Geomagnetically correlated autoregression model forshort-term prediction of ionospheric parameters

Abstract

A new method for short-term prediction of ionospheric parametersis developed by incorporating the cross-correlation between theionospheric characteristic of interest and the Ap index intothe autocorrelation analysis. We consider the hourly timeseries of an ionospheric characteristic as composed of aperiodic component and a random component. The periodiccomponent containing the average diurnal variation is removed by using its relative deviations from the median values (Φ), which in the case of the critical frequency of the F2layer, foF2, has the form: Φ = (foF2 - foF2med)/foF2med. The geomagnetically correlated autoregression model (GCAM) is an extrapolation model based on theweighted past data. The new term in the regression equation expresses linearly the dependence of Φ on magnetic activity byintroducing a synthetic geomagnetic index G, which approximatesthe average dependence of Φ on hourly interpolated Kp.Using parametric expressions of the auto- and cross-correlationfunctions ensures the statistical sufficiency in GCAM; theparameters are then obtained by data fitting. Data from 2years of high solar activity (1981-2) and 2 years of lowsolar activity (1985-6) were used to evaluate the predictionaccuracy of GCAM. The mean square error in per cent of the1-day prediction of foF2 relative to the median shows a largegain of accuracy of GCAM in the first 8-10 h of prediction relative to the median based prediction, a diurnal variation oferrors and a steady offset of the GCAM prediction error fromthe median based prediction error. The GCAM error at the firsthour is lowest, but gradually approaches the median error witha timescale of 8-10 h. A new error estimate, called`prediction efficiency' that is a good indicator of prediction performance during disturbed ionospheric conditions is defined.