Abstract Determining the bathymetric properties of a river, lake or reservoir is important for many reasons because it identifies location of sections with sedimentation or erosion which indicate the key processes and biophysical habitat status. This study used observational data of a 5 km section of the Danube River (transformed into a reservoir after 1972) to determine the best GIS interpolation method and the optimal number of sampling points to get an accurate description of the riverbed. Nine interpolation methods were used to generate bathymetric maps of the Danube course, between km 955 and km 950, in the area of Iron Gates reservoir, using a variable number of points (n = 1943; n = 4328; n = 17139), obtained by single beam echo sounder. Six methods of deterministic interpolation have been used: Inverse Distance Weighting (IDW) – criteria automatically determined (CAD) and manually determined (MD); Local Polynomial Interpolation (LPI) – criteria automatically determined (CAD) and manually determined (MD); Radial Basis Function (RBF) – fully regularized spline CRS; Radial Basis Function RBF – spline with tension SWT, and three geo-statistics methods: Ordinary Kriging – OK, Simple Kriging – SK, Universal Kriging – UK. Cross-validation and bathymetric Digital Elevation Models (DEM s) analysis were used to evaluate the accuracy of the estimated data and maps. The two interpolation methods, Inverse Distance Weighting (IDW-MD) and Simple Kriging (SK), provide satisfactory results, even in the case of a small number of points (n = 1943). If the number of measured points grows and a section grid (perpendicular/longitudinal) is developed, the DEM s bathymetric accuracy improves significantly, especially in the case of complex morphology, even if the positive and negative forms of the bed lead to sudden changes in the measured values. The obtained results revealed differences between the interpolation methods used, particularly in the case of a small number of points measured per perpendicular trajectories, but by the addition of points on longitudinal paths, a significant improvement of the bathymetric maps is obtained.

中文翻译：

---

点数、横断面位置和插值技术在测深测量分析中的重要性

摘要 确定河流、湖泊或水库的测深特性很重要，原因有很多，因为它确定了沉积或侵蚀部分的位置，表明关键过程和生物物理栖息地状态。本研究利用多瑙河（1972 年后改建为水库）一段 5 公里的观测数据，确定了最佳 GIS 插值方法和最佳采样点数，以准确描述河床。九个插值方法被用来生成多瑙河航道的测深图，在 955 公里和 950 公里之间，在铁门水库区域，使用可变数量的点（n = 1943；n = 4328；n = 17139），获得通过单波束回声测深仪。已经使用了六种确定性插值方法：反距离加权 (IDW) – 自动确定 (CAD) 和手动确定 (MD) 的标准；局部多项式插值 (LPI) – 自动确定 (CAD) 和手动确定 (MD) 的标准；径向基函数 (RBF) – 完全正则化样条 CRS；径向基函数 RBF – 带张力 SWT 的样条曲线和三种地质统计方法：普通克里金法 – OK、简单克里金法 – SK、通用克里金法 – 英国。交叉验证和测深数字高程模型 (DEM) 分析用于评估估计数据和地图的准确性。两种插值方法，即反距离加权 (IDW-MD) 和简单克里金法 (SK)，即使在点数较少 (n = 1943) 的情况下也能提供令人满意的结果。如果测量点的数量增加并开发了截面网格（垂直/纵向），DEM 的测深精度显着提高，特别是在复杂形态的情况下，即使床的正负形式导致测量值的突然变化。获得的结果揭示了所使用的插值方法之间的差异，特别是在每个垂直轨迹测量的点数较少的情况下，但通过在纵向路径上添加点，获得了测深图的显着改进。