Three-dimensional artificial neural network model of the dayside magnetopause

Artificial neural networks (ANN) from the package NeuroShell 2 are applied to the modeling of the dayside magnetopause. The model data set is based on magnetopause crossings and on the corresponding hour-averaged measurements of the solar wind plasma and the interplanetary magnetic field. The ANN mo del represents the three-dimensional shape and size of the dayside magnetop ause as a function of the solar wind dynamic pressure and the B-z, and B-y. components of the interplanetary magnetic field. The model describes the d ynamics of the magnetic cusp and global asymmetry of the dayside magnetopau se under a wide range of external conditions. Small variations in the size of the magnetopause as a function of the absolute value of B-y are presente d quantitatively in the form of an analytic expression. The ANN model shows three regimes of magnetopause dynamics that are controlled by the B-z, com ponent: a pressure balance regime (B-2>0 nT), a reconnection regime (0>B-z> -10 nT), and a regime of reconnection saturation (B-z<-10 nT). Also, three different regimes of magnetopause dynamics at the subsolar point are descri bed by a modified pressure balance equation obtained from the ANN model.