In this study we apply geometric and gravimetric observations from various Earth observation satellites in order to estimate the variability in a lake with respect to its geometrical extent and water storage. Our test case is the Aral Sea, located in the arid zone of central Asia. Due to the diversion of its primary inlet rivers for irrigation purposes the lake suffered a devastating decline until its south eastern part had almost dried out in 2009. The study is focused on the period of the satellite gravity field mission GRACE from 2002 onwards. We present the change of the lake's surface extent based on optical remote sensing data from Landsat images that were analyzed for spring and autumn each year. Height variations of the lake surface were computed from multi-mission satellite altimetry. Both the surface extent and the water stage of the lake reached an absolute minimum in autumn 2009. However in 2010 a clear reversal of the negative trend of the previous years is visible. A geometrical intersection of the water level with a digital elevation model allows for estimating water volume changes. The resulting volume changes are subsequently analyzed with respect to satellite-based estimates of mass variations observed by GRACE. The results reveal that water storage variations in the Aral Sea are indeed the principal contributor to the GRACE signal of mass variations in this region. The different observations from all missions agree very well with respect to their temporal behavior.