Iterative data assimilation approach for the refinement of marine geoid models using sea surface height and dynamic topography datasets

The modelling errors of marine geoid models may reach up to a few decimetres in the shorter wavelength spectrum due to gravity data void areas and/or inaccurate data. Various data acquisition methods can provide sea surface heights more accurately. Similarly, hydrodynamic model data in conjunction with tide gauge readings allow the derivation of reliable dynamic topography. Geometrical marine geoid heights, independent of the usual gravity-based marine geoid models, can be obtained by removing the estimated dynamic topography from sea surface height measurements. This study exploits such geometry information to refine marine geoid models. A data assimilation approach was developed that iteratively combines sea surface height and dynamic topography datasets with an initial gravimetric geoid model. A case study is presented using sea surface heights from shipborne GNSS campaigns and an airborne laser scanning survey for refining the EIGEN-6C4 global geopotential model. Comparisons with a high-resolution regional marine geoid model reveal that the initial discrepancies of up to around two decimetres reduce to sub-decimetre within the study area. It is concluded that the developed iterative data assimilation approach can significantly improve the accuracy of marine geoid models, especially in regions where gravity data are of poor quality or unavailable.