Summary The seafloor topography (ST) of most sea areas is predicted using altimetry-derived gravity data. Traditional ST inversion methods suffer from two primary limitations: suboptimal accuracy in locally rugged areas due to linear approximation, and failure to fully leverage the advantages of gravity anomalies (GA) and vertical gravity gradient anomalies (VGG) at different wavelengths. In this study, we present a novel nonlinear inversion framework to address these two problems. The proposed approach introduces a new seafloor density contrast (SDC) that accounts for sediments for the derivation of nonlinear ST through forward modeling. The new method determines the optimal combination of different wavelengths of GA and VGG to recover the final ST. We apply the new method to invert the ST in a local area of the South China Sea (SCS). The results show that the nonlinear ST model has the highest standard deviation (STD), mean absolute percentage error (MAPE) and correlation coefficient of 51. 35 m, 0. 83 per cent and 99. 82 per cent among all ST models. The accuracy of the nonlinear ST model shows improvements of 2. 30 per cent and 3. 02 per cent relative to the two linear models derived from GA and VGG, respectively. Furthermore, it shows even greater improvements of 17. 93 per cent, 32. 24 per cent and 12. 39 per cent over the topo₂5. 1, GEBCO₂025, and SDUST2023BCO, respectively. The spectral analysis results show that the nonlinear ST model has higher energy at wavelengths less than 50 km, which demonstrates that nonlinear topography carries more short-wavelength information. Results from this study demonstrate that the nonlinear ST model can recover richer shortwave topographic features, and show that the nonlinear effects of ST are non-negligible in local rugged sea areas.
Wang et al. (Fri,) studied this question.