Finding Solutions with Land Use Cover and Change Models

Land Use Cover and Change models are complex systems designed to simulate the effects of land use on a catchment area. This includes the profitability of different land use choices and their subsequent effect on the environment. In particular, some land use choices will increase salinity which in turn degrades the land, reducing crop yields and profitability in future years. The models are complex systems due to their size and the interdependence of their components. Computer-based models are useful in predicting the long term efects of land use. However, finding land use choices that match specific requirements can be difficult because of the large search space involved. This paper describes a genetic algorithm for finding solutions. The algorithm is implemented on a Beowulf cluster using Open MPI. Catchment data are distributed to nodes that simultaneously compute solutions using a genetic algorithm. These solutions are then collated by a central node.