Improving the Reconstruction Accuracy of Marine Gravity Anomaly Encrypted Reference Map Using the New Mean Sea Surface 3-D Correction Method

Abstract

The reconstruction methods of marine gravity anomaly encrypted reference map are studied in this study, which can improve the accuracy of underwater gravity matching navigation. Firstly, the mean sea surface as the third dimension component was introduced into the semivariance function of the traditional Kriging two-dimensional (2-D) interpolation method for the first time. After that, a new mean sea surface three-dimensional (3-D) correction method was proposed in order to improve the spatial resolution of the marine gravity anomaly reference map and minimize the loss of accuracy. Secondly, encrypting the spatial resolution of the marine gravity anomaly reference map from 4′×4′, 3′×3′ and 2′×2′ to 1′×1′ as an example, respectively, under the same conditions, the reconstruction accuracy of the new proposed method improved 62.25%, 140.92% and 104.93%, respectively. Thus, compared with the traditional Kriging 2-D interpolation method, we verified that the new proposed method had better reconstruction accuracy. Thirdly, based on the proposed new method, the spatial resolution of the marine gravity anomaly reference map was firstly encrypted from 1′×1′ to 0.25′×0.25′. After that, transforming the spatial resolution of 0.25′×0.25′ into 1′×1′, compared with the original model data of 1′×1′, its the final RMSE is 4.143×10-1 mGal, which further verified the validity and reliability of the new mean sea surface 3-D correction method in application. INDEX TERMS the new mean sea surface 3-D correction method, marine gravity anomaly encrypted reference map, reconstruction accuracy, semivariance function.

An Improved Coastal Marine Gravity Field Based on the Mean Sea Surface Height Constraint Factor Method

“…Related research showed that there are implications for satellite altimeter inversion of the marine gravity field in some areas with significant variation in the seafloor topography [22]. In addition to the density and reliability of the satellite altimeter data, topographical factors may also lead to inconsistency in the trend of the distribution of values during the interpolation process [32]. In two figures, there are significant differences between the marine…”

“…However, even for the same spatial resolution, varying interpolation methods make a significant difference in the accuracy of gridded marine gravity fields from the same data source [30]. Furthermore, when commonly used interpolation methods were compared, including inverse distance weighted (IDW), radial basis function (RBF), Shepard, and ordinary kriging, kriging has the highest accuracy but with an unacceptable accuracy loss in areas with sparse data density [31,32]. From the perspective of statistical analysis, adding data correlated with the predicted variables can improve the weight of the kriging predicted points [33].…”