Geographia Technica, Vol 20, Issue 1, 2025, pp. 161-178
ASSESSMENT OF DAM BATHYMETRY USING GIS AND GEOSTATISTICAL METHODS. CASE STUDY OF THE GHRIB DAM (ALGERIA)
Souhila BENKACI 
, Boualem REMINI, Salah Eddine CHECHAM
ABSTRACT: Assessing variations in water depth at a dam is highly beneficial for dam managers, as it enables more effective management of sluice gates and improves the planning and coordination of emptying and dredging operations. In this context, our study aims to analyze the spatial variability of water depth based on a sample of 44922 bathymetric survey data points measured at the Ghrib dam level. Initially, an exploratory analysis was performed, during which outliers were removed. Four semivariograms fitting models were then tested: Exponential, Spherical, Circular and Gaussian. Model performance was assessed using cross-validation tests, including mean error (ME), mean square error (MSE), root mean square error (RMSE), standardized mean error (MSE) and standardized root mean square error (RMSSE). The Gaussian model yielded the best results and was thus selected for interpolation using ordinary kriging. The resulting bathymetric map highlighted distinct classes representing estimated water depths within the Ghrib dam reservoir. The minimum water depth was estimated at approximately 18 meters, with a maximum of around 53 meters, which corresponds to the maximum water depth observed at the dam wall (near the dike). This bathymetric map serves as a valuable decision-making tool. However, while the kriging technique proved effective in this study, it would be advisable to generate multiple bathymetric maps using alternative interpolation methods, such as universal kriging, IDW, or spline, and to validate the most accurate map by comparing the results of these different approaches.
Keywords: Bathymetry, GIS, Geostatistical analysis, Kriging, Semivariogram
Full article here