SOLAR WIND
PATHWAYS

Line_Lt_Dk_Olive.gif (228 bytes)

Back to...
MAIN RESEARCH PAGE

Line_Lt_Dk_Olive.gif (228 bytes)

Back to...
MAIN PAGE

        RESEARCH...

Solar Wind Coupling

Solar wind coupling is the name applied to empirical studies of the relation between phenomena on the Sun and in the solar wind to various measures of geomagnetic activity. This field of research often treats the earth's magnetosphere as a black box whose properties can be ascertained from records of its past behavior. These properties are then used in conjunction with measurements of the Sun and solar wind to predict future geomagnetic activity. Such predictions are frequently used in empirical methods of space weather forecasting. Measures of geomagnetic activity are called magnetic indices. The most popular indices include the planetary range index, Kp, the polar cap index, PC, the auroral electrojet index, AE, and the ring current index, Dst. Many important phenomena have been correlated with these indices and if the values of the indices are know, the correlations can be used to estimate or predict their development. For example, the electric potential applied to the polar cap by the solar wind is linearly related to the Kp index. Magnetic indices have been related to the solar wind dynamic pressure and electric field by a variety of methods that include linear regression, cross correlation functions, linear prediction filters, neural networks, and local linear prediction filters.

Prof. McPherron's group was one of the first to recognize that techniques such as linear prediction filtering could be used to encapsulate the behavior of the magnetosphere for prediction of activity measures, and that these encapsulations could be used to study the physical properties of the solar wind-magnetosphere system. A major contribution to this field was the work of Bargatze et al. [button to IGPP List] which established that the response of the westward auroral electrojet (substorm associated ionospheric current) to the solar   wind was bimodal within certain ranges of

geomagnetic activity. This then led to the modern idea that the magnetosphere is a nonlinear dynamic system driven by the solar wind. Current research in this field is focused on the idea that the properties of the magnetosphere are locally linear in time, but that as the internal state of the system changes, the transfer function between the solar wind and the magnetosphere changes. Past records of activity are used in real time to update the system transfer function. Recent contributions by McPherron's group include a review paper [button to IGPP List] summarizing the results of linear prediction filtering studies, and a paper with one of Prof. McPherron's principle collaborators, Dr. Dan Baker of the Laboratory for Atmospheric and Space Physics of the Univ. of Colorado on the factors controlling the intensity of magnetospheric substorms [button to IGPP List]. A graduate student in Prof. McPherron's group, Mr. Gerard Blanchard, has investigated how well a bimodal model of the system transfer function represents individual substorms and finds it is a remarkably good representation [button to IGPP List]. This result suggest that the magnetosphere behaves as a low dimensional system and that its behavior should be predictable if the proper differential equations can be found. Prof. Gene Stringer of Southern Oregon State College, a NASA JOVE (Joint Venture) scholar working with Prof. McPherron, is completing a manuscript describing the use of neural networks to predict the flux of relativistic electrons at synchronous orbit from the Kp index. These electrons are a major cause of failure in electronic systems within synchronous satellites.