10:45 AM - 12:15 PM
[PEM12-P30] Extraction of Magnetic Field Variations excited by Atmospheric Tides with Independent Component Analysis
Keywords:Atmospheric wave, Independent component analysis, Atmospheric tide
Ground magnetic field data are complexly superimposed the Sq-EEJ current system induced by solar radiation and currents excited by various modes of atmospheric tides. Therefore, no method has yet been established to separate each phenomenon from the observed data.
In this study, we used Independent Component Analysis (ICA), which separates and extracts independent components from multivariate signals, to extract variations due to atmospheric waves from ground magnetic field data of the MAGnetic Data Acqusition System/Circum-pan Pacific Magnetometer Network (MAGDAS/CPMN) on magnetically quiet days for 1992-2004. As a first step, we reduce the dimension of the data with Principal Component Analysis (PCA) to remove noise. Since [Gurubaran, 2002] suggests that the first principal component is the Sq-EEJ current system and atmospheric tidal variations appear in the addition of the 2nd through 5th principal components. Therefore, we applied ICA to the 2nd through 5th principal components as four variables. Also, we visualize the horizontal current structure from each independent component as equivalent current. We report that which mode has a strong influence on the ionospheric current.
In the future, we will reproduce each atmospheric tidal mode influenced on the ionospheric currents with an atmosphere-ionosphere coupled model (GAIA) and compare with the current structure of each independent component.
In this study, we used Independent Component Analysis (ICA), which separates and extracts independent components from multivariate signals, to extract variations due to atmospheric waves from ground magnetic field data of the MAGnetic Data Acqusition System/Circum-pan Pacific Magnetometer Network (MAGDAS/CPMN) on magnetically quiet days for 1992-2004. As a first step, we reduce the dimension of the data with Principal Component Analysis (PCA) to remove noise. Since [Gurubaran, 2002] suggests that the first principal component is the Sq-EEJ current system and atmospheric tidal variations appear in the addition of the 2nd through 5th principal components. Therefore, we applied ICA to the 2nd through 5th principal components as four variables. Also, we visualize the horizontal current structure from each independent component as equivalent current. We report that which mode has a strong influence on the ionospheric current.
In the future, we will reproduce each atmospheric tidal mode influenced on the ionospheric currents with an atmosphere-ionosphere coupled model (GAIA) and compare with the current structure of each independent component.