The aim of this paper is to present an alternative method that can be used to improve existing gravity field models via the application of gradient data from Gravity field and Ocean Circulation Explorer (GOCE). First, the proposed algorithm used to construct the observation equation is presented. Then methods for noise processing in both time and space domains aimed at reducing noises are introduced. As an example, the European Improved Gravity model of the Earth by New techniques (EIGEN5C) is modified with gradient observations over the whole lifetime of the GOCE, leading to a new gravity field model named as EGMGOCE (Earth Gravitational Model of GOCE). The results show that the cumulative geoid difference between EGMGOCE and EGM08 is reduced by 4 centimeters compared with that between EIGEN5C and Earth Gravitational Model 2008 (EGM08) up to 200 degrees. The large geoid differences between EGMGOCE and EIGEN5C mainly exist in Africa, South America, Antarctica and Himalaya, which indicates the contribution from GOCE. Compared to the newest GOCE gravity field model resolved by direct method from European Space Agency (ESA), the cumulative geoid difference is reduced by 7 centimeters up to 200 degrees. @en

Bathymetry is the measurement of the water depth similar to underwater topography. Gravity anomaly and vertical gravity gradient are important data used to inverse bathymetry. In this study, we investigate the sensitivities of the gravity anomalies and vertical gravity gradient data to bathymetry inversion. Firstly, the formulas of the gravity anomaly and vertical gravity gradient signals produced by the seabed terrain are derived, and then a series of numerical studies are designed to investigate the sensitivities. The results show that gravity anomalies and vertical gravity gradients have different sensitive bands to bathymetry inversion. Therefore, accuracy requirements for the gravity anomaly and vertical gravity gradient data are different for bathymetry inversion with different spatial resolutions and accuracies. With decreasing resolution, the accuracy requirement of the gravity anomaly gradually decreases, whereas the accuracy requirement of the vertical gravity gradient first decreases and then increases, indicating that the vertical gravity gradient is more sensitive to the short-wavelength signal than the gravity anomaly, especially in the 0–20 km wavelength ranges. In addition, we perform a theoretical analysis that shows that the vertical gravity gradient has a band that is most sensitive to the seafloor topography. This study is capable to provide a reference for the design of future satellites to retrieve the bathymetry.@en