Mahamadou Traoré, Laurent Esclade, Ousmane Thiaré, and Bernard Pottier
Physical modeling, Synchronous simulation, Cellular automata
Desert locusts cause severe damages to crops in several regions of the world, particularly in west Africa where the last 2004-2005 plague has cost US $ 400 million. Despite efforts to control population, the risk of invasion remains present. This paper presents a modeling/simulation work directed to monitor such risk. With a set of tools developed in laboratory, a cellular automaton parametric model is built for simulating population dynamics in relation with geographic area. Simulations are time driven, synchronous and can be executed on multi-core processors or graphic accelerators from the same set of tools. The model is based on geographic cells that hold insect population physical state following insect life cycle, and can handle neighborhood exchanges. Cell data can be complemented by real time data such as vegetation cover, climate or weather. Synchronous simulations from observed distributions show how the model can predict population evolution.