Upstream and downstream controls on aggradation in a small-scale physical river model: implications for fluvial stratigraphy

Abstract:

Fluvial stratigraphy has the potential to record changes in accommodation generated by both downstream drivers (i.e. sea-level change) and-or upstream drivers (i.e. water and/or sediment discharge changes). Hence, the hypothesis is that changes in either of these drivers can potentially cause rivers to respond and adjust to a new equilibrium river profile by aggradation or degradation. Fluvial stratigraphy, therefore, remains a major archive of past environmental changes. However, we still lack the ability to differentiate the manifestation of upstream and downstream driven changes when faced with fluvial stratigraphic successions. Here we explore this problem using physical modelling, which has the advantage of being able to independently control both upstream and downstream parameters.

In our ongoing study we have designed and manufactured a narrow (0.05 m), long (2.25 m) flume with an initial gradient of zero. The narrow width of the flume allows us to have a quasi 1-D system and additionally decreases experimental running time. We have developed an algorithm that enables us to analyze the resultant sand wedge from side-profile photos and fit a linear regression to investigate slope. Furthermore, we have positioned a camera so that the top view of the experiment is also recorded and changes in channel width can be measured. In our experiments we: (i) investigate the role of changes in water discharge and sediment supply on equilibrium river profiles, and (ii) carry out a series of perturbation experiments varying downstream drivers (i.e. sea-level) which theoretically produce the same amount of aggradation as the upstream parameters we have used. This enables us to compare any similarities or differences in morphology. Our preliminary findings suggest that changes in sediment concentration are the dominant factor in controlling slope (when using the same single grain-size distribution for all experiments). Ultimately, we will migrate these experiments to a fully unconfined flume which will allow us to compare the changes we have seen in the ‘1-D’ to the 3-D.