Mulitiscale 3D structure of substorm currents obtained by mining spaceborne magnetometer data
Abstract
Substorms represent a key explosive process in the space weather chain, rapidly releasing energy stored in the Earth's magnetic tail into the ionosphere. Even after several recent multi-probe missions aimed to understand substorm mechanisms, at any given moment in time the magnetosphere is still sparsely observed. Data-mining (DM) approach allows one to also utilize data during different but similar substorm events that when combined spatially populate the magnetosphere. Here, the state of the magnetosphere during a substorm is defined using phase space consisting of solar wind parameters, the global indices AL and Sym-H, and their time derivatives. Using a nearest-neighbors technique, a historical database of magnetometer data from the Geotail, Cluster, Polar, GOES 8, 9, 10, and 12, IMP-8, Van Allen Probes, THEMIS, and MMS missions is mined to form a virtual constellation of millions of spaceborne magnetometer data records, which are then ingested into a newly developed empirical magnetic field picture. This new approach also uses a regular expansion to describe the equatorial currents, a flexible description of field-aligned currents capable of reproducing their spiral structure at low altitudes, and an additional equatorial current expansion to characterize the current sheet thinning. It allows one to reconstruct the 3D global structure and evolution of the magnetic field and underlying electric currents for substorms. This includes the specific shape of the stretched nightside magnetic field in the growth phase, its dipolarization during the expansion phase, and the resulting reconfiguration of the field-aligned and ring currents. The DM approach helps also to resolve the formation of the magnetic flux accumulation in the growth phase, earthward propagation of the dipolarization signal through geosynchronous orbit, the buildup of a thick current sheet in the expansion phase, and its persistence in the recovery phase, in a drastic contrast to the thin current sheet evolution.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2018
- Bibcode:
- 2018AGUFMSM31D3534S
- Keywords:
-
- 1942 Machine learning;
- INFORMATICSDE: 7924 Forecasting;
- SPACE WEATHERDE: 7959 Models;
- SPACE WEATHERDE: 7999 General or miscellaneous;
- SPACE WEATHER