Modeling Deltaic Stratigraphy Through Superposition of Discrete Networks
Previously we have developed a model capable of generating deltaic networks using a partly correlated random walk algorithm. The model can render distributary networks that can vary in terms of overall shapes, number of individual channels, channel sinuosity, distribution of internal bifurcation nodes and of channel outlets (i.e., delta mouths). This versatility is suitable for reproducing the variability of natural deltas. Here we tap into a powerful trait of the model: its speed of execution. By constraining some of the parameters, a large set of thousands of networks, similar in terms of overall size, shape, and number of major channels, can be rapidly generated. Deltaic stratigraphy can be built by stacking a series of networks drawn from the large set. Individual channel dimensions were assigned as functions of the discharge (Andren, 1994) which is routed and conserved from the apex to the shoreline. For simplicity the coarse sediment was assumed to be concentrated within the channel (sand fill) while fine sediment characterizes the inter-channel space. Using the cosine similarity as a simple method of comparing networks, continuous stratigraphic sequences can be obtained by orderly stacking networks, starting with a random outcome, followed by the most similar network in the large set and so forth. Although the channels were assumed to have simple, symmetrical shapes, complex channel fill geometries arise from the sequential stacking (Sylvester et al, 2011). The vertical ordering of similar networks is a proxy for discrete evolution of a deltaic system characterized largely by progradation and lateral channel migration. The series of similar networks are separated by interspersed randomly chosen networks that correspond to the more rapid reconfiguration of the deltaic system through lobe switching or avulsion of a dominant channel. Stratigraphic arrangements at basin scale are first explored by assuming constant aggradation and subsidence rates, with little variation in shoreline position. A full eustatic cycle is imposed to observe more complex variability of stratigraphic outcome for different net to gross ratios. In sum, the ability to create large sets of deltaic networks sidesteps the complex long term hydrodynamic modeling of a channel network, offering an effective alternative tool for generating and exploring detailed stratigraphic arrangements.
AAPG Datapages/Search and Discovery Article #90291 ©2017 AAPG Annual Convention and Exhibition, Houston, Texas, April 2-5, 2017