Free satellite data can now reliably estimate small dam storage capacity
Researchers have validated that freely available satellite elevation maps can accurately measure water storage in small dams, potentially saving developing countries millions in surveying costs. The finding opens a cost-effective path for water infrastructure planning in regions where budget constraints have forced reliance on outdated or incomplete data.
Originaltitel: Assessing SRTM one Arc second DEM accuracy for small dam volume-elevation curves using terrain metrics
<p>Accurate reservoir storage estimation is fundamental to sustainable water resources management; however, small dam projects are often hindered by the prohibitive costs and time required for high-precision topographic surveys. In this study, a rigorous validation of freely available one-arc-second SRTM DEMs was conducted as an alternative approach for estimating volume-elevation relationships at ten small dams in Iraq. High-precision surveys served as benchmarks, enabling statistical validation of DEM-derived estimates using absolute relative error (ARE), root mean square error (RMSE), mean absolute error, and the coefficient of determination (R²). Reservoir basin morphology was further characterised through planimetric indices, including area-to-volume ratio (AVR), shape factor, and solidity. In parallel, terrain complexity within a 5 km buffer zone was quantified using slope variability, curvature, vector ruggedness measure (VRM), and terrain ruggedness index (TRI). A strong structural agreement was demonstrated (R² > 0.98), although substantial variation in volumetric precision was observed. A global sensitivity analysis using the Morris Method identified the standard deviation of the Terrain Ruggedness Index (TRI) as the dominant predictor of accuracy, with µ* values of 83–86, while all other metrics showed minimal influence (µ* ≈ 0–24). These results establish a clear accuracy threshold for one-arc-second SRTM DEMs: they are sufficiently reliable for preliminary planning (< 20% error) in low-ruggedness terrain (TRI SD < 0.1) but become highly unreliable in rugged landscapes, where errors exceed 150% (TRI SD > 0.1). These findings provide a predictive framework for assessing DEM suitability, supporting the integration of satellite topography into small-scale reservoir planning.</p>