Date Published: April 6, 2017
Publisher: Public Library of Science
Author(s): Ahmed Bilal, Wenhong Dai, Magnus Larson, Qaid Naamo Beebo, Qiancheng Xie, Jun Xu.
Sediment-dynamics modeling is a useful tool for estimating a dam’s lifespan and its cost–benefit analysis. Collecting real data for sediment-dynamics analysis from conventional field survey methods is both tedious and expensive. Therefore, for most rivers, the historical record of data is either missing or not very detailed. Available data and existing tools have much potential and may be used for qualitative prediction of future bathymetric change trend. This study shows that proxy approaches may be used to increase the spatiotemporal resolution of flow data, and hypothesize the river cross-sections and sediment data. Sediment-dynamics analysis of the reach of the Tenryu River upstream of Sakuma Dam in Japan was performed to predict its future bathymetric changes using a 1D numerical model (HEC-RAS). In this case study, only annually-averaged flow data and the river’s longitudinal bed profile at 5-year intervals were available. Therefore, the other required data, including river cross-section and geometry and sediment inflow grain sizes, had to be hypothesized or assimilated indirectly. The model yielded a good qualitative agreement, with an R2 (coefficient of determination) of 0.8 for the observed and simulated bed profiles. A predictive simulation demonstrated that the useful life of the dam would end after the year 2035 (±5 years), which is in conformity with initial detailed estimates. The study indicates that a sediment-dynamic analysis can be performed even with a limited amount of data. However, such studies may only assess the qualitative trends of sediment dynamics.
Sediment transport modeling is a process of using a numerical or physical model to reproduce the transport processes of a real river system in a controlled environment. Mathematical modeling provides useful knowledge about a river and dams built on it and can able to predict the effect of different external factors on bathymetric changes and reservoir sedimentation. This knowledge assists efficient and sustainable development related to rivers . Reservoir sedimentation not only negatively affects the storage capacity of a reservoir , but it also alters the current biogeochemical and ecological cycles[4–7]. Sediment-dynamics analysis provides information about bathymetric changes upstream of a reservoir. A numerical model may be used for this analysis to predict the upcoming trends of bathymetric changes under different scenarios. This process is a complex task that requires a significant amount of observed data. Among other factors, the prediction quality of a numerical model also depends on the spatiotemporal resolution of the observed records, which include flow time series, sediment inflow and outflow time series, river cross sections, and river bed soil data .
HEC-RAS is designed to perform 1-D steady and unsteady flow, sediment transport, and water quality modeling. Modeling sediment transport behavior is not an easy task, as in addition to theoretical derivations it also has some form of empiricism, i.e. based on experimental data and is not derived from laws of physics. Further, it also contains a broad range of very sensitive physical parameters . HEC-RAS uses a hydrologic simplification called the quasi-unsteady flow method, which approximates a continuous hydrograph with a series of discrete steady flow profiles .
In general, the model performed satisfactorily once all parameters were set within their effective ranges. Despite many simplifications, HEC-RAS has produced reasonable results with the data available for the period of study. The model results have R2-values ranging from 0.68 to 0.96, and the relative difference in river bed profile change between the modeled and observed data is in the range 8%–75%. The relative difference is greater at the beginning of the simulation period when the changes are smaller, but it decreases toward the end of the simulation period. The relative loss of storage volume due to sedimentation is also similar in the observations and the model results.
Conventional methods for gathering data are costly, and historical flow and sediment data are not always available. Even if the data are available, it may not be in the desired spatiotemporal resolution, particularly for transboundary river basins in developing countries. This study explored the possibility of developing river hydraulic models using simplifications, hypotheses, and freely available data and tools (e.g., Google Earth) to overcome a lack of observed data.