Sediment transport in analogue flume models compared with real-world sedimentary systems: a new look at scaling evolution of sedimentary systems in a flume

Evolution of sedimentary systems at large temporal and spatial scales cannot be scaled down to laboratory dimensions by conventional hydraulic Froude scaling. Therefore, many researchers question the validity of experiments aiming to simulate this evolution. Yet, it has been shown that laboratory experiments yield stratigraphic responses to allocyclic forcing that are remarkably similar to those in real-world prototypes, hinting at scale independency with strong dependence on boundary conditions but weak dependence on the actual sediment transport dynamics. This paper addresses the dilemma by contrasting sediment transport rules that apply in the laboratory with those that apply in real-world geological systems. It is demonstrated that the generation of two-dimensional stratigraphy in a flume can be simulated numerically by the non-linear diffusion equation. Sediment transport theory is used to demonstrate that only suspension-dominated meandering rivers should be simulated with linear diffusion. With increasing grain-size (coarse sand to gravel) and shallowness of river systems, the prediction of long-term transport must be simulated by non-linear, slope-dependent diffusion to allow for increasing transport rates and thus change in stratigraphic style. To point out these differences in stratigraphic style, three stages in infill of accommodation have been defined here: (i) a start-up stage, when the system is prograding to base level (e.g. the shelf edge) with no sediment flux beyond the base-level point; (ii) a fill-up stage, when the system is further aggrading while progressively more sediment is bypassing base level with the progression of the infill; and (iii) a keep-up stage, when more than 90% of the input is bypassing the base level and less than 10% is used for filling the accommodation. By plotting the rate of change in flux for various degrees of non-linearity (varying the exponent in the diffusion equation) it was found that the error between model and real-world prototype is largest for the suspension-dominated prototypes, although never more than 30% and only at the beginning of the fill-up stage. The error reduces to only 10% for the non-linear sandy-gravelly and gravelly systems. These results are very encouraging and open up ways to calibrate numerical models of sedimentary system evolution by such experiments.