Nonlinear analysis of the ground-based magnetometer network

Abstract

When the first magnetometer was created by Frederick Gauss in 1833, scientists gained a powerful tool for studying the structure, dynamics, and strength of the Earth's magnetic field: the magnetosphere. Since Gauss' time, the world's scientific community has established ground-based magnetometer stations around the globe in an effort to study local and global perturbations and patterns of the Earth's magnetic field. The main focus of this network has been monitoring the magnetic flux and impact from the Sun's constant outflow of radiation and particles known as the solar wind, as well as its more violent eruptive events. There has been little work, by comparison, into the signals and correlations within the network itself. Since the Earth's field can roughly be mapped to a dipole and disturbances often have a large scale structure, one can surmise there should be some correlation between stations based on their relative positions to one another. What that correlation is or represents is not clear. To investigate this possible correlation and its nature, a set of nonlinear analytic methods were conducted on magnetic data collected from stations scattered across North America over an 18 year period. The analysis was focused on searching for spatial and temporal correlations of nonperiodic signals in the magnetometer network. The findings from that analysis suggest there exist nonlocal correlations between stations that are dependent on position, which could be useful in the development of a space weather risk assessment.