Simulating Turbidity Currents Through a Cellular Automata Model

Salles, Tristan, University of Bordeaux I- Geology and Oceanography Department, Talence Cedex, France

Cellular Automata (CA) represent a formal frame for dynamical systems, which evolve on the base of local interactions. Some types of gravity flows, such as turbidity currents match well this requirement. For CA simulation purposes, turbidity currents can be viewed as dynamical systems, subdivided into elementary parts, whose state evolves exclusively as a consequence of local interactions and internal transformations within a spatial and tempo­ral discrete domain. The model is developped for unsteady, two-dimensional, depth-aver-aged particle-laden turbulent underflows driven by non-uniform and non-cohesive sediment. Space is discretized in square cells. The attributes of each cell describe physical character­istics. For computational reasons, the natural phenomenon is ‘decomposed’ into a number of elementary processes, with a particular composition that makes up the transition func­tion of the CA. The evolution of the phenomenon can be simulated by applying this function to all the cells simultaneously. The transition function includes effects of water incorpora­tion at the suspension-ambient fluid interface, transport equation for the particles volume concentration and toppling rule for the deposited sediments. Simulations of real turbidity currents, occured in Cap-Breton canyon in December 1999, have been performed for model calibration purposes. A comparison of the nature, the amount of sediment supply and the geometry of the basin between theoretical predictions and real event is described. Simple and flexible, it constitutes a first step towards quantitative comprehension of the impact of external parameters on turbidity current dynamics and on the organisation of subsequent depositional sequences (turbidites).