Skip to main content# Prediction scheme for an Ap index based on Fourier extrapolation.

Presentation #110.23 in the session Data Analysis Techniques Posters.

Published onSep 18, 2023

Prediction scheme for an Ap index based on Fourier extrapolation.

Romashets and Bothmer (2001) suggested that the function Ap(t) can be approximated by a sum of sin and cos with periods T, T/2, T/3.., T/N where T = 27 days and N=100. In our analysis, we use the data from 24 days preceding the moment at which the forecast should be issued. The least squares method is utilized in order to determine the coefficients a_{n} and b_{n} at sin and cos functions. The measure of success is the root mean square of the difference between actual and predicted values of the planetary index Ap in the next three days, divided by its averaged value in these days. We tested the approach on data from 2006-2023 and found that the percentage error is within the range of 15-20 %. The number of harmonics N is the free parameter, which can be adjusted, depending on level of geomagnetic activity. In quiet times, it can be kept much smaller of the order of 30.