Inverting Vertical Gravity Anomaly Gradients Using Multidirectional Data from Mean Sea Surface Model: A Case of Arabian Sea

Abstract Advancements in satellite altimetry have significantly enhanced high-resolution mean sea surface (MSS) models, enabling the computation of high-resolution vertical gravity anomaly gradient (VGAG) models. This study focused on the methodology for computing VGAG models using MSS models, introducing innovative improvements to established techniques. Using the SDUST2020 MSS model within the Arabian Sea research area, the DTU22 and CNES-CLS22 mean dynamic topography (MDT) models, and the XGM2019e₂159 Earth gravity field model for the remove-restore process, the short-wavelength geoid was derived. To harness the extensive marine gravity field information within the MSS model, the study considered the complex marine environment and calculated the second-order derivatives of the geoid in multiple directions. These derivatives were then used to determine their north-south and east-west components through the least squares method, resulting in the computation of the short-wavelength vertical gravity anomaly gradient. By restoring the long-wavelength vertical gravity anomaly gradient, a VGAG model for the study area was established. Finally, the results were analyzed using the SIO V32. 1 vertical gravity anomaly gradient model. Experimental results demonstrated that this approach effectively extracted marine gravity field information from the MSS model using multidirectional data, mitigating the amplification of geoid uncertainties caused by second-order derivatives.