Statistical comparison of different methods for filtering solar activity in foF2 and hmF2 for long-term analysis at Juliusruh station

Abstract

The ionospheric F2 layer plays a crucial role in radio wave propagation and is significantly influenced by various factors. Understanding its long-term variations is essential for analyzing Solar-Terrestrial dynamics and improving ionospheric models. This study uses the critical frequency of the F2 layer (foF2) and the height of the peak electron density (hmF2) from 1964 to 2019. Both parameters are affected by solar activity; thus, modeling their response to solar activity is essential. An analysis of extreme ultraviolet (EUV) proxies, including Mg II, F30, and F10.7, was carried out to identify the proxy with the highest correlation with the ionospheric parameters. Furthermore, a comparison of five methods for modeling the impact of solar activity on foF2 and hmF2 was conducted using hourly data from Juliusruh. The methods considered were three polynomial regressions based on data clustered by solar cycle, month and hour of the day, and two Fourier Series applied to data clustered by hour of the day. Finally, statistical tools such as the coefficient of determination (R²), mean absolute percentage error (MAPE), Akaike Information Criterion (AIC), and Bayesian Information Criterion (BIC) were utilized to select the most appropriate method. The results show that F30 is the EUV proxy with the highest correlation with foF2, while for hmF2 is MgII. The third-degree polynomial regression was found to be the most effective method for modeling the ionospheric response to solar activity. This analysis only makes a statistical comparison, and further exploration is needed to assess the explanatory power of each method.