顾及空间各向异性的海岸带水下地形构建方法——以黄河三角洲为例

A Method for Constructing Underwater Topography in Coastal Zones While Accounting for Spatial Anisotropy:The Case of the Yellow River Delta

测深点是获取水下高质量数字高程模型的主要数据源,海岸带区域低密度的测深点会导致水下数字高程模型精度较低。针对该问题,结合反距离权重法与粒子群算法,提出一种顾及各向异性的多参数协同优化的点位加密算法,获取在全局意义下的补插点位水深最优解。以黄河三角洲1992年、2007年和2015年3期测深数据为研究对象,采用4种地学常用的插值方法对提出的加密方法进行验证,结果表明该方法处理后的数据插值精度有显著提高,绝对误差与相对误差均降低12%,能够对水下地形进行更为准确的重建。

Bathymetric points are the main data source for obtaining high-quality underwater Digital Elevation Models(DEMs).The low density of bathymetric points in coastal zones can lead to low underwater DEM accuracy.To address this problem,we propose a multiparameter collaborative optimization algorithm for point densification that considers spatial anisotropy.First,the particle swarm optimization was employed to collaboratively optimize the four parameters affecting the computational accuracy to achieve the overall tuning and determine the optimal solutions of the four parameters.Subsequently,the four determined parameters were applied to the inverse distance weighting method to obtain in the global sense the optimal value of the complementary point depth.Finally,the proposed densification method was validated using four common interpolation methods in geosciences and bathymetric data from the coastal zone of the Yellow River Delta from 1992,2007,and 2015.The experimental results showed that this method significantly improves the interpolation accuracy of the processed data,reducing both the absolute and relative errors by 12%.This method overcomes the problem of large interpolation errors caused by the density of bathymetric points and allows for a more accurate underwater topography reconstruction.