Expansion of ionospheric characteristics fields in empirical orthogonal functions

Abstract At the reproduction of empirical models of ionospheric characteristics we propose to use an expansion after orthogonal functions. Compared with any other system of expansion after prescribed functions this method achieves optimum convergence. It represents the variations of the given data field by a minimum number of expansion components; these are uncorrelated, i.e. linearly independent of each other. Empirical descriptive functions are defined as eigen-functions of the autocorrelation matrix of the field so that they have universal character and at the same time take account of the particularities of the given field. The method can be applied to the parameters which are mainly needed for global empirical models of the ionosphere, namely the critical frequencies of the E and F2 layers, the peak height and the semi-thickness of the F2 layer. Such models might be used as such or for updating other models such as the International Reference Ionosphere. Further discussed are two possibilities for constructing an empirical model of the full electron density profile. We deal with a finite set of data points so that either linear interpolation or splining must be carried out when intermediate parameter values are asked for. The method can, of course, only reproduce such variations as are contained in the original data set.