Building Synthetic Bathymetries for Unsurveyed Reservoirs: Hydrologically Conditioned Cubic Spline Interpolation

Despite their importance in social, economic, and hydroecological terms, numerous dams and reservoirs spread worldwide lack bathymetric or even storage data of any kind. This scarcity of minimal storage and morphology data represents an obstacle to evaluating the current functions and impacts of these artificial reservoirs, hindering planning efforts for their future development or removal and limiting hydrologic, hydrodynamic, or water quality modeling efforts on dammed watersheds. Although some global databases provide records with basic information on the location and surface extent of large reservoirs—including attributes such as dam height, total storage capacity, or even storage curves—bathymetries remain rarely available. Surveying bathymetric data is challenging and expensive, especially in the least-developed countries, regions experiencing war conflicts, or remote areas exposed to natural risks. This article presents a method for deriving synthetic bathymetries using cubic spline spatial interpolation. The method uses a digital terrain model (DTM) surrounding the reservoir as the only required input, assuming geomorphological and hydrological continuity and imposing longitudinal depth controls to the interpolation process. In comparison with a sample of 12 vessel-based reservoir bathymetric surveys, available from the United States Bureau of Reclamation (USBR), the synthetic bathymetries resulting from this interpolation method show bias performances within a range of $\pm 25\%$ and Kling–Gupta efficiencies (KGE) over 0.5. Except for vessel-based bathymetric surveys, any indirect depth estimation method that applies remote sensing or interpolation is susceptible to some level of uncertainty and varied reliability. However, cubic spline spatial interpolation can produce bathymetric models with a reasonable level of accuracy depending on the projected application. The performance levels suggest that the proposed method can produce synthetic bathymetries that offer valuable information for early-stage hydrological or environmental studies or assessments. Some potential applications include initial evaluations involving problems like dam failure and dam removal or restoration, where bathymetric data play a relevant role as an input of hydrological, hydraulic, environmental, and risk simulations. In prefeasibility or conceptual stages, these types of projects and assessments can tolerate higher levels of uncertainty in the input data before making decisions to proceed to advanced stages of assessment, design, construction, restoration, or demolition. The method offers a low-cost and time-efficient alternative solution if vessel-based bathymetric surveys or remote sensing techniques turn out to be unfeasible for future studies or assessments in data-scarce regions. Reliable synthetic bathymetric interpolation turns out to be especially valuable in remote or dangerous sites posing physical risks, social conflicts, or any other safety threats for field surveys.