It's all about statistics: Global gravity field modeling from GOCE and complementary data

Since October 2009, ESA's dedicated satellite gravity mission GOCE (Gravity Field and Steady-State Ocean Circulation Explorer) observes the global gravity field of the Earth. The estimation of the model parameters from the original GOCE observations requires the application of tailored tools of geomathematics and statistics. One of the main constraints is to compute pure GOCE models, which are independent of any other external gravity field information. Up to now, four releases of global GOCE gravity field models have been computed and released. Their continuously increasing accuracy is validated by external gravity field information. A key prerequisite for achieving high-quality results is the correct stochastic modeling of all input data types in the frame of a least-squares adjustment procedure based on the rigorous solution of full normal equation systems. Together with the global gravity field models, parameterized as coefficients of a spherical harmonic series expansion, also the related error variance-covariance matrix is generated, which turns out to describe the true errors of the solutions very accurately. The fourth release achieves global geoid height accuracies of 3.5 cm and gravity anomaly accuracies below 1 mGal at a spatial wavelength of 100 km. Further improvements are expected, also because of the GOCE satellite's orbit lowering in its final mission phase, which will further improve the spatial resolution. In addition to these pure GOCE-only models, in the frame of the GOCO initiative consistent combined gravity field models are processed by including GRACE and SLR data (improving the long wavelengths), as well as terrestrial gravity information and satellite altimetry (improving the high-frequency component). Also for the computation of these optimum combinations, the tools developed for the GOCE processing can largely be applied. Numerous fields of application in geodesy, oceanography, and geophysics can benefit already now from the new GOCE models. As an example, the derivation of global ocean transport processes from a combination of satellite altimetry and global gravity information demonstrates that GOCE can contribute significantly to an improved understanding of processes in system Earth.