Premium
Underlying scaling relationships between solar activity and geomagnetic activity revealed by multifractal analyses
Author(s) -
Yu ZuGuo,
Anh Vo,
Eastes Richard
Publication year - 2014
Publication title -
journal of geophysical research: space physics
Language(s) - English
Resource type - Journals
eISSN - 2169-9402
pISSN - 2169-9380
DOI - 10.1002/2014ja019893
Subject(s) - multifractal system , scaling , earth's magnetic field , physics , series (stratigraphy) , statistical physics , mathematics , geometry , mathematical analysis , magnetic field , fractal , geology , quantum mechanics , paleontology
This paper identifies some scaling relationships between solar activity and geomagnetic activity. We examine the scaling properties of hourly data for two geomagnetic indices ( a p and A E ), two solar indices (solar X‐rays X l and solar flux F 10.7), and two inner heliospheric indices (ion density N i and flow speed V s ) over the period 1995–2001 by the universal multifractal approach and the traditional multifractal analysis. We found that the universal multifractal model (UMM) provides a good fit to the empirical K ( q ) and τ ( q ) curves of these time series. The estimated values of the Lévy index α in the UMM indicate that multifractality exists in the time series for a p , AE , X l , and N i , while those for F 10.7 and V s are monofractal. The estimated values of the nonconservation parameter H of this model confirm that these time series are conservative which indicate that the mean value of the process is constant for varying resolution. Additionally, the multifractal K ( q ) and τ ( q ) curves, and the estimated values of the sparseness parameter C 1 of the UMM indicate that there are three pairs of indices displaying similar scaling properties, namely a p and X l , AE and N i , and F 10.7 and V s . The similarity in the scaling properties of pairs ( a p , X l ) and ( A E , N i ) suggests that a p and X l , AE and N i are better correlated—in terms of scaling—than previous thought, respectively. But our results still cannot be used to advance forecasting of a p and AE by X l and N i , respectively, due to some reasons.