Premium
Decadal period external magnetic field variations determined via eigenanalysis
Author(s) -
Shore R. M.,
Whaler K. A.,
Macmillan S.,
Beggan C.,
Velímský J.,
Olsen N.
Publication year - 2016
Publication title -
journal of geophysical research: space physics
Language(s) - English
Resource type - Journals
eISSN - 2169-9402
pISSN - 2169-9380
DOI - 10.1002/2015ja022066
Subject(s) - empirical orthogonal functions , spherical harmonics , amplitude , inversion (geology) , earth's magnetic field , geodesy , latitude , harmonic analysis , climatology , geology , physics , magnetic field , mathematics , mathematical analysis , seismology , quantum mechanics , tectonics
We perform a reanalysis of hourly mean magnetic data from ground‐based observatories spanning 1997–2009 inclusive, in order to isolate (after removal of core and crustal field estimates) the spatiotemporal morphology of the external fields important to mantle induction, on (long) periods of months to a full solar cycle. Our analysis focuses on geomagnetically quiet days and middle to low latitudes. We use the climatological eigenanalysis technique called empirical orthogonal functions (EOFs), which allows us to identify discrete spatiotemporal patterns with no a priori specification of their geometry—the form of the decomposition is controlled by the data. We apply a spherical harmonic analysis to the EOF outputs in a joint inversion for internal and external coefficients. The results justify our assumption that the EOF procedure responds primarily to the long‐period external inducing field contributions. Though we cannot determine uniquely the contributory source regions of these inducing fields, we find that they have distinct temporal characteristics which enable some inference of sources. An identified annual‐period pattern appears to stem from a north‐south seasonal motion of the background mean external field distribution. Separate patterns of semiannual and solar‐cycle‐length periods appear to stem from the amplitude modulations of spatially fixed background fields.