Solar activity forecasting is an important topic for numerous scientific and technological areas, such as space mission operations, electric power transmission lines, power transformation stations and earth geophysical and climatic impact. Nevertheless, the well-known difficulty is how to accurately predict, on the basis of various recorded solar activity indices, the complete evolution of future solar cycles, due to highly complex dynamical and stochastic processes involved, mainly related to interaction of different components of internal magnetic fields. There are two main distinct classes of solar cycle prediction methods: the precursor-like ones and the mathematical-numerical ones. The main characteristic of precursor techniques, both purely solar and geomagnetic, is their physical basis. Conversely, the non-precursor methods use different mathematical and/or numerical properties of the known temporal evolution of solar activity indices to extract useful information for predicting future activity. For current solar cycle #24 we obtained fairly good statistical performances from both precursor and purely numerical methods, such as the so-called solar precursor and nonlinear ones. To further check the performances of these prediction techniques, we compared the early predictions for the next solar cycle #25. Preliminary results support some coherence of the prediction methods considered and confirm the current trend of a relatively low solar activity.
تحميل البحث