Dead Sea Transform Research Articles

Challenges to reconstruct chert availability in tectonically highly modified environments: Examples from the Dead Sea Transform (Gesher Benot Ya’aqov, Wadi Hammeh, Greater Petra Region)

In the Near East, more particularly in the Southern Levant, the Rift Depression, also called the Dead Sea Transform (DST), reveals unique natural phenomena that have deeply marked the landscape, notably its topographic and hydrographic features, since the Miocene, about 5 million years ago (Mya).In this paper, we present four geo-archaeological case studies associated with the Central Jordan Valley (i.e. Gesher Benot Ya'aqov, Wadi al-Hammeh and the Greater Petra Region). These examples intend to illustrate the importance of taking the context and development of the geological substratum into account for the reconstruction of prehistoric lithic procurement strategies. Chert features and availability have been profoundly impacted by these diverse and complex processes. Continental and marine erosion, associated with plate tectonics and lateral plate displacements, have altered silicifications in their primary matrix and have also resulted in the accumulation of secondary deposits rich in chert nodules.Prehistoric peoples throughout the Pleistocene have extensively exploited these various sources. At Acheulian Gesher Benot Ya'aqov the presence of a non-local chert provided the opportunity to focus on and better understand the battered chert only available south of the Sea of Galilee, in the Central Jordan Valley (Case Study #1). The Natufian inhabitants of Wadi Hammeh 27 mainly exploited a yellow chert collected in secondary outcrops of limited geographical distribution that have vanished with the Holocene erosion (Case Study #2). In the Petra Region the lithic analysis of Neolithic assemblages was challenged by the presence of cherts of supposedly non-local origin. Lithic analysts working in the Dead Sea Rift should be aware that the chert spectrum of “the other side” of the depression might be different and that knowledge on local or regional chert availability cannot simply be transferred (Case Study #3). In the Wadi Sabra lateral and vertical displacements as well as tilting processes have affected primary chert-bearing formations resulting in heavily fissured and cracked cherts badly suitable for knapping. In contrast, the tectonically unaffected outcrops in the highlands show no internal damage of the material. This difference is clearly mirrored in the procurement behaviour of the Upper Palaeolithic groups occupying the wadi (Case Study #4).

Seismic Observations of Microearthquakes from the Masada Deep Borehole

AbstractA better understanding of the Dead Sea basin (DSB) tectonics and hazards can be achieved through high-sensitivity observations of its seismic activity. Accordingly, we have taken advantage of the Masada deep borehole (MDBI), an abandoned oil well on the western shore of the Dead Sea rift, to install a seismometer at 1516 m below sea level—a 2 Hz three-component seismometer. The MDBI station is near the east Masada fault, part of the western boundary fault of the DSB. Seismological measurements conducted at these depths can provide more information than possible with surface observations—in this case, recording 1491 high-quality seismograms as compared to 103 by the local national network. The MDBI seismograms also have significantly higher signal-to-noise ratio than the on-surface observations of nearby Israel Seismic Network (ISN) stations. This lowers the detection threshold down to well below M = 0 on the ISN scale. Many of these earthquakes, sometimes occurring in clusters, take place almost under MDBI, at depths of 10–25 km. Using cross-correlation techniques, we found several sequences of earthquake clusters underneath both the station and along the Dead Sea fault (DSF). The b-value obtained by the frequency–magnitude relationship of these events is somewhat higher than the value for the DSF area taken as a whole.

Early-to-mid Miocene erosion rates inferred from pre-Dead Sea rift Hazeva River fluvial chert pebbles using cosmogenic <sup>21</sup>Ne

Abstract. In this work, we utilize a novel application of cosmogenic 21Ne measurements in chert to compare exposure times measured in eroding surfaces in the central Jordanian Plateau with exposure times from chert pebbles transported by the Miocene Hazeva River. The Miocene Hazeva River was a large fluvial system (estimated catchment size > 100 000 km2) that drained the Arabian Plateau and Sinai Peninsula into the Mediterranean Sea during the early-to-mid Miocene. It was established after the rifting of the Red Sea uplifted the Arabian Plateau during the Oligocene. Following late-Miocene-to-early-Pliocene subsidence along the Dead Sea rift, the Hazeva drainage system was abandoned and dissected, resulting in new drainage divides on either side of the rift. We find modern erosion rates derived from cosmogenic 21Ne, 26Al, and 10Be in exposed in situ chert nodules to be extremely slow (between 2–4 mm kyr−1). Comparison between modern and paleo-erosion rates, measured in chert pebbles, is not straightforward, as cosmogenic 21Ne was acquired partly during bedrock erosion and partly during transport of these pebbles in the Hazeva River. However, 21Ne exposure times calculated in Miocene cherts are generally shorter (ranging between 0-0+59 and 242±113 kyr) compared to exposure times calculated in the currently eroding chert nodules presented here (269±49 and 378±76 kyr) and other chert surfaces currently eroding in hyperarid environments. Miocene exposure times are shorter even when considering that they account for bedrock erosion in addition to maintained transport along this large river. Shorter exposure times in Miocene cherts correspond to faster paleo-erosion rates, which we attribute to a combination of continuous surface uplift and significantly wetter climatic conditions during the early-to-mid Miocene.

A non-active fault within an active restraining bend: The case of the Hasbaya fault, Lebanon

The Hasbaya fault is a 50-km-long fault branch of the Dead Sea Transform Fault within the Lebanese Restraining Bend. It lies within the intersection zone of Mount Lebanon and Anti-Lebanon chains, along the western flank of Mount Hermon. It is little studied, its tectonic behavior is unknown, and its role among the other active fault branches of the restraining bend is the least defined. It was mapped as a discontinuous array of fault segments, with minor displacements and tectonically undisturbed basaltic cover. We studied the Hasbaya fault with detailed field mapping, and combining geomorphology, structural geology, and active tectonic investigations. The results reveal little and localized faulting along the entire length of the fault, with no signs of tectonic correlation between the drainage network evolution and the tectonic activity of the fault. The absence of active-tectonic surface features along the Hasbaya fault trend indicates that it does not manifest enough surface evidence to be classified as a tectonically active structure within the Lebanese Restraining Bend, and subsequently does not appear to be a major source of seismic hazard in the region. It may be related to the subsurface structure that controls the monocline that forms the western flank of Mount Hermon, or it may represent the remnant of an old suture zone between Mount Lebanon and Anti-Lebanon as a part of the regional compression imposed across the restraining bend. Our results shed light on the role of the non-active Hasbaya fault in a zone of active faults within a restraining bend, which may provide a case with broader implications on comparable tectonic settings worldwide.

Displacement of springs and changes in groundwater flow regime due to the extreme drop in adjacent lake levels: The Dead Sea rift

Lake-level fluctuations brought on by climatic changes and anthropogenic factors may affect the flow regime in adjacent aquifers that discharge toward those lakes. Such fluctuations may also cause displacement of springs that discharge these aquifers. Using a calibrated numerical model, an extreme example of such phenomenon is observed in the Dead Sea rift valley and the adjacent Eastern Mountain Aquifer of the Judea and Samaria Mountains. Lake levels along the Dead Sea rift have changed dramatically and rapidly by hundreds of meters, followed by changes in the lake areas by hundreds of square kilometers. Simultaneously, the aquifer exhibited spring displacements, both on large and small scales. Currently, 50% of the aquifer water discharge in the Zuqim zone, and 10% north of the Dead Sea. But in the past, only 30% discharged at Zuqim and 40% north of the Dead Sea. There is evidence for such an occurrence in the past, and it is likely to recur in the future, based on the predicted progressive decline of the current Dead Sea level. This may have an impact on wetland habitats along the coast.

Fragmentation of the Sinai Plate indicated by spatial variation in present-day slip rate along the Dead Sea Fault System

SUMMARYA comprehensive GPS velocity field along the Dead Sea Fault System (DSFS) provides new constraints on along-strike variations of near-transform crustal deformation along this plate boundary, and internal deformation of the Sinai and Arabian plates. In general, geodetically derived slip rates decrease northwards along the transform (5.0 ± 0.2 to 2.2 ± 0.5 mm yr−1) and are consistent with geological slip rates averaged over longer time periods. Localized reductions in slip rate occur where the Sinai Plate is in ∼N–S extension. Extension is confined to the Sinai side of the fault and is associated with prominent changes in transform geometry, and with NW–SE striking, left-lateral splay faults, including the Carmel Fault in Israel and the Roum Fault in Lebanon. The asymmetry of the extensional velocity gradients about the transform reflects active fragmentation of the Sinai Plate along the continental margin. Additionally, elastic block modelling of GPS velocities requires an additional structure off-shore the northern DSF segment, which may correspond with a fault located along the continental margin, suggested by prior geophysical studies.

An updated parametric catalog of historical earthquakes around the Dead Sea Transform Fault Zone

In this study, we present the work done to review the existing historical earthquake information of the Dead Sea Transform Fault Zone (DSTFZ). Several studies from various sources have been collected and reassessed, with the ultimate goal of creating of new homogenized parametric earthquake catalog for the region. We analyze 244 earthquakes between 31 BC and 1900, which are associated with the geographical buffer extending from 27 N to 36 N and from 31 E to 39 E. Of these, 93 were considered real seismic events with moment magnitude (Mw) greater than 5 that indeed occurred within this zone. While we relied on past parametric data and did not assign new macroseismic intensities, magnitude values, or epicenters for several controversial events, we did however resort to the primary sources to obtain a more critical perspective for the various assigned macroseismic intensities. In order to validate the derived parametric information, we tried to associate the events present in the historical records, with any evidence coming from past field investigations, i.e., geological or archaeological studies. Acknowledging the uneven quality and quantity of data reporting each event, we provided each entry with an uncertainty range estimate. Our catalog lists 33 events of Mw ≥ 6 absent from the latest published compilation with compatible time span and areal coverage. The whole catalog is considered complete down to Mw 7 and in certain areas down to Mw 6 after the year 1000, with majority of the larger earthquakes located in the part of DSTFZ, which extends from the southeast part of Dead Sea lake till Antioch.

Ancient earthquake’s ruins may signal future quakes near Dead Sea Fault

Ancient earthquake’s ruins may signal future quakes near Dead Sea Fault

A Comparative Study on the VS30 and N30 Based Seismic Site Classification in Kahramanmaras, Turkey

Assessment of seismic site classification (SSC) using either the average shear wave velocity (VS30) or the average SPT‐N values (N30) for upper 30 m in soils is the simplest method to carry out various studies including site response and soil‐structure interactions. Either the VS30‐ or the N30‐based SSC maps designed according to the National Earthquake Hazards Reduction Program (NEHRP) classification system are effectively used to predict possible locations for future seismic events. The main goal of this study is to generate maps using the Geographic Information System (GIS) for the SSC in Kahramanmaras city, influenced by both East Anatolian Fault and Dead Sea Fault Zones, using both VS30 and N30 values. The study also presents a series of GIS maps produced using the shear wave velocity (VS) and SPT‐N values at the depths of 5 m, 10 m, 15 m, 20 m, and 25 m. Furthermore, the study estimates the bed rock level and generates the SSC maps for the average VS values through overburden soils by using the NEHRP system. The VS30 maps categorize the study area mainly under class C and limited number of areas under classes B and D, whereas the N30 maps classify the study area mainly under class D. Both maps indicate that the soil classes in the study area are different to a high extent. Eventually, the GIS maps complied for the purpose of urban development may be utilized effectively by engineers in the field.

Impact of the Dead Sea Transform Kinematics on Adjacent Volcanic Activity

AbstractThe Dead Sea Transform (DST) cuts through the northwestern edge of the Harrat ash Shaam volcanic field of western Arabia. The relation of the DST to adjacent volcanic activity has varied during the last 5 Ma. While Pliocene volcanism erupted on both shoulders of the DST and thick sections of volcanics accumulated along its axis, Pleistocene volcanic activity gradually abandoned the transform area while continued farther east. This was concomitant with the increase of transform‐normal (E‐W) contraction along the adjacent transpressive Metulla block at the DST northern sector. New geochronology of Pleistocene volcanic rocks from the Golan Heights (Ar‐Ar) and of syntectonic and post‐tectonic calcite precipitates (U‐Th) from Metulla block reveals temporal association of volcanic dormancy and intensity of the transpressive regime. This includes volcanic cessations during 600–220 ka and post 100 ka, which coincided with growth episodes of several ~E‐W striking veins within the adjacent Metulla block. New high‐resolution U‐Th ages of calcite‐filled vein record the exact timing of these growth episodes to 340–240 and 53–16 ka, which is in accordance with the renewed activity along N‐S strike‐slip faults during 220–60 ka, manifested by U‐Th ages of calcite precipitates on striated fault planes. The temporal association of volcanic dormancy and enhanced transpression, while volcanic activity and increased strike‐slip activity, suggests that continental transforms may act both as inhibitors and as facilitators to magmatic activity. Switching between intermediate and least principal stresses might be the underline mechanism for stress dissipation and relaxation at a time scale of 100 ka.

Palynological correlation of the Arqov and Saad formations of the Negev, Israel, with the Umm Irna Formation of the eastern Dead Sea, Jordan

Palynological study of the Arqov and Saad formations of the Negev, Israel, in Avdat-1 borehole has allowed comparison with the assemblages of the Umm Irna Formation of the Dead Sea, Jordan. Core 7 of Avdat-1 (Saad Formation) contains common Hamiapollenites dettmannae and Distriatites insolitus indicating that it correlates with the Arabian Peninsula OSPZ5 Biozone and therefore has a likely Roadian–Wordian age, rather than a Westphalian age as previously suggested. Core 6 contains Indotriradites mundus, whose first appearance indicates the base of OSPZ6 (Wordian, extending into the Capitanian), but also contains common Protohaploxypinus uttingii, Pretricolpipollenites bharadwajii and Thymospora spp. As such, the assemblages of Core 6 are most similar to those of the Umm Irna Formation, while those of Core 7 are older, as implied by the correlation of Core 7 with OSPZ5. An interesting difference between typical OSPZ6 assemblages and those of Avdat-1 Core 6 is the common presence of Falcisporites stabilis in the latter. This pollen was produced by the corystosperm plant Dicroidium and its presence in Core 6 further indirectly confirms previous suggestions that Dicroidium (once considered confined to the Triassic) existed in the pre-Triassic in the Middle East. The difference in thickness in Permian sediments between Avdat-1, eastern Negev boreholes, and the Umm Irna Formation may be due to greater accommodation space provided by a southwestward extension of the Palmyrid depocenter, or perhaps subsidence related to a fault in a similar position to the present Dead Sea Fault, or a fault ancestral to the Dead Sea Fault.

Surface deformation processes in the Carmel Fault based on 17 years of GPS measurements

Tectonic activity and crustal deformation in northern Israel are mainly related to the Dead Sea Fault (DSF) and the Carmel–Gilboa Fault System (CGFS). The CGFS is composed of several NW–SE trending faults while the main faults are the Carmel Fault (CF) and Gilboa Fault (GF). The CGFS divides the Sinai Sub-Plate into two tectonic domains. In this study, we geodetically investigate surface deformation processes in the Carmel Fault region. Beside the processing and analysis of GPS measurements, we highlight geodetic aspects in the process of deformation analysis in geodetic monitoring networks. We implement the Extended Free Network Adjustment Constraints solution to calculate the velocities of 24 sites that were measured eight times between 1999 and 2016 using Global Positioning System (GPS). The regional site velocities were estimated with respect to a local datum that was defined by a stable cluster of sites on one side of the fault. We introduced the site velocities into the estimation of surface deformation parameters by using affine transformation also with respect to a local datum. The coordinates of network sites can be transformed to any desired datum by using extended similarity transformation. Examination of the velocity field in relation to a datum defined by points in the Galilee region raises the suggestion that the velocities in the Yizre’el Valley region are due to activities along the GF or similar trending faults on the northern side of the valley which are halted by the Tivon Hills. The best set of deformation parameters, the one which better describes the velocity field, was determined by the second-order Akaike Information Criterion (AICc). The results show significant sinistral deformations of less than 1 mm/year along the Carmel Fault accompanied with extensions and shear strain.

Seismic Activity and Energy Release of Earthquakes along the Gulf of Aqaba, Egypt

The Gulf of Aqaba is situated along the southern part of the Dead Sea Rift Area transform (DST), 1000 km (620 miles), the boundary between the African plate and the Arabian plate. It is situated toward the east of the Sinai landmass and west of Saudi Arabia. It is one of the joints interfacing the Asian and African landmasses. The Gulf expands 180 km from Eilat and Aqaba and joins the Red Sea in the Strait of Tiran, with the most extensive purpose of 28 km. On November 22, 1995, the biggest seismic tremor estimating 7.0 degrees happened along parts of the dead left line in the Gulf of Aqaba close to the port of Nuweiba in Egypt. This tremor made extraordinary harm Egypt and neighboring nations. Consequently, a few point by point investigations of seismology and geophysics have been led in the Gulf of Aqaba. In the territory researched, the vast majority of the ongoing seismic action revolved around the Gulf of Aqaba is portrayed by the contribution of moderate central quakes, an immediate aftereffect of the relative development between the plates of Africa, Sinai and the Arabian Peninsula. Seismic Energy along the Gulf of Aqaba locale is assessed utilizing a brief period (50 tests for every second) executed by the Egyptian National Seismic Network amid the period 1998‒2017. The connection between the dispatch and greatness of quakes in the Gulf of Aqaba demonstrated that the vast majority of the seismic Energy discharge originates from the biggest occasions and is packed in the inside, while it is moderately low in the southern and northern parts of the Gulf.

Temporal and Spatial Patterns of Seismic Activity Associated with the Dead Sea Transform (DST) during the Past 3000 Yr

AbstractHistorical reports of earthquakes occurring before the twentieth century along the Dead Sea Transform (DST) are available for the past 3000 yr. Most of them are organized in various catalogs, reappraisals, and lists. Using a comprehensive and consistent compilation of these reports, the historical seismicity associated with the DST as a complete tectonic unit was examined. The compilation, supported by paleoseismic and archeoseismic evidence, resulted in 174 reliable historical earthquakes and 112 doubtful ones. The reliable earthquakes, along with 42 post‐nineteenth century instrumental earthquakes, are an up‐to‐date evaluation of the DST seismicity starting from the mid‐eighth century B.C.E. until 2015 C.E. Additionally, the scenario of historical earthquakes such as the 363 C.E. and 1033 C.E. events was resolved. The characterization of temporal and spatial patterns of DST seismicity, classifying them into four geographical zones, raised that most of the northern destructive earthquakes are clustered while clustering at the central and southern zones is less abundant.

TerraceM-2: A Matlab® Interface for Mapping and Modeling Marine and Lacustrine Terraces

The morphology of marine and lacustrine terraces has been largely used to measure past sea- and lake-level positions and estimate vertical deformation in a wealth of studies focused on climate and tectonic processes. To obtain accurate morphometric assessments of terrace morphology we present TerraceM-2, an improved version of our Matlab® graphic-user interface that provides new methodologies for morphometric analyses as well as landscape evolution and fault-dislocation modelling. The new version includes novel routines to map the elevation and spatial distribution of terraces, to model their formation and evolution, and to estimate fault-slip rates from terrace deformation patterns. TerraceM-2 has significantly improves its processing speed and mapping capabilities, and includes separate functions for developing customised workflows beyond the graphic-user interface. We illustrate these new mapping and modelling capabilities with three examples: mapping lacustrine shorelines in the Dead Sea to estimate deformation across the Dead Sea Fault, landscape evolution modelling to estimate a history of uplift rates in southern Peru, and dislocation modelling of deformed marine terraces in California. These examples also illustrate the need to use topographic data of different resolutions. The new modelling and mapping routines of TerraceM-2 highlight the advantages of an integrated joint mapping and modelling approach to improve the efficiency and precision of coastal terrace metrics in both marine and lacustrine environments.

BUILDING COLLABORATIONS WITH BIOBANKS AND HARNESSING ELECTRONIC HEALTH RECORD DATA FOR PSYCHIATRIC GENOMICS RESEARCH

Paleoseismology, archaeoseismology, and paleotsunami studies are disciplines of modern seismology and active tectonics, that open the window to past earthquakes; earthquakes that occurred before seismometers were available and which possibly were not mentioned in any written sources. Field investigation techniques and new laboratory and modeling approaches have significantly advanced these disciplines in the past two decades. The methods applied require the trace of permanent deformation at the Earth surface or in man-made structures caused by ground movements of past earthquakes. The methods include the identification of active faults, trench excavations, shallow geophysical prospecting, coring techniques, laser scanning and photogrammetry, virtual models of structural behavior, advanced archeological techniques, and different kinds of physical dating of relevant samples. In this article, three selected examples of active faulting, i.e., the North Anatolian Fault, the Himalaya Thrust Front and the Dead Sea Fault are illustrated by published results of recent case-studies with dating of the record of past earthquakes and their successive occurrence through space and time. Catastrophic inundations on coastal regions following tsunamigenic earthquakes are preserved in the succession of deposits considered as geological archives. Earthquake damage to structures at archeological sites document the effects of seismic shaking and case specific models provide some constraint on ground motion parameters. The application of the new approaches in the study of past earthquakes has become an essential part of the seismic hazard and risk assessment.

The Bisri dam project: A dam on the seismogenic Roum fault, Lebanon

The Roum fault is a fault branch of the Dead Sea Transform Fault within the Lebanese Restraining Bend. It is an active seismogenic fault that was associated with the 1 January 1837 (Ms 7.1) earthquake, and whose extent has been the subject of great controversy within the restraining bend. In this paper, the Roum fault is revisited after a dam project (the Bisri dam) has been planned in one of the fault's most critical locations, i.e. in the Bisri valley that hosted the epicenter of the 16 March 1956 (ML 5.8) earthquake. The fault and its associated structures in the proposed dam area are discussed, and highlights on the dam site are made as a geologically complex and tectonically active area. The future water body behind the planned Bisri dam will overlie an active fault, with huge potential for fault movement underneath the dam footprint, liquefaction around the dam foundations, mass movements near the dam site, and direct effect on inducing seismicity on a delicate fault system that can potentially trigger a major earthquake that can affect the whole Lebanon and its surroundings. This requires special attention and in-depth considerations especially of the potential of reservoir-induced/triggered seismicity on the seismogenic Roum fault, and makes of the Bisri area a very unfavorable site for the proposed Bisri dam in Lebanon.

Evolution of the stress field near the Arava basin located along the Dead Sea Fault system as revealed by joint sets

To reveal the development of the stress field near the Arava basin located along the Dead Sea Fault system, we characterized joint sets that were exposed mainly in the Santonian and Eocene chalk rock outcropping in the study area.In total, ~2600 strike orientations of single-layer joints (SLJ) and multilayer joints (MLJ) mirrored four main stress fields: 1) σHNW(J1-326°), 2) σHNNW(J2-343°), 3) σHNNE(J3-033°), and 4) σHN (J4-360°).The results showed that the relatively lower values of the fracture spacing ratio (FSR) and their standard deviations in J1 implied that J1 was formed during the burial stage of the Eocene chalk formations. The inconsistency between the statistical parameters of the log-normal distributions of joint spacing and FSR raised doubt about the applicability of the “saturation joint density model” in our studied area.The joint sets characteristics suggest that the regional σHNW operated during the Eocene. During the Miocene, the regional stress field changed clockwise to σHNNW, probably in association with the early stage of transform motion along the Dead Sea Fault. In the Pliocene-Pleistocene, the σHNNE and σHN local stress fields both operated and were accompanied by a tensional stress field that can be identified in the intensive normal faulting evolved along the western Arava basin margins.

Erq el Ahmar Elephant Site – A mammoth skeleton at a rare and controversial Plio-Pleistocene site along the mammal migration route out of Africa

Early sites along the Dead Sea Transform (southern Levant), among them the Erq el Ahmar Elephant Site, are key points in understanding hominin and mammal migration out of Africa and into Eurasia. The late Prof. Tchernov had begun an intensive campaign to expose the faunal remains at the site, but unfortunately was unable to conclude his study. Based on interim reports and geomorphological descriptions, we were aware of numerous elephant remains found and left in situ. The Erq el Ahmar Elephant Site is a controversial site. There are those who see it as the earliest Pleistocene hominin site in the area, while others consider it a paleontological site without any hominin involvement. We returned to the site to try to resolve this controversy. In a systematic excavation, we succeeded in exposing the previously uncovered elements, exposed more material and currently better understand the deposition sequence. However, the task was very challenging, since the skeletal elements were very fragile and required careful exposure and conservation, both in situ and in the laboratory, before they could be studied. A series of elements were found partially superimposed. Several elements of the skull, an almost complete tusk, vertebrae, ribs, a scapula and limb bones were found. Mammoth diagnostic traits were identified in the teeth and tusk. However, very few skeletons of early mammoths are known from the region. Have we exposed the most complete Mammuthus rumanus skeleton? Tooth microwear indicates leaf-browsing dietary traits, similar to that of other M. rumanus of this period. In addition, the recent excavations have revealed the potential of the site in understanding the evolution and dispersal of proboscidean species out of Africa during the Plio-Pleistocene, adding another focal point to the southern Levant along this route.

Paleoclimatology of the Levant from Zalmon Cave speleothems, the northern Jordan Valley, Israel

The Levant region of Israel, located along the Eastern Mediterranean Sea, is characterized by Mediterranean climate, but sixty km eastwards towards the Dead Sea Rift Valley (DSRV), the region rapidly becomes a ‘rain shadowʼ desert. Speleothems from the Mediterranean climate zone continuously grew throughout several glacial/interglacial cycles, indicating that water was always available in the unsaturated zone. Their δ18O variations match global and regional climate changes, with the Eastern Mediterranean Sea being the major control on their δ18O values, as evident from the similarity between the speleothems and the planktonic foraminifer G. ruber records. In contrast, speleothems along the central and southern segments of the DSRV grew mainly during last glacial periods coinciding with the higher stands of Lake Lisan, the precursor of the present-day Dead Sea. This paper discusses the newly discovered Zalmon Cave speleothems, located in the northern segment of the DSRV shedding light on the hydrological conditions along the rift during last glacial. Unlike speleothems located further south along the DSRV, Zalmon Cave speleothems grew both during glacial and interglacial periods. However, during last glacial their δ18O values, shifted to lower values by ∼1–2‰ relative to speleothems from central Israel coinciding also with the largest difference in δ18O values between Zalmon Cave speleothems and the marine record. The data suggest that a change of the westerlies' storm tracks occurred during most of the last glacial period, which resulted in increased rainfall in the northern DSRV providing freshwater input during the otherwise relatively ‘drierʼ glacial.