The Response of Equivalent Ionospheric Currents to the External Drivers via Spatiotemporal Prediction by Neural Network
Abstract
The geomagnetic field disturbances can be generated at high latitudes due to strong ionospheric electrojet currents during a substorm and/or a geomagnetic storm. These large magnetic disturbances, usually accompanied by a large rate-of-change in the magnetic field dB/dt, produce geoelectric fields and geomagnetically induced currents (GIC). Understanding how the magnetospheric dynamics, and the Earth's ionosphere coupled with each other is essential. However, the conventional statistical analysis is incapable of providing a quantitative prediction and reproduction of the EICs and SEC amplitudes with relatively high accuracy due to the high nonlinearity of this system. In order to study the variations of ionospheric currents (i.e., the Equivalent Ionospheric Currents (EICs) and the Spherical Elementary Current (SEC) amplitudes) and their response to external drivers, we developed an ANN-SEC model based on the feedforward neural network to reproduce the ionospheric current which is derived from the SEC technique. The data utilized are measured by multiple spacecraft and ground-based observations, and the target of the ANN-SEC model is the ionospheric current obtained from the SEC technique, e.g., each component of the EICs and the SEC amplitudes. We will demonstrate the modeling results for the Jy component to reveal the westward/eastward electrojet patterns. The input parameters include the locations of the measurements (longitude and latitude, local time, or MLT/MLAT), the geomagnetic indices (e.g., AL, AU, SYM-H, and F10.7 index). Based on our ANN-SEC model, our model is promising in predicting the Earth's ionospheric currents using the driving and coupling mechanism of the magnetosphere. Our model can provide a spatial and temporal reconstruction of the ionospheric currents within and beyond the North America and Greenland sectors.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2022
- Bibcode:
- 2022AGUFMNG51A..09C