Radial basis function method for refining regional gravity field from satellite gravity gradient

Aiming at the problem that global gravity field model constructed by spherical harmonic function ignores the differences of local gravity signals, a radial basis function method and data processing scheme were presented for refining the regional gravity field based on satellite gravity gradient data．The global gravity field joint inverted by GOCE (gravity field and steady-state ocean circulation explorer) satellite gravity gradients and ITG-GRACE2010S normal equation was used as the background model．The low-frequency errors and colored noise of gravity gradient data were suppressed by the cascaded filtering method with frequency domain weighting．Based on the spherical splines function and sub-region regularization method, the regional gravity field models matching for local characteristics of gravity signals were constructed in the North America． To verify the effectiveness of our method, the high-precision gravity field model GOCO06S and GPS (global positioning system) leveling data were used for validation．The results show that the refined regional gravity field model has higher accuracy than the global gravity field model solved by the same period observations, and the inversion accuracy of the new preprocessed GOCE data is better than that of old version． © 2022 Huazhong University of Science and Technology. All rights reserved.