formation. In the northern basin of the DS, maximum thickness of salt (~ 23 m) is found where salt top and bottom elevations are below ~ 440 meters below sea level (mbsl) and ~ 465 mbsl, respectively. Above these elevations the salt layer gradually thins out until 416 mbsl, above which it is no longer found. These relationships suggest that thermohaline stratification, with a thermocline at 25–30 m depth, similar to the present day dynamics of the DS, developed annually during the salt-precipitation period, giving rise to uniform salt accumulation below the thermocline and partial to full dissolution above it. Salt accumulation was controlled by the bathymetry of the lake and its configuration relative to the thermocline, and locally hampered by discharge of subaqueous under-saturated groundwater. The truncation of the salt layer at elevation of 416 mbsl is attributed to salt dissolution down to this elevation by a relatively diluted upper water layer that developed following inflow of fresh surface water at the end of the salt period. This event also marks the change to a positive water balance and lake level rise from its lowest stand of ~ 405 mbsl, as determined from limnological considerations.


Introduction
The Dead Sea (DS), the lowest place on the continental Earth, is located along the DS Transform, a ~ 1000 km-long plate boundary separating the Arabian plate from the African plate (Freund 1965;Garfunkel 1997). The DS (Fig. 1) is a hyper-saline terminal lake that currently occupies the northern part of the DS basin. Its bottom elevation is at ~ 730 meter below sea level(mbsl) and it is separated from the southern sub-basin by an elevated sill (the Lynch strait; Fig. 1) at ~ 400 mbsl. The southern sub-basin, with bottom elevation of ~ 410 mbsl is currently occupied by evaporation ponds of the  (Table 1). In red-boreholes where groundwater-level-drop rates were determined.
b En Gedi subset. Solid white line marks the location of the cross section shown in Fig. 6. c Ze'elim subset Israeli and Jordanian potash companies. In recent decades the DS water level has been dropping at a rate of about 1 m/yr, reaching the current (8/2022) elevation of ~ 437 mbsl (Israel Hydrological Service 2022). In the late 1970s, when the DS level dropped to about 400 mbsl, the water bodies of the two basins were disconnected, and at about the same time halite began precipitating from the brine (Steinhorn 1983;Gavrieli 1997).
Since the Pliocene the DS basin was occupied by water bodies of different characteristics and areal extents (Zak 1967;Stein 2001). The sediments deposited from these water bodies are excellent archives of the climatic, environmental and seismic history of the Levant region as well as the limnological evolution of the lakes. During high stands and positive water balance, the hypersaline lakes were meromictic with stratification that may have lasted for hundreds to thousands of years. Following negative water balance and lake-level drop the water column underwent overturn and homogenization and the meromictic regime thereby changed to a monomictic regime (Stiller and Chung 1984;Stein 2001;Torfstein et al. 2008;Levy et al. 2017). Thick halite intervals recovered by the ICDP Dead Sea Deep Drilling Project (DSDDP) cores show evidence for severe arid conditions in the eastern Mediterranean during the last three inter-glacials, with detailed chronological constraints (Neugebauer et al. 2014;Kiro et al. 2017). Some of these layers have been extensively studied and correlated to global and local climatic events (Torfstein et al. 2015;Levy et al. 2017). Based on water and salt budgets derived from the Dead Sea brine composition and the thickness of salt in the cores, Kiro et al. (2017) suggested major lake-level drops of ∼170 m from lake levels of 320 and 310 mbsl down to lake levels of ∼490 and ∼480 mbsl, during MIS 5e and the Holocene, respectively, reflecting severe droughts. Of particular interest is the 11-10 kyr old (radiocarbon dating; Stein et al. 2010) salt layer that lies at the base of the Holocene Ze'elim Formation (Yechieli et al. 1993). This layer marks one of the major lake-level drops (Kiro et al. 2017), but more importantly, it has significant environmental impact, as its present dissolution by groundwater and floodwater is the major cause for the formation of > 5000 collapse sinkholes along the DS shorelines during the last decades (Frumkin and Raz 2001;Abelson et al. 2006;Yechieli et al. 2006;Nof et al. 2013;Watson et al. 2019;Al-Halbouni et al. 2021). We thus focus our current study on this 'Sinkholes Salt' layer.
An important advance toward our understanding of salt deposition in hypersaline lakes comes from numerous studies of the modern (post 1980) DS water body (Steinhorn 1985;Anati et al. 1987;Anati and Stiller 1991;Anati 1997;Gavrieli 1997;Gertman and Hecht 2002;Lensky et al. 2005;Arnon et al. 2016;Sirota 2020;Armon et al. 2020). The basic observations recurring in these studies are the seasonal transition from relatively uniform profiles of temperature  Arnon et al. (2016) and Sirota et al. (2020)) showing seasonal changes from the development of an early summer stratification with an upper, warm layer (epilimnion), a lower, cooler layer (hypolimnion) and a transitional layer (metalimnion), through an abrupt temperature step (thermocline) in the late summer, to a winter uniform temperature profile. The width of the metalimnion (pink rectangles), as well as its upper and lower boundaries vary locally and temporally. b Schematic illustration of the lake floor showing the late summer thermocline depth, above which salt precipitates mainly during the winter and dissolves during the spring and summer stratification, and below which salt precipitates yearround Table 1 Location (Fig. 1), elevation and thickness of the salt layer, and rate of groundwater-level drop in boreholes along the Dead Sea *-Dragot 4 borehole did not reach salt bottom and is thus excluded from the analysis #-BHCR-23 borehole is excluded from the analysis due to the existence of an 8 m thick clay intercalation between the salt layers, resulting in an ambiguous determination of its bottom level and salinity over the entire water column during the winters to thermohaline stratification with vertical temperature and salinity variations during the summer periods (Fig. 2). During these months, the lakewater body may be divided to an upper, warm and saline layer (epilimnion), a lower, cooler and slightly less saline layer (hypolimnion), and a transitional layer (metalimnion) in between. The thickness of the latter, the depths of the boundaries of the three layers, as well as the detailed shapes of the temperature and salinity profiles in the lake are highly variable among different parts of the DS and develop during the stratification period (Arnon et al. 2014;Sirota et al. 2020). Daily changes are also recorded in response to internal waves (Arnon et al. 2014). These seasonal variations in temperature and salinity affect halite saturation and consequently, the dynamics of halite deposition and accumulation on the DS floor (Sirota et al. 2016). The current study expands our comprehension of the extent and precipitation process of the 'Sinkholes Salt' layer by analysis of a partly new, valuable database of lithological logs retrieved from more than 40 boreholes that penetrated the salt layer at the western coast of the DS (Fig. 1). On the eastern side of the DS only two boreholes were drilled in the vicinity of the sinkholes (Polom et al. 2018) but did not encounter any salt layer. Detection of the 'Sinkhole Salt' layer by geophysical methods on both sides of the DS has shown to be ambiguous or less accurate, and is thus excluded from this study. We first show systematic elevation-thickness variations of the salt layer along the entire shoreline of the DS. Then, by analogues from the modern DS, we explain these systematic variations (including some outlying cases), and reconstruct the lake level at the end of the salt period, its stratification and its paleo-bathymetry during salt deposition. Finally, we comment on the role that the elevation of this layer may play in future sinkhole formation along the DS.

Borehole data
Boreholes were drilled along the DS since the 1990s to depths of 30-100 m, reaching the 'Sinkholes Salt' layer in almost 50 boreholes and penetrating it from top to bottom in 40 boreholes ( Fig. 1 and Table 1). The border between salt-and no salt-bearing boreholes marks the western extension of the salt layer. From this wide dataset, we pick two subsets of densely distributed boreholes to examine local salt elevation and thickness variations (Fig. 1). The first subset of 19 boreholes was drilled in the Ze'elim Fan by the Dead Sea Works Ltd. (DSW) between 2013 and 2016 (boreholes 15-33 in Table 1), and was recently released for research use. The second subset includes five boreholes drilled along the Arugot Wadi near En Gedi (5-9 in Table 1). The complete dataset (40 boreholes) is used to examine elevation-thickness variations of the salt layer along the entire western shoreline of the DS.

Sinkhole and subsidence maps
For a more precise determination of the western boundary of the salt layer, we use sinkholes and subsidence maps, considering that both features form mostly along the margins of the salt layer (Abelson et al. 2006;Ezersky et al. 2017). Sinkholes and subsidence maps are available in the recently published sinkhole early warning website of the Geological Survey of Israel https:// egozi. gsi. gov. il/ webap ps/ hazar ds/ sinkh oles_ subsi dence/ (Nof et al. 2019).

Groundwater levels
Groundwater levels were measured in most boreholes during the drilling process; however, repeated measurements of groundwater levels in the following years are available from only 12 boreholes (Table 1). Sinkholes form where salt is in contact with undersaturated groundwater. With the drop of the DS and the associated groundwater levels below the salt, new sinkholes are less likely to form by groundwater dissolution. Thus, the elevation of the salt-layer bottom, combined with the local rate of water-level drop, could be important factors in estimating the time when groundwater level will drop below the salt layer, after which salt dissolution by groundwater will decrease.

Interpolation, extrapolation and contouring
To map the spatial variations in salt elevation and thickness and rates of groundwater-level drop we use Geographic Information System (GIS) tools. Basic interpolation and contouring tools perform well where the measurement points are uniformly scattered within the area of interest. However, the irregular scatter of the boreholes along the DS limit programmed interpolation tools to the close vicinity of the boreholes, beyond which the interpolation outputs become less reliable. To overcome this limitation, we performed manual interpolation, extrapolation, and contouring between and beyond the measurement points, applying geological considerations of sedimentary trends instead of machine-based contouring, and then converted the contour maps to raster images using GIS tools. Clearly, these maps also become less reliable as the distance from the boreholes grows.

Results
The 'Sinkholes Salt': texture, elevation and thickness The salt layer has been recovered from continuous cores in the Ze'elim boreholes and from intervallic cores in other DS boreholes. The salt is white to grey, coarse, massive to friable, and in some intervals interbedded with wavy layers of very fine white salt (Fig. 3a). Coarse transparent cubic salt crystals up to 3 cm in size cemented by clayey matrix are found in places. The salt may be porous with voids up to 6 cm large. Salt is interbedded with clastic sediments and laminated clay-silt and aragonite, several mm to more than 1 m thick (Fig. 3b). These clastic intercalations comprise less than 10% of the entire thickness of the salt interval, and their relative abundance is lower where the salt layers are thin, and higher in the thick layers. The salt layer is generally mantled by thick sequences of laminated clay-aragonite. In a few locations (mostly within the main riverbeds) salt is underlain and/or overlain by coarse clastic material.
The relatively dense borehole distribution in the Ze'elim fan enables a spatial analysis (contouring and cross sections) of the salt layer within this area (Fig. 4). The western boundary of the salt layer (zero thickness) is marked between salt-and no-salt boreholes or at the westernmost margins of sinkholes and subsidence areas. The maps and profiles (Figs. 4 and 5) show significant elevation and thickness variations, with shallow and thin salt in the central part of the fan and deeper (salt bottom at 460-470 mbsl) and thicker (up to 23 m) salt at the northeastern and southwestern margins of the fan. The salt top map is rather similar to the salt bottom map (Fig. 4b, c), yet moderated by variations in salt thickness (Fig. 5).
The En Gedi subset consists of five boreholes that reached the bottom of the salt layer, and display significant thickness and elevation variations (Figs. 1 and 6 and Table 1). This practically hinders construction of a contour map as done for Ze'elim, and we thus present the spatial elevation-thickness variations along a  (Table 1). a Coarse salt (left) and wavy layers of very fine white salt (right). b Salt interbedded with laminated clay-silt-sized clastic sediments and aragonite WNW-ESE transect (Fig. 6). The area also includes a borehole with no salt at the west (EG-19) and a borehole that reached the top of the salt layer only (EG-17). As in Ze'elim, the thickest salt in En Gedi boreholes has the lowest bottom elevation (borehole M-1 in Fig. 6).
Systematic elevation-thickness variations of the salt layer occur also among most other boreholes in the northern basin of the DS (Table 1; Fig. 7a, b). The salt layer in these boreholes gradually thickens as bottom elevation deepens from ~ 420 to ~ 460 mbsl (Fig. 7b). Below salt bottom of ~ 460 mbsl (and top elevation of ~ 440 mbsl), thickening of the layer with depth is indistinct (Fig. 7). Two regions, however, show exceptional elevation-thickness relationships (Fig. 7): 1. En Gedi, where three out of five sites show salt thickness significantly lower compared to sites at similar elevations elsewhere (grey circles in Fig. 7). 2. Zohar-Boqeq (which are located in the southern DS basin), where salt thickness is exceptionally high (up to 35 m) compared to all other DS sites (green circles in Fig. 7).
Also noteworthy (but not included in Fig. 7) is a 19-m thick salt layer, with ~ 6 m thick clayey intercalations, encountered at elevations between 792 and 811 mbsl in the deep Dead Sea Drilling Project (DSDP) borehole at the center of the northern DS basin, which has a similar age range as the onshore 'Sinkholes Salt' (Neugebauer et al. 2014). Its thickness is comparable to values encountered in the shoreline boreholes, suggesting that the elevation-thickness relationships described above apply for the deeper part of the DS basin as well.

Discussion
The borehole data presented above provide a wealth of information pertaining to the depth and thickness of the subsurface 'Sinkholes Salt' layer, which has been formerly overlooked. These two parameters are most valuable for reconstructing the conditions prevailing during the precipitation of that layer, 11-10 kyr ago, and on the other end, for our capability to predict future sinkhole formation along the DS. Below we discuss these two aspects.
Origin and paleogeographic implications of the depth-thickness relationships of the 'Sinkholes Salt' The presence of halite layers in the Pleistocene to Holocene sedimentary sequence in the DS basin is interpreted to represent periods of negative water balance, lake-level drop and increased salinity of the saline lakes that occupied the basin (Stein et al. 2010;Torfstein et al. 2015;Kiro et al. 2017;Levy et al. 2018). The information on the late Pleistocene to Holocene halite layers is derived from the core retrieved from the center of the lake in the framework of the Dead Sea Drilling Project (DSDP; Neugebauer et al. 2014;Levy et al. 2017). The numerous shoreline boreholes to the relatively young 11-10 kyr old salt layer hereby analyzed, add important insight into the mechanism of salt precipitation and dissolution in the DS basin.
The elevation-thickness relationships of the salt layer (Fig. 7) that precipitated during the transition from Lake Lisan, the late Pleistocene precursor of the DS, to the Holocene DS, illuminate two overall features: 1. The 'Sinkholes Salt' layer gradually thickens with depth from 0 to 5 m at salt-bottom elevations of 421-426 mbsl to a maximal thickness of 23 m at salt top and bottom elevations of ~ 440 mbsl and ~ 465 mbsl, respectively. Below these elevations the salt layer does not thicken (Fig. 7a, c). 2. The 'Sinkholes Salt' is not encountered at elevations above 416 mbsl in any borehole (Fig. 7a).
To explain these two observations we draw analogues from modern DS annual limnological cycles and their associated salt-precipitation processes. During the winter, the thermal profile in the DS is depth-independent (Fig. 2a). In spring, as air temperature rises, a thermal stratification begins developing. Consequent heating and evaporation from the surface water during the coming months of spring and summer leads to a warmer and slightly more saline epilimnion, and cooler, less saline hypolimnion (Anati and Stiller 1991;Arnon et al. 2016;Sirota et al. 2016). The thermocline with the transitional zone of intermediate salinity and temperatures (metalimnion) develops between the two layers. The thickness and depth of the metalimnion changes throughout the season, and geographically, from north to south (Arnon et al. 2014;Sirota 2020). Daily variations are also recorded in response to internal waves (seiches; Arnon et al. 2014). Towards mid-summer, the salinity and temperature profiles develop into a relatively sharp step at depth of 25-30 m (halocline and thermocline, respectively; Fig. 2a). The elevated temperatures of the epilimnion result in minor undersaturation with respect to halite (Gavrieli 1997;Sirota 2020), hindering salt precipitation and enabling limited salt dissolution at the shallow lake floor (Fig. 2b).
Concurrently, halite may continue to precipitate in the hypolimnion due to double diffusion processes . During the autumn and winter months the epilimnion cools and the thermocline becomes less pronounced, until the water column overturns and homogenizes. This leads to oversaturation with respect to halite, and salt precipitation and accumulation at all depths throughout the coming winter months (Gavrieli 1997; Arnon et al. 2016;Sirota 2020). The overall result of this annual cycle is reduced or no salt accumulation at the shallow lake floor within the epilimnion and metalimnion and uniform salt accumulation on the lake floor of the hypolimnion (Fig. 2). With ongoing lake-level drop, salt previously deposited below the thermocline, within the hypolimnion finds itself within the epilimnion, i.e. above the seasonal thermocline, and undergoes partial to full dissolution during the summer months, depending on its thickness and local conditions. Thus, at any given time during periods of lake-level drop and salt precipitation, the salt thickness on the shallower lake floor, within the epilimnion, is smaller compared to the maximal salt thickness found on the lake floor of the hypolimnion. Given the temporal and vertical development in the depth and shape of the thermocline (Fig. 2) and its geographic north-south location, the transition to the maximum salt thickness is expected to be found at the bottom of the metalimnion, at water depth of ~ 30 m. If similar conditions prevailed during the 'Sinkholes Salt' period, then the highly variable thickness of salt at top depths of 435 mbsl and above, and the more uniform salt thickness below top depth of ~ 440 mbsl suggest that at the end of the salt period, the metalimnion extended between 425 and 435 mbsl (Fig. 7a). This implies that the lowest lake stand at the end of the salt-precipitation period was at ~ 405 mbsl (Fig. 7). This minimum-elevation estimate is slightly above the reconstructed lake level compiled by At elevations between 405 and 435 mbsl (within and above the metalimnion), alternating seasonal salt precipitation and dissolution may result in variable salt thicknesses. However, the abrupt absence of salt above the 416 mbsl line, regardless of the local salt thickness in each site, suggests that it was truncated at this particular elevation throughout the lake. Here again, we draw an analogue with a similar process that occurred in the modern DS. The current massive halite precipitation from the DS began at the end of 1982 (Steinhorn 1983;Stiller and Sigg 1990;Gavrieli 1997) after the Dead Sea brine attained saturation with respect to halite. Since then DS water level has dropped by more than 35 m. However, in 1991/1992 a particularly rainy winter led to increased inflow of freshwater to the lake, resulting in a water-level rise of 2 m, and the formation of a temporary diluted and highly under-saturated surface layer (Anati and Stiller 1991;Beyth et al. 1993Beyth et al. , 1997. As a result, halite precipitation from the DS ceased for a few years and instead, halite was dissolved on the lake's floor down to a depth of ~ 20 m, the depth of the under-saturated water column in summer 1992 (Beyth et al. 1993). We suggest that the truncation of the 'Sinkholes Salt' layer at elevation of 416 mbsl is due to a similar process: increased inflow of freshwater which resulted in lake-level rise of a few meters, formation of a diluted epilimnion down to elevation of 416 mbsl and full dissolution of halite above this elevation by the diluted epilimnion water. This event also marks the switch from a monomictic regime to a meromictic regime, characterized by a stable and longterm stratification. Halite was preserved below 416 mbsl because in the coming years, as the lake maintained its stratification, it remained submerged in the halite-saturated brines. The rate of water-level rise and increased inflows volumes are likely to have been more gradual than the rates described above for 1991/2 since it involved no anthropogenic intervention such as opening of dams. The ~ 10 kyr dilution event also marks the end in the decline of the lake-water level that was responsible for the precipitation of the 'Sinkholes Salt'.  A similar dissolution-truncation surface and change in sedimentary sequence, albeit on a much larger scale was recently proposed to explain the intra-Messinian truncation surface at the top of the tilted salt layers of the Messinian Salinity Crisis in the eastern Mediterranean (Gvirtzman et al. 2017). Here, halite precipitation from evaporated seawater ceased as halite-under-saturated, yet still evaporated seawater, reached the Levant. The density of the halite-precipitating brine was clearly higher than that of the inflowing seawater, resulting in stratification of the Levant water column and dissolution of the halite down to the depth of the halocline. Following this event halite ceased to precipitate in the eastern Mediterranean and the sedimentary sequence up to the Messinian top erosional surface (TOS) consists mostly of anhydrite as well as detrital sands and shales (Gvirtzman et al. 2017).
Two regions, En Gedi and Zohar-Boqeq, show exceptional thickness values with respect to the pattern shown in Fig. 7. The En Gedi salt accumulated at bottom elevations of 440-465 mbsl, while the lake surface at the end of the period was at least 40 m above (Fig. 7). Thus, according to the abovementioned considerations its thickness is expected to be > 20 m. However, while two boreholes in En Gedi (M-1 and M-2) indeed encountered 20-23 m thick salt, the other three boreholes (EG-7, EG-13 and EG-20) show only 2-5 m thick salt (Table 1). There may be several possible explanations to the missing salt in En Gedi. The first is post-precipitation erosion. However, a close examination of all the available borehole logs shows that in all the western boreholes the salt is mantled by thin clay layers, and only one borehole in the east is mantled by coarse clastic sediments. Thus, erosion alone cannot explain these observations. A more likely scenario involves salt dissolution or partial precipitation due to contact with under-saturated water at the lake floor during the period of 'Sinkhole Salt' precipitation. Possible sources for such under-saturated water are submarine groundwater discharge (SGD; Judd and Hovland 2007), analogous to those described at depths of 10-30 m below the current DS surface (Ionescu et al. 2012;Munwes et al. 2020). Such seeping water will generate local rising columns of halite under-saturated water which may dissolve salt and/or hinder salt precipitation and result in significant variations in salt thickness over short distances. The currently high volumetric freshwater sources in the En Gedi region as well as the thick tufa outcrops in the area may indicate that this particular region was rich in fresh groundwater during the 'Sinkholes Salt' period as well, lending additional support to this explanation.
The anomalously high salt thickness at the Zohar-Boqeq boreholes, which are located in the southern DS basin, may reflect different limnological conditions between the deep northern basin and the shallow southern basin, where evaporation and its impact on the limited water volume may have been more extensive.
On the role of the 'Sinkhole Salt' elevation in future sinkhole formation along the DS Sinkholes along the DS are triggered by dissolution of the 'Sinkhole Salt' layer by under-saturated groundwater with respect to halite. The pathways of such water may be either lateral, invading the salt layer at its western boundary (Yechieli et al. 2006), upward along vertical faults (Abelson et al. 2006), or by drainage of floodwater in riverbed sinkholes (Avni et al. 2016). Salt dissolution by the first two mechanisms will continue as long as the groundwater level (or the hydraulic head of confined aquifers) remains within or above the salt layer. With ongoing drop of the DS level, the associated groundwater levels will drop below the salt layer, salt dissolution will decrease and new sinkholes will gradually stop forming by these processes. Three parameters determine where and when this transition will occur: (1) the elevation of the bottom of the salt layer, (2) the elevation of groundwater level or the hydraulic head at a specific time, and (3) the rate of groundwater-level drop. These parameters are thus important in evaluation of future sinkhole formation. It should, however, be noted that sinkholes will continue forming by drained floodwater even after groundwater drops below the salt layer. Thus, future assessment of sinkhole formation along the DS should consider the combination of all its governing processes. The database presented in this study will definitely contribute to such assessment.

Summary and conclusions
We analyzed the elevation and thickness variations of the ~ 10 kyr old 'Sinkholes Salt' in 40 boreholes along the western side of the DS. In the northern basin of the DS, these variations suggest that thermohaline stratification developed annually during the time of DS water-level decline that led to the deposition of the salt. This resulted in a relatively uniform salt layer that accumulated on the lake floor below the metalimnion, whereas salt dissolution in the summer months by the warm epilimnion resulted in the thinning of the shallower salt layer. By the end of the 'Sinkholes Salt' period, lake level dropped to an elevation of ~ 405 mbsl. This was followed by a change to a positive water balance of the lake, rise in water level, dilution of the surface water and the establishment of new long-term stratification which led to full dissolution and truncation of all salt above the halocline at 416 mbsl.
The relatively dense concentration of boreholes in the Ze'elim fan enables reconstruction of the local paleo-bathymetry of the lake in the near offshore. The salt layer is shallow and thin in the central part of the fan and thickens and deepens significantly towards its northeast and southwest margins, suggesting that the Ze'elim fan delta at the end of the Pleistocene-early Holocene had a similar morphology as at the present day. In Arugot fan (En Gedi), the elevation of the salt bottom reflects the local paleo-bathymetry, while local variations in the salt-layer thickness may be interpreted as the result of under-saturated submarine groundwater discharge (SGD) that impede salt precipitation in their close vicinity.
Overall, our results and analysis demonstrate how basic parameters, such as depth and thickness of a subsurface-salt layer, provide a wealth of information on paleolimnology, paleogeography, previous lake levels, and future sinkhole susceptibility. Such information is important along the DS and may be useful in other sinkhole environments globally.