基于重力地质法的南中国海海底地形反演

根据重力地质法（GGM），利用南中国海海域内63179个船测控制点水深将测高自由空间重力异常划分为长波参考场和短波残差场，并反演出了该海域112°E-119°E，12°N-20°N范围的1’×1’海底地形模型，该过程中使用的海水和海底洋壳密度差异常数1.32 g·cm^-3通过实测水深估计得到.利用反演得到的GGM模型对剩余的10529个检核点船测水深插值计算后与实测水深进行比较，其较差结果的均值为-1.64 m，标准差为76.95 m，相对精度为4.06%.此外，根据船测点数量、分布和海底地形的不同，选择了三个海域进行统计，结果表明：在船测控制点分布均匀的海域，GGM模型精度优于ETOPO1模型，在控制点过于分散的海域其精度会有所下降，但好于船测水深的直接格网化结果.为进一步探究检核点的较差结果中出现较大数值的成因，本文对精度较差的点位进行了单独分析，选择了两条船测航迹剖面进行了研究，并分析了检核点的水深较差、相对精度与水深和重力异常的关系，结果表明：GGM模型精度受水深和重力异常的相关性影响较小，受海底地形复杂程度影响较大，地形坡度变化平缓海域的预测精度明显高于海山地区.最后，综合GGM模型和ETOPO1模型优势，利用所有船测水深作为控制，生成了综合的海底地形模型.

Predicting bathymetry in South China Sea using the gravity-geologic method

The gravity-geologic method (GGM) was implemented for 1'×1' bathymetry determination in 112°E—119°E, 12°N—20°N region in the South Sea .63179 ship-derived data (as the control points) were used in this method to divide the observed gravity anomalies into long- and short-wavelength gravity anomalies, the density contrast between the seawater and the ocean bottom topographic mass is 1.32 g·cm^-3, which was estimated by ship data in this area. The residual 10529 ship-derived data(as the check points) were used for estimating accuracy of the predicted bathymetry model, differences between the interpolation depths of GGM model and the ship-data of check points reached approximate mean value of -1.64 m, standard deviations of 76.95 m and relative precision of 4.06%. In addition, according to the differences of the quantity, distribution and ocean bottom topographic conditions of the control points, three areas were chosen for accuracy statistics, results showed that:in areas with denser and better distributed shipborne measurements, the bathymetric model predicted by GGM was more accurate than ETOPO1 model, in other areas, the accuracy of GGM model may be lower at a certain extent, but which was still more accurate than directly gridded model from dispersed ship data. In order to determine the reason of big differences, check points with worse precision were analyzed along, profile of shipborne depths from two cruises were selected for researching, what's more, relationships between the differences, relative precisions and the bathymetry and gravity anomaly were also showed, respectively, results showed that: In term of the GGM model's accuracy, the influence of correlations with bathymetry and gravity anomaly was small, conversely, the influence of the complexion's degree of ocean bottom topography was large, topography gradient varied smoother resulted in a better predicted accuracy. As a final result, the respective advantages of GGM model and ETOPO1 model were considered and then a comprehensive bathymetry model was presented by integrating all available ship data.