P 1 . 7 Genetic Algorithms in Geophysical Fluid Dynamics

The genetic algorithm (GA) is an optimization tool that has shown great success at solving problems not amenable to easy solution via more traditional means (such as the traveling salesman problem, solved by Koza 1992). Since most CFD problems have not been traditionally posed in terms of optimization, GAs have not yet been widely used in this field. The goal of this work is to demonstrate the ability of the GA to provide interesting solutions to problems difficult to solve in more traditional ways. One need only to be a bit creative in posing the problem as one in optimization. For instance, boundary value problems can be seen as minimizing the discretized version of the magnitude of a partial differential equation (PDE). Similarly, inversion problems often involve finding the best fit parameters to an assumed version of a model. For nonlinear model forms, it is often difficult to analytically minimize the mean square difference between model and data. These are cases where the GA has potential to help in making strides forward.