Using neural networks to determine the optimum solar input for the prediction of ionospheric parameters Original Research Article

Neural networks (NNs) are proving to be ideal tools for modeling the behaviour of the ionosphere. The NNs are trained using a database of archived data describing the relationship between the output parameter and an input space. The input space is designed from knowledge of those variables that affect the behaviour of the output parameter. For ionospheric parameters this input space would always include a solar variable due to the strong influence that the sun has on ionospheric behaviour. This paper uses the critical frequency of the F2 layer, foF2, which provides an indication of the ionospheric maximum electron density, to demonstrate how NNs can be used to determine the optimum solar input variable for use in predicting foF2. Amongst the solar inputs used are the daily sunspot number, the F10.7 cm solar radio flux, and the solar irradiance. Varying averaging time periods, of 1 day to 12 months, of these solar parameters are also investigated. The criteria used in determining the optimum input are the root mean square (RMS) error between the measured and predicted output parameters. Discussions on the extension of this technique to other ionospheric parameters will also be included. This paper will interest the International Reference Ionosphere (IRI) community since the use of the 12 month running mean sunspot number in the global IRI model has been the subject of debates in the past.