Incorporating bedload transport and bed evolution in soil erosion modelling
Event details
| Date | 29.03.2010 |
| Hour | 16:15 |
| Speaker | Sander, Graham (Loughborough University, UK) |
| Location |
GR B3 30
|
| Category | Conferences - Seminars |
Most soil erosion models consider only sediment transport in suspension; bedload transport is commonly neglected and bed evolution is rarely, if ever, considered. Experimental data on sediment transport under both net erosion and net deposition conditions was recently published in the literature. The inflow boundary condition cycled between net erosion (zero sediment flux) and net deposition (sediment flux greater than the transport capacity) conditions, with periods of 15 minutes per cycle. Application of the Hairsine-Rose (HR) soil erosion model to this data shows that it can reproduce the total outflowing sediment concentrations with time, across the periodic cycles in the boundary condition, as well as the steady-state particle size distribution at the end of each cycle. The HR model also predicts the spatial and temporal evolution of the mass of deposited sediment on top of the original soil surface. While physically plausible estimates of the deposited layer were obtained downstream of the inflow boundary, significant overestimation occurred in a very narrow zone around the inflow boundary itself during the net deposition cycle. The magnitude of the overestimation was such that it would have blocked the flow cross the boundary. As this did not occur during the experiments, an additional transport mechanism must have been active to move this deposited sediment downstream. Visual observations of particles rolling along the bed suggests that bedload transport could be this additional mechanism. However changes in surface slope due to significant deposition can also provide an alternative mechanisms for driving the sediment downstream. An extension of the HR model to include a both these processes allows physically consistent predictions of the deposited masses to be obtained for all space and time. It is commonly thought that a sufficient and demanding test of an erosion model is whether it can reproduce spatial and temporal experimental data on the size distribution of suspended sediment. This is in fact not the case, and that accurate predictions of this type of data can be obtained with models that ignore important transport mechanisms. Finally an application of this extended model to flumed data obtained or supercritical flow into a nail bed (representing a filter or buffer strip) gives good agreement with both the deposited mass within, and transported mass through the nail bed.
Links
Practical information
- General public
- Free
Contact
- andrew. [email protected]