Long-Term Forecasting of 10.7 cm Solar Radio Flux Using Neurodynamics

The solar radio index F_10.7 is a critical indicator of solar activity intensity. Accurate forecasting of F_10.7 is essential for advancing many fields. A promising direction for addressing such complex forecast problems is known as neurodynamics, which incorporates dynamic perspectives into neural networks. In this study, we introduce a forecast model based on neurodynamics to achieve high-precision, long-term forecasting of the F_10.7 index. First, we construct an F_10.7 dataset making up for the missing period of F_10.7 measurements by converting sunspot numbers, and we propose a new fitting method, improving the accuracy of converting sunspot number to F_10.7 index. For the forecast modeling, we employ a neurodynamics model to capture the variation characteristics of historical datasets selected by clustering. This approach enhances the objectivity of long-term F_10.7 forecasting, enabling accurate forecast spanning even an entire solar cycle. In the cycle used to validate the forecasting method, the model effectively captures the long-term trend of F_10.7 index, and the forecasted values closely match the observed values. To simplify forecasting, we develop a method for calculating F_10.7 for an entire solar cycle using only the Modified Julian Day (MJD), thereby expanding the usability of the forecasts.