[M-IS19] Atmospheric electricity

[MIS19-P04] Predication of ULF geomagnetic field based on nonlinear system identification approach

*Hayato Iibuchi¹, Hendy Santosa¹, Yasuhide Hobara¹, Michael A Balikhin², Richard Boynton² (1.The University of Electro-Communications, 2.The University of Sheffield, UK)

Keywords:ULF, Geomagnetic field variation, Nonlinear system identification, Prediction

Nonlinear Auto Regressive Moving Average Model with Exogenous Inputs (NARMAX) was applied to the time series of terrestrial ULF geomagnetic field. The one-step-ahead (OSA) prediction model was built with Orthogonal Least Square (OLS) methodology, which can unveil significant and important quantities for geomagnetic field variation. As a result, the correlation coefficient between predicted and observed values was found to be around 0.8. Moreover, the model shows a good prediction performance. Furthermore, the model suggests some controlling parameter in relation with solar activity and inner radiation belt.