Loading effects in Siberia and Northern Canada determined from a combination of hydrological model data and GRACE gravity field observations

Mass redistributions in various components of the Earth system, such as the atmosphere and the hydrosphere, exert time variable surface loads on the solid Earth that lead to temporal variations of its gravity field and its geometry. Recent satellite gravity field missions and modern space geodetic techniques allow for the observation of such effects with high accuracy. In order to interpret the observations geophysically, independent approaches from theory and modeling are required. Atmospheric effects can be described well by models and reanalysis data; they are commonly removed from the gravity field variations as seen by GRACE. But as hydrologic processes are far less understood, a clear discrepancy between respective model estimates and residual gravity field variations is evident, especially in regions that feature large seasonal variations. In our study, monthly fields of water storage and snow coverage from the Land Dynamics Model (LaD) will be compared with monthly gravity field solutions of GRACE for the period between 2003 and 2005. We focus on the northern landmasses of Siberia and Northern Canada, which are characterized by strong melting effects of snow that may be underestimated by LaD so far. From the comparison of the model results with observations, corrections for the LaD model are assessed. In order to validate the improved hydrological model with independent data, the resulting water storage values are converted into surface pressure variations that are subsequently superposed with atmospheric surface pressure variations from NCEP reanalyses. From the resulting pressure fields, load deformations of the Earth's crust are computed. For this purpose, an algorithm is applied which accounts for the regional discrepancies of the rheological properties of the crust due to its composition of heterogeneous materials with differing structures and densities. This procedure has already been shown to be more effective than the traditional approach in which the loading effects on crustal deformations are computed by means of a weighting function based on site-independent load Love numbers (Green's function). The resulting vertical displacements are finally compared with time series of GPS sites of the IGS network in the respective regions.