Multi-scale simulation of fluvio-deltaic and shallow marine stratigraphy

Abstract

SimClast is a basin-scale 3D stratigraphic model, which allows several interacting sedimentary environments. We developed it from 2004 to 2008 at Delft University of Technology and implemented part of the Meijer (2002) code for accounting, loading and storing algorithms. SimClast is a fully plan view 2D, depth-averaged model, allowing the complex interaction between fluvial and wave influences on deltaic and shoreface development to be studied. It focuses on theoretical experiments, as quantitative experiments are intrinsically difficult to recreate in real world settings. Yet there lies the great strength of numerical modelling, as we can improve upon the understanding of these systems by focussing on the process forming and removing the deposits. The modelling applications focus especially on the erosional and nondepositional events as these probably represent the greatest amount of “stratigraphic time”. Short-term, high-resolution processes are coupled with the long-term stratigraphic model by nesting a parameterised version of the high-resolution processes. We extrapolate physical and empirical relationships of the geomorphological development and implement these. A necessary constraint on these long-term models is a relatively large grid sizing (i.e. km scale), as the area to be modelled is on the scale of continental margins and the modelling time is on the scale of many millennia. Areas of special importance are modelled by implementing sub-grid scale processes into a large-scale basin-filling model; this refines the model dynamics and the resulting stratigraphy. Processes included are; fluvial channel dynamics and overbank deposition, river plume deposition, open marine currents, wave resuspension, nearshore wave induced longshore and crosshore transport. This combined modelling approach allows insight into the processes influencing the flux of energy and clastic material and the effect of external perturbations in all environments. Many governing processes work on relatively small scales, e.g. in fluvial settings an avulsion is a relatively localised phenomenon, yet they have a profound effect on fluvial architecture. This means that the model must mimic these processes, but at the same time maintain computational efficiency. Additionally, long-term models use relatively large grid sizing (km scale), as the area to be modelled is on the scale of continental margins. We solve this problem by implementing the governing processes as sub-grid scale routines into the large-scale basin-filling model. This parameterization greatly refines morphodynamic behaviour and the resulting stratigraphy. SimClast recreates realistic geomorphological and stratigraphic delta behaviour in river and wave-dominated settings.