Global Gravity Field Determination by Combination of terrestrial and Satellite Gravity Data

A multitude of impressive results document the success of the satellite gravity field mission GOCE with a wide field of applications in geodesy, geophysics and oceanography. The high performance of GOCE gravity field models can be further improved by combination with GRACE data, which is contributing the long wavelength signal content of the gravity field with very high accuracy. An example for such a consistent combination of satellite gravity data are the satellite-only models GOCO01S and GOCO02S. However, only the further combination with terrestrial and altimetric gravity data enables to expand gravity field models up to very high spherical harmonic degrees and thus to achieve a spatial resolution down to 20-30 km. First numerical studies for high-resolution global gravity field models combining GOCE, GRACE and terrestrial/altimetric data on basis of the DTU10 model have already been presented. Computations up to degree/order 600 based on full normal equations systems to preserve the full variance-covariance information, which results mainly from different weights of individual terrestrial/altimetric data sets, have been successfully performed. We could show that such large normal equations systems (degree/order 600 corresponds to a memory demand of almost 1TByte), representing an immense computational challenge as computation time and memory requirements put high demand on computational resources, can be handled. The DTU10 model includes gravity anomalies computed from the global model EGM08 in continental areas. Therefore, the main focus of this presentation lies on the computation of high-resolution combined gravity field models based on real terrestrial gravity anomaly data sets. This is a challenge due to the inconsistency of these data sets, including also systematic error components, but a further step to a real independent gravity field model. This contribution will present our recent developments and progress by using independent data sets at certain land areas, which are combined with DTU10 in the ocean areas, as well as satellite gravity data. Investigations have been made concerning the preparation and optimum weighting of the different data sources. The results, which should be a major step towards a GOCO-C model, will be validated using external gravity field data and by applying different validation methods.