GOCE Long-Wavelength Gravity Field Recovery from 1s-Sampled Kinematic Orbits Using the Acceleration Approach

The acceleration approach is an efficient and accurate tool for the estimation of the low-frequency part of GOCE (Gravity field and steady-state Ocean Circulation Explorer) gravity fields from GPS-based satellite-to-satellite tracking (SST). This approach is characterized by second-order numerical differentiation of the kinematic orbit. However, the application to GOCE-SST data, given with a 1s-sampling, showed that serious problems arise due to strong amplification of high frequency noise. In order to mitigate this problem, we developed a tailored processing strategy in a recent paper which makes use of an extended differentiation scheme acting as low-pass filter, and empirical covariance functions to account for the different precision of the components and the inter-epoch correlations caused by orbit computation and numerical differentiation. However, also a more “brute-force” strategy can be applied using the standard unextended differentiation scheme and data-weighting by error propagation of the provided orbit variance-covariance matrices (VCMs). It is shown that the direct differentiator shows a better approximation and the exploited method benefits from the stochastic information contained in the VCMs compared to the former strategy. A strong dependence on the maximum resolution, the arc-length and the method for data-weighting is observed, which requires careful selection of these parameters. By comparison with alternative GOCE hl-SST solutions we conclude that the acceleration approach is a competitive method for gravity field recovery from kinematic orbit information.