Regional gravity field recovery from GRACE using position optimized radial base functions

Global gravity solutions are generally influenced by degenerating effects such as insufficient spatial sampling and background models among others. Local irregularities in data supply can only be overcome by splitting the solution in a global reference and a local residual part. This research aims at the creation of a framework for the derivation of a local and regional gravity field solution utilizing the so-called line-of-sight gradiometry in a GRACE-scenario connected to a set of rapidly decaying base functions. In the usual approach, the latter are centered on a regular grid and only the scale parameter is estimated. The resulting poor condition of the normal matrix is counteracted by regularization. By contrast, here the positions as well as the shape of the base functions are additionally subject to the estimation process. As a consequence, the number of base functions can be minimized. The analysis of the residual observations by local base functions enables the resolution of details in the gravity field which are not contained in the global spherical harmonic solution. The methodology is tested using simulated as well as real GRACE data.