RECONSTRUCTING PRECIPITATION AND WATER SOURCES TO THE DEAD SEA BASED ON SEDIMENTS AND THEIR GEOCHEMISTRY

The ICDP Dead Sea Deep Drilling Project recovered a continuous ~460 m record covering the past ~220,000 years from the deepest basin of the Dead Sea. Its watershed includes the Mediterranean climate zone and the Saharan-Arabian desert belt and its sediment facies reflect the climate. The sediments are comprised of mainly aragonite, gypsum, halite and detritus, reflecting changes in the fresh water runoff into the lake. The wettest intervals are represented by alternating aragonite and detritus layers and occur during glacials when lake level was high. The driest intervals are represented by thicker halite-rich layers that were deposited during interglacials when lake level was low. The total thickness of the halite, together with changes in the lake chemistry recorded by pore waters and halite fluid inclusions, were used to reconstruct the initial and final lake levels and runoff during the driest climate intervals. We have generated a high-resolution lake level history, based on the thicknesses of halite layers and the concentration of aragonite, interpreted from XRF scanning of the cores. Because the Dead Sea brine is enriched in Ca²⁺ and depleted in HCO3^-, aragonite precipitates when fresh water containing HCO3^- enters into the lake. From the composition of the brine in the halite fluid inclusions, we estimate that 14-16 cm of halite precipitates per 1 m lake level drop and therefore individual halite layers can be used to calculate lake level changes. The relationship between the XRF chemistry and the fresh water discharge was obtained empirically after calibrating the XRF counts to the lake level curve. In addition to the average changes in runoff, ²³⁴U/²³⁸U ratios in the lake, reflected in the authigenic minerals (aragonite, gypsum and halite), document changes in the water sources to the lake during its driest periods. ²³⁴U/²³⁸U activity ratios decrease from typical values of ~1.5 (reflecting the main water source – the Jordan River) to values down to ~1.0 during the driest interval at the end of MIS 5e at ~118-116 ka. These changes reflect major shifts in the lake’s hydroclimate whereby the inflow of Jordan River water became insignificant and eastern and southern water sources from the usually dry sections of the watershed became wetter.