10:00 AM - 10:15 AM
[ACG42-05] Morphodynamics of sea floor: challenges in modeling suspended sediment transport
★Invited Papers
Keywords:sediment gravity flow, suspended load, deap sea
The deep marine is often considered a stable environment, but in reality, various events generate erosional or depositional currents and dynamically change the topography. Sedimentation and erosion on the seabed are risk factors for facilities installed in the sea, especially offshore wind turbines. Turbidites (deposits of turbidity currents) on the deep-sea floor generated by large earthquakes and tsunamis are attracting attention as a record of past disaster events. It is known that sedimentation from suspended sediment plays an essential role in the depositional process from river flooding flows, not only on the deep-sea floor. Therefore, modeling suspended sediment transport is necessary to perform inverse analysis of geologic records of disasters and future predictions.
However, there are several challenges in modeling suspended sediment transport. When modeling large-area and -scale suspended sediment transport, suspended sand transport is often combined with a 1D or horizontal 2D layer-averaged model. One of the most challenging aspects to model within this framework is the behavior of sediment particles near the bed. The flux of suspended sediment particles near the bottom has to rely on an empirical law (sediment entrainment function) based on experiments and field observations. The same applies to vertical 2-D or 3-D models. Thus, the accuracy and precision of the entrainment function is the key to modeling suspended sediment transport. There are two unresolved problems for establishing this function, particularly with regard to (1) high-velocity flow and (2) handling of mixed grain sizes.
(1) Firstly, the current functions do not cope with high-velocity flows (high bottom shear stress) exceeding 3.0 m/s. The entrainment functions used in many existing studies predict unrealistic values of suspended sediment concentrations (in some cases more than 100%) for flows in the high-velocity range. Why do entrainment functions established by the experiments of the low-speed regime predict abnormally high concentrations? In the high-velocity regime, the reduced shear stress associated with this enhanced density stratification prevents the suspended sediment concentration from increasing abnormally. We show that, considering the conservation law of turbulent kinetic energy and hindered settling effects, there can be two equilibrium states (high and low-density equilibrium) for the suspended sand concentration, depending on the initial value. Tsunamis and turbidity currents start from a low concentration state, so the suspended sediment concentration cannot reaches high concentration state. On the other hand, debris flows and hyperconcentrated flows start from a high-concentration state and are assumed to flow in another high-concentration equilibrium state. It is expected that future observations and experiments will verify these theoretical assumptions.
(2) It is known that the actual landform development by suspended sediment is strongly related to the grain-size sorting during suspended sediment transport. However, the current sediment entrainment functions need more basic data to consider mixed grain size. When applying the entrainment function to mixed grain size sediments, the sediment grain size distribution near the bottom surface has to be taken into account. Only a few studies exist on how the thickness of the active layer (the region of the surface layer where suspended and bottom sediments are exchanged) should be set. In addition, when dealing with mixed grain sizes, it is necessary to consider inter-particle interactions, known as hiding effects, which have also been studied in very few cases. In the future, conducting experiments and observations focusing on the mixed grain size and establishing a new entrainment function using a data-driven scientific approach will be necessary.
However, there are several challenges in modeling suspended sediment transport. When modeling large-area and -scale suspended sediment transport, suspended sand transport is often combined with a 1D or horizontal 2D layer-averaged model. One of the most challenging aspects to model within this framework is the behavior of sediment particles near the bed. The flux of suspended sediment particles near the bottom has to rely on an empirical law (sediment entrainment function) based on experiments and field observations. The same applies to vertical 2-D or 3-D models. Thus, the accuracy and precision of the entrainment function is the key to modeling suspended sediment transport. There are two unresolved problems for establishing this function, particularly with regard to (1) high-velocity flow and (2) handling of mixed grain sizes.
(1) Firstly, the current functions do not cope with high-velocity flows (high bottom shear stress) exceeding 3.0 m/s. The entrainment functions used in many existing studies predict unrealistic values of suspended sediment concentrations (in some cases more than 100%) for flows in the high-velocity range. Why do entrainment functions established by the experiments of the low-speed regime predict abnormally high concentrations? In the high-velocity regime, the reduced shear stress associated with this enhanced density stratification prevents the suspended sediment concentration from increasing abnormally. We show that, considering the conservation law of turbulent kinetic energy and hindered settling effects, there can be two equilibrium states (high and low-density equilibrium) for the suspended sand concentration, depending on the initial value. Tsunamis and turbidity currents start from a low concentration state, so the suspended sediment concentration cannot reaches high concentration state. On the other hand, debris flows and hyperconcentrated flows start from a high-concentration state and are assumed to flow in another high-concentration equilibrium state. It is expected that future observations and experiments will verify these theoretical assumptions.
(2) It is known that the actual landform development by suspended sediment is strongly related to the grain-size sorting during suspended sediment transport. However, the current sediment entrainment functions need more basic data to consider mixed grain size. When applying the entrainment function to mixed grain size sediments, the sediment grain size distribution near the bottom surface has to be taken into account. Only a few studies exist on how the thickness of the active layer (the region of the surface layer where suspended and bottom sediments are exchanged) should be set. In addition, when dealing with mixed grain sizes, it is necessary to consider inter-particle interactions, known as hiding effects, which have also been studied in very few cases. In the future, conducting experiments and observations focusing on the mixed grain size and establishing a new entrainment function using a data-driven scientific approach will be necessary.