De-Leakaging and De-Aliasing of GRACE-based Surface Mass Distributions by Regularized Basis Functions and Additional Geometric Information

In this study, a method is developed which allows to reduce the leakage effect in GRACE gravity fields The leakage effect means that mass changes are falsely assigned to nearby regions due to the limited resolution of the gravity field. These assignments are partly physically unreasonable.

In our method, we separate the entire Earth’s surface according to their surface type as e.g. sea and land surface. Across these regions, individual basis functions (e.g. Slepian functions) are applied. In a constrained least squares adjustment, a priori surface mass trends are rearranged to the respective regions while the basis functions’ coefficients are constrained differently according to their surface type. Therefore, we assume different variabilities of possible mass change. With the inclusion of a filter matrix, the resulting field of surface mass changes is linked to the a priori distribution which results directly from the input gravity field.

The procedure is tested for unfiltered and DDK-filtered GRACE gravity fields as well as SLR-based gravity fields.  Furthermore, geometric information on the changing sea level is introduced in order to improve the de-leakaging and de-aliasing of the input GRACE distributions of surface mass change. For this purpose, products of ESA’s marine service are applied. In our study, we want to show benefits and applications concerning ice mass estimation.