Abstract A numerical model was developed to simulate sedimentary architectures created by turbidity currents over geological time. The model is based on the cellular automata paradigm. The automata exchange matter and energy and are built to reproduce the physical processes which govern turbidity-current behavior. The simulated architecture is the result of a given set of geological events. For each of these events a steady state is computed. This steady state is assumed to be representative of the average effect of a turbidity current on the construction of the sedimentary architecture. Using the model, we studied the impact of external controls on deep-water depositional systems. Topographic control on geological deposits is studied using various slopes. Several architectures are also reproduced thanks to spatial and temporal variations in the frequency and magnitude of the turbidity-current events. The role of suspended-sediment concentration and grain-size distribution on the transport efficiency are presented. The model results show that small variations of flow inputs may have strong controls on deposit evolution. This numerical approach allows a better identification and understanding of key physical parameters and may provide a better prediction of reservoir architecture in deep-sea clastic systems.

Abstract In the middle slope section of the Lower Congo Basin a late Miocene channel system was tracked on 3D seismic profiles over more than 350 km. Along its course, between shelf and basin, the system encountered four regional tectonic structures that induced local slope modifications, either by uplift or by subsidence. The turbidite deposits of this channel system were influenced strongly by these structures, in terms of both character and morphology. From the proximal to the distal part of this channel system, variations in parameters such as the sinuosity, the width and depth of basal incision, the presence of splay and levee deposits, the location of vertical aggradation zones and channel avulsion, all correlate with changes in longitudinal slope gradient. Thus, along a conventional sigmoidal slope, the convex regions are subjected to erosional processes whereas the concave regions are depositional. The direct relationship observed between sedimentary deposits and changing slope gradient highlights an important control in deep-water turbidite systems. This apparent response to local gradient changes on the slope suggests the existence of a sedimentary equilibrium profile similar to that defined for fluvial systems.

A numerical model useful for the prediction of the runout distance, deposit dimensions, and internal properties of turbidity current deposits is proposed. The model uses an integrated formulation of the Chezy equation and includes geotechnical properties (shear strength) of seafloor sediments. The model is applied to a historical flood of the Saguenay River (Quebec, Canada) that created a month-long turbidity current that flowed along the seafloor of the Saguenay Fjord. Model simulation of plume acceleration includes the effects of fluid drag at both the bottom and top of the plume, and grain friction within the plume is modelled as a function of volumetric concentration. Entrainment of sea water is computed to increase the volume of the plume by a factor of five. The model also includes the effects of plume pushing as a function of the rising and the falling limbs of the flood event. Sediment deposition is assessed using removal rate values published by Syvitski and Lewis (1992). Deposition of sediment depends very much on the phase of the river flood event. Measured peaks in the thickness of the turbidite deposit are simulated by the model and correspond to a progressive shift seaward in the location where deposition begins. During the initial phase of the flood event, and until the peak discharge, the turbidite is inversely graded and coarse particles (sand and core silts) represents more than 75% of the weight of the sediment deposited. During the decreasing discharge phase, initial deposition progressively shifts landward and deposition of the finer size fractions (medium silts and finer) dominate. The simulated turbidite is then well graded and matches both core lithology and deposit thickness and runout distance. Our numerical model can also be applied to large ignitive turbidity current events and their deposits, providing a useful tool in the prediction of reservoir properties of submarine fans, e.g. porosity and permeability.