The project is directed toward the advancement of empirical techniques to model the magnetospheric dymamics as characterized by various ground-based indices derived from magnetic data. Effort will be focused in the development of nonlinear dynamical models which address the spatial structure of agnetospheric disturbances obtained from the analysis of observational time series data. Coupling between different regions of the magnetosphere will be studied to build models to predict the local behavior. The project will be novel in the use of multi-point groundbased satellite data for the development of a multivariate model of the dynamic behavior of different regions of the magnetosphere and for the analysis of this model to identify dominant physical processes which may complement the development of first principle models.