Glacial-isostatic adjustment and the viscosity structure underlying the Vatnajokull Ice Cap, Iceland

We examine the dependence of glacial-isostatic adjustment (GIA) due to changes in the Vatnajokull Ice Cap, Iceland, on the underlying viscosity structure. Iceland offers a unique case study for GIA research, with a thinner elastic lithosphere underlain by a low-viscosity zone or asthenosphere, as opposed to regions such as Fennoscandia or North America described by a thicker lithosphere, while not necessarily featuring an asthenosphere. A laterally homogeneous spherical earth model is used consisting of an elastic lithosphere, a viscoelastic asthenosphere, a viscoelastic upper and lower mantle and a fluid core. We examine the response of the earth model to three ice models with circular plans and cross-section profiles based on the assumption of perfectly plastic material, but with different load histories. These are: (1) A history where the ice cap grows from a AD 900 minimum to a maximum at 1890, followed by a uniform decrease until 1991, continuing to the present day at an average rate based on recent mass-balance measurements, (2) a history that is the same as the first, except for constant ice volumes prior to 1890, and (3) a history that is again the same as the first model, except that the post-1991 changes correspond to the measured mass-balance values. We first compare the response to each ice model using typical earth-model parameters for Iceland presented in the literature. We then undertake a parameter-space search, where we assess the importance of lithosphere thickness, asthenosphere viscosity and basal asthenosphere depth, to predicted vertical-displacement rates, and compare them to rates determined from GPS measurements obtained from campaigns conducted between 1991 and 1999. The earth-viscosity structure that provides the optimum predictions with respect to the GPS-derived vertical-displacement rates consists of an elastic lithosphere with a thickness of between 20 and 30 km, an asthenosphere viscosity between 1 and 2 x 10(18) Pa s, and a basal asthenosphere depth between 250 km and possibly greater than 400 km. We find that the very low asthenosphere viscosity values of ca. 10(17) Pa s sometimes suggested in the literature are not necessary to account for the rapid vertical-displacement rates observed, which are the result of the contemporary decrease in the mass of the ice cap not considered previously.