Cellular Automata (CA) can be efficiently applied in the simulation of complex natural processes. Debris flows can be described in terms of local interactions among their elementary portions, and can be efficiently modeled through CA. In this paper, the general frame of the cellular automata model SCIDDICA (Simulation through Computational Innovative methods for the Direction of Debris flow path using Cellular Automata) for simulating Debris flow phenomena are presented. this paper is to present a model for simulating Debris flows with MATLAB programming tools. The program has been validated through the reconstruction of the initial topographic and geomorphologic conditions of a typical phenomenon which occurred at Bayanlou. Thereafter, the simulation results show good agreement with the actual observed debris flow path.
Cellular Automatas (CAs) are decentralized spatially extended systems consisting of large numbers of simple identical components with local connectivity. Such systems have the potential to perform complex computations with a high degree of efficiency and robustness, as well as to model the behavior of complex systems in nature (Barks, 1970: Wolfram, 1986). Some types of landslides, such as Debris flows (Johnson and Rodine, 1984), match well this requirement. CAs are based on regular division of space in cells. Each cell has an identical shape (square or hexagonal in SCIDDICA) and embeds an identical finite automaton (fa) whose state at the beginning of the simulation defines the CA initial configuration. Input for each fa is given by the state of neighboring cells. The state of each cell such as the altitude, dept of soil cover, thickness of landslide debris, landslide energy describe the physical characteristics of the corresponding portion of space. Elementary processes constitute the transition function (σ) of the model which is made of a set of rules. These rules simultaneously applied, step by step, to each of the cellular space. At t=0 cell states are initialized by means of input matrices. Model parameters are assigned in this phase. By applying the transition function to all cells, at discrete steps, states changed and the behavior of the phenomenon can finally be simulated. SCIDDICA was originally developed for simulating flow-like landslides. The model SCIDDICA release "T" applied to the landslides of Tessina (Avolio et al., 2000). The release "O", applied to Mt. Ontake landslide, is an extension of the T-model (Di Gregorio et al., 1999). In this case, which was extremely rapid (20–26 m/s), they considered kinetic energy of the landslide and the possibility of up-slope displacement. Therefore, they added "run-up" to the model and this debris avalanche successfully analyzed. After the May 1998 debris-flow disaster at Sarno (Southern Italy) family of SCIDDICA was developed for simulating slope movement, which was characterized by strong soil erosion along landslide path. Then, the S1 release obtained (D'Ambrosio et al., 2003b). By critically analyzing the first result of simulation, some minor modifications to the model applied and the S2 release developed (D'Ambrosio et al., 2002).
