An ant colony optimization is applied to estimate the hydrogeologic parameters of aquifers for steady state groundwater flow models according to measured water heads and boundary fluxes. The ill-posedness of the inverse problem as characterized by the instability and the nonuniqueness is controlled by making use of computational intelligence. Compared with gradient-type methods

the ant colony optimization is able to converge to global minima for estimating model parameters. A set of numerical experiments are conducted to illustrate the methodology. The results achieved are compared to those previously obtained by a simulation procedure. The research shows that the parameter identification procedure proposed has the characteristics of robustness

global convergence and the ability of fitting measurement noises.