Numerical Modelling of Basin Margin Stratal Geometries: Implications for Sequence Stratigraphy

Sequence stratigraphy is an established methodology for the study of stratal relationships within a chronostratigraphic framework. The method is based on a number of simplifying assumptions, including that changes in relative sea-level (RSL) are the major control on stratal geometry formation. However, because stratal geometries are now understood to be controlled by a number of processes, not simply RSL variations, the application of sequence stratigraphic models and methods for various theoretical and stratigraphic problems is unlikely to be as simple as proposed.

This thesis uses numerical stratigraphic forward models to investigate how different combinations of parameter values for multiple processes can control formation of basin
margin stratal geometries. To achieve this, many hundreds of stratal geometries are generated in two- and three-dimensional model runs covering a wide range of parameter values for different controlling processes. Stratal geometries are then characterised or examined to elucidate the likely controls on formation of various stratal architectures. How the numerically modelled stratal geometries in this work, including stacking patterns, stratigraphic surfaces and shoreline trajectories, can impact sequence stratigraphic interpretation is discussed and illustrated.

Stratal geometries generated from 1264 numerical simulations on a simple ramp topography and on a topography with a shoreline break of slope demonstrate that topset aggradation during RSL fall occurs across a wide range of values of terrestrial diffusion coefficients (representing sediment transport in the model), and of amplitudes and durations of RSL fall. Topset aggradation during falling RSL is particularly prevalent in models with sediment transport rates at the low end of what is observed in modern delta systems. These results suggest that falling-stage topset aggradation is likely to be common in ancient strata, and consequently distinguishing between forced and unforced regressions in the ancient record could be difficult.

Sediment transport rates are also shown to be an important control on shoreline trajectories generated in the numerical model. With constant supply and duration of RSL fall, model runs with relatively high sediment transport rates lead to erosion at the shoreline and increased progradation, whereas model runs with low sediment transport rates lead to aggradation at the shoreline and reduced progradation. Other numerically modelled shoreline trajectories in this thesis also suggest that different shoreline trajectories can result from different sediment transport rates when all other controlling factors are constant. Reliably distinguishing between horizontal, ascending and descending shoreline trajectory classes requires knowledge of the paleo-horizontal surface. This surface can be difficult to identify when tectonic tilting or differential compaction has occurred. Modelled stratal geometries presented show how differential compaction can alter shoreline trajectory geometries to the extent that it is not possible to distinguish between the different trajectory classes. This suggests that it would be difficult to reliably distinguish between strata generated during rising or falling RSL using shoreline trajectories without very detailed two- and three-dimensional backstripping.

Since stratal geometries on basin margins a consequence of multiple different controls, correct sequence stratigraphic interpretation requires an understanding of how multiple controls can generate similar stratal geometries. Numerical model runs are executed in this thesis to explore the impact of time-variable sediment supply and different sediment transport rates on stratal geometries. Four common types of stratal geometry are shown to form by more than one set of parameter values and are thus likely to be non-unique. A maximum transgressive surface can occur in the model due to an increase in rate of RSL rise during constant sediment supply, and due to a reduction in rate of sediment supply during a constant rate of RSL rise. Similarly, sequence boundaries, topset aggradation and shoreline trajectories are also examples of non-unique stratal geometries. If these model results are realistic and non-unique stratal geometries occur in the ancient record, this suggests that the sequence stratigraphic method requires a shift towards an approach that is capable of considering and evaluating multiple hypotheses and scenarios.

Shoreline trajectories interpreted directly from stratal geometries generated in three-dimensional numerical model runs in this work demonstrate that shoreline trajectories interpreted from two-dimensional outcrop or subsurface data may be limited in respect to describing three-dimensional clinoform development. The limitation exists because stratal architectures on basin margins contain significant variability along strike and are thus likely to exhibit different shoreline trajectories along various dip sections. A three-dimensional model run of a basin margin with time variable sediment supply from three sources illustrates how two-dimensional shoreline trajectories interpreted from different dip sections along a basin margin can look different. In order for shoreline trajectory analysis to account for this variability, multiple two-dimensional dip sections should be interpreted and compared.

Results presented in this thesis highlight the limitation of sequence stratigraphic models and methods for extracting reliable information from the ancient record. Stratigraphic correlation using particular stratal geometries, making predictions regarding volume and style of sediment bypass, and reconstructing RSL histories based on observed stratal geometries are probably more complex than sequence stratigraphic methods and models suggest. The results highlight the requirement for a sequence stratigraphic method based on constructing and evaluating multiple hypotheses and scenarios.