Premium
Exploring the cross correlations and autocorrelations of the ULF indices and incorporating the ULF indices into the systems science of the solar wind‐driven magnetosphere
Author(s) -
Borovsky Joseph E.,
Denton Michael H.
Publication year - 2014
Publication title -
journal of geophysical research: space physics
Language(s) - English
Resource type - Journals
eISSN - 2169-9402
pISSN - 2169-9380
DOI - 10.1002/2014ja019876
Subject(s) - solar wind , physics , magnetosphere , earth's magnetic field , coronal mass ejection , geodesy , geophysics , plasma , geology , magnetic field , quantum mechanics
The ULF magnetospheric indices S gr , S geo , T gr , and T geo are examined and correlated with solar wind variables, geomagnetic indices, and the multispacecraft‐averaged relativistic‐electron flux F in the magnetosphere. The ULF indices are detrended by subtracting off sine waves with 24 h periods to form S grd , S geod , T grd , and T geod . The detrending improves correlations. Autocorrelation‐function analysis indicates that there are still strong 24 h period nonsinusoidal signals in the indices which should be removed in future. Indications are that the ground‐based indices S grd and T grd are more predictable than the geosynchronous indices S geod and T geod . In the analysis, a difference index ∆ S mag ≈ S grd − 0.693 S geod is derived: the time integral of ∆ S mag has the highest ULF index correlation with the relativistic‐electron flux F . In systems‐science fashion, canonical correlation analysis (CCA) is used to correlate the relativistic‐electron flux simultaneously with the time integrals of (a) the solar wind velocity, (b) the solar wind number density, (c) the level of geomagnetic activity, (d) the ULF indices, and (e) the type of solar wind plasma (coronal hole versus streamer belt): The time integrals of the solar wind density and the type of plasma have the highest correlations with F . To create a solar wind‐Earth system of variables, the two indices S grd and S geod are combined with seven geomagnetic indices; from this, CCA produces a canonical Earth variable that is matched with a canonical solar wind variable. Very high correlations ( r corr = 0.926) between the two canonical variables are obtained.