A new method for ionospheric short-term forecast using similar-day modeling

The conception of similar-day and its modelling of ionospheric short-term forecast are presented. The similar-day method not only has the merits of simpleness and facility, but also improves the forecast accuracy. A test of the method for 1h∼24h f
0
F
2
forecast of Manzhouli, Beijing, Urumchi and Haikou during the periods of the high level solar activities (in 1980) and a comparison with the autocorrelation method developed by LIU R show that the method is realistic and reliable. The similar-day method betters the forecast result obviously for more than N h ahead (N=1 for winter and equinoxes, N=3 for summer), and especially in winter, the relative error can be usually reduced by 2 percent or so. Finally, a new method for ionospheric short-term forecast is formed by combining the similar-day method and the autocorrelation method.