BRIDGING EVAPORITES AND CLIMATE: TOWARDS A GENERAL MODEL FOR HYPERSALINE LAKES

Hypersaline lakes are usually found in depressions, where inflows and their solutes build up. In these closed settings, the only exits for water and ions are evaporation and mineral precipitation, respectively. The dependence on inflow and evaporation makes hypersaline lakes highly sensitive to hydrological and climatic fluctuations. In turn, the precipitation of saline minerals provides a detailed record of the limnological variables that respond to these fluctuations. In particular, during their growth, minerals trap microdroplets of lake water from which, thanks to a series of recent methodological breakthroughs, it is now possible to access the lake temperature, chemical composition, density, and changes in volume. However, interpreting these lacustrine variables in terms of hydrological and climatic fluctuations remains a challenge. Here, we present a new numerical lake model that aims to fill this gap. The model is based on the SIMSTRAT code, a 1D hydrodynamic model for freshwater lakes. Using climatic, hydrological and bathymetric parameters as an input, the model simulates accurate timeseries of the surface heat fluxes and of the vertical profile of lake temperature and density stratification. In this work, we extend the SIMSTRAT code by adding the necessary ingredients for the accurate prediction of the limnological variables in halite-saturated lakes. This includes an equation of state covering a large range of salinities, a surface water activity-dependent evaporation equation, and a physically realistic code for the precipitation of halite. Tests on the Dead Sea for the period 1979-2015 yield excellent results, with the deviation between simulated and monitored hypolimnion temperature, thermocline depth, lake level, salinity and halite sedimentation rate within a few percent of the total variability. The code, albeit restricted to the precipitation of halite, is designed for easy addition of other minerals in future developments. It thus constitutes a first step towards a general model for hypersaline lakes. The code will be open access.