Comparison of GRACE Data With Superconducting Gravimeter Data at Ny-Alesund in Svalbard, Norway.
The Svalbard archipelago, Norway, is affected by the post-glacial rebound subsequent to the last deglaciation. Moreover, the glaciers experience large thinning (approx. 50 cm/yr) due to present-day ice-melting. These two effects make Svalbard a very interesting zone to study and understand better the consequences of ice-melting at different time scales. We analyze eight years of gravity data obtained with a superconducting gravimeter (SG) installed at the Global Geodynamics Project station of Ny-Alesund. We compare the measured secular gravity variation with that obtained from the modeling of the gravity variation due to present-day ice-melting. We can then extract out a trend that can be associated with the post-glacial rebound. The SG time series are also confronted with GRACE time series. Finally, we discuss the correct use of the GRACE data for this spatial scale.
Constraints of GRACE on the Ice Model and Mantle Rheology in Glacial Isostatic Adjustment Modeling in North-America
GRACE satellite data offer homogeneous coverage of the area covered by the former Laurentide ice sheet. The secular gravity rate estimated from the GRACE data can therefore be used to constrain the ice loading history in Laurentide and, to a lesser extent, the mantle rheology in a GIA model. The objective of this presentation is to find a best fitting global ice model and use it to study how the ice model can be modified to fit a composite rheology, in which creep rates from a linear and non-linear rheology are added. This is useful because all the ice models constructed from GIA assume that mantle rheology is linear, but creep experiments on rocks show that nonlinear rheology may be the dominant mechanism in some parts of the mantle. We use CSR release 4 solutions from August 2002 to October 2008 with continental water storage effects removed by the GLDAS model and filtering with a destriping and Gaussian filter. The GIA model is a radially symmetric incompressible Maxwell Earth, with varying upper and lower mantle viscosity. Gravity rate misfit values are computed for with a range of viscosity values with the ICE-3G, ICE-4G and ICE-5G models. The best fit is shown for models with ICE-3G and ICE-4G, and the ICE-4G model is selected for computations with a so-called composite rheology. For the composite rheology, the Coupled Laplace Finite-Element Method is used to compute the GIA response of a spherical self-gravitating incompressible Maxwell Earth. The pre-stress exponent (A) derived from a uni- axial stress experiment is varied between 3.3 x 10-34/10-35/10-36 Pa-3s-1, the Newtonian viscosity η is varied between 1 and 3 x 1021 Pa-s, and the stress exponent is taken to be 3. Composite rheology in general results in geoid rates that are too small compared to GRACE observations. Therefore, simple modifications of the ICE-4G history are investigated by scaling ice heights or delaying glaciation. It is found that a delay in glaciation is a better way to adjust ice models for composite rheology as it increases geoid rates and improves sea level fit at some sites.
New Relative Sea-level Observations From the Northern Cascadia Subduction Zone and Implications for Cordilleran Ice Sheet History and Mantle Rheology
New data from Barkley Sound on the west coast of Vancouver Island, British Columbia, and from the mainland coast north of Vancouver add to the inventory of constraints on relative sea level during and after the collapse of the Cordilleran Ice Sheet (CIS). Previously published observations lie on a northwest-southeast profile along the east coast of Vancouver Island. The new sea-level observations extend existing data further back in time and create a southwest-northeast profile across Vancouver Island and the Strait of Georgia that is approximately perpendicular to the strike of the Cascadia Subduction Zone. The two profiles intersect in the central Strait of Georgia, where the sea level history was described by Hutchinson et al. (2004). At the western end of the study area, sea level in Barkley Sound dropped from a high-stand above 30 m elevation just before 15 000 cal BP to below the present shoreline around 14 500 cal BP, consistent with the timing of a previously documented, well-constrained low stand. Relative sea level in the central Strait of Georgia dropped from well over 150 m to about -15 m from 14 000 cal BP to 11 500 cal BP. Near Sechelt Inlet, on the mainland coast, sea level closely followed the mid-strait trend, dropping from a high-stand of at least 150 m just prior to 14 000 cal BP. Sometime between 12 500 and 11 000 cal BP, sea level near Sechelt fell past present levels to a poorly constrained low-stand between about 11 000 and 10 000 cal BP. The data are consistent with a previously noted trend of relatively early and deep low-stands on the periphery of the former ice sheet and later and shallower low-stands towards the interior. The mean rate of glacial isostatic rebound near Sechelt was similar to that of the mid-strait (about 11 cm/yr), while the rate was somewhat lower in Barkley Sound. The new data will refine existing models of the ice-sheet history during the rapid CIS deglaciation and may reveal spatial variations in shallow mantle viscosity related to the structure of the Cascadia Subduction Zone.
Influence of the Structural Dichotomy of Antarctic Lithosphere on Regional Glacial-Isostatic Adjustment
The strong structural dichotomy between East and West Antarctica is related to the West Antarctic Rift. The rheological implications are a reduction of the elastic-lithosphere thickness by a factor of more than 2 from East to West Antarctica as well as a strongly reduced mantle viscosity below West Antarctica and the Antarctic Peninsula. For modelling glacial-isostatic adjustment, we use a global viscoelastic earth model and apply the spectral finite-element method for the solution of the field equations. Ice models ICE-5G and IJ05 are used for parameterizing the last Pleistocene deglaciation. Lateral viscosity variations in the upper mantle are derived from variations in seismic velocity by applying scaling laws. Considering also lateral variations in the lithosphere structure, we study the implications of lateral variability on the glacial-isostatic adjustment of Antarctica.