Principal Components and Independent Component Analysis of Solar and Space Data

Abstract

Principal Components Analysis (PCA) and Independent Component Analysis (ICA) are used to identify global patterns in solar and space data. PCA seeks orthogonal modes of the two-point correlation matrix constructed from a data set. It permits the identification of structures that remain coherent and correlated or which recur throughout a time series. ICA seeks for maximally independent modes and takes into account all order correlations of the data. We apply PCA to the interplanetary-magnetic-field polarity near one AU and to the 3.25R source-surface fields in the solar corona. The rotations of the two-sector structures of these systems vary together to high accuracy during the active interval of solar cycle 23. We then use PCA and ICA to hunt for preferred longitudes in northern hemisphere, Carrington maps of magnetic fields.

20 pages, 13 figures

References in corpus (1)

• Phase Coherence Analysis of Solar Magnetic Activity

Cited by in corpus (6)

• Faster independent component analysis by preconditioning with Hessian approximations
• The Width Distribution of Loops and Strands in the Solar Corona -- Are we Hitting Rock Bottom ?
• Solar Cycle Related Changes at the Base of the Convection Zone
• Rotational quasi periodicities and the Sun - heliosphere connection
• Statistical relationship between the succeeding solar flares detected by the RHESSI satellite
• Image patch analysis of sunspots and active regions. II. Clustering via matrix factorization