Historical reconstruction of the Aral Sea shrinking by a full 3-D wetting and drying model ECOSMO

Abstract. Several decades ago, the Aral Sea was the fourth out of the biggest lakes all over the world. Due to intense irrigation in the Aral Sea basin an imbalance in the water budget occurred and caused dramatic shrinking of the Aral Sea. With the present paper a sea-ice model for the Aral Sea has been developed, aiming in resolving the long-term development of the sea including drastic volume and surface area changes. The necessary requirement for such a model is a mass conservative wetting and drying scheme which has been developed in the application. By performing sensitivity experiments on different implementations of wetting and drying schemes, a significant influence of waggling on the volume and salt budget was identified for a hindcast under realistic forcing and the need for a volume and mass conserving scheme was demonstrated. Using a mass conserving implementation a model hindcast was performed for 1979–1993 by using the ECMWF re-analysis (ERA-15) meteorological data and modeled evaporation. Consideration of surface area changes in a temporal resolution of the model time step allowed for a realistic description of non-linear interactions between volume loss and evaporation changes. The sea-ice model was able to successfully simulate sea level and sea surface area decrease as well as sea-ice dynamics on a multiyear to decadal time scale. Model predicted evaporation was found to result in significantly improved volume budgets for the Aral Sea and in realistic predictions of the Aral Sea shrinking. Furthermore, by comparing the simulated average salinities to observed ones, it could be shown that the present model concept neglecting additional salt flux effects such as salt precipitation and salinity deposition on the dried sea bed, is adequate to reproduce observed mean salinities during the period of investigation.