Neural Network Recovery of Gaps in Geomagnetic Field Records

Abstract

The paper demonstrates the capabilities of neural network recovery of ground-based geomagnetic field records at a selected magnetic station using similar magnetic field data at another station. By the example of the restoration of disturbance records made at the magnetic stations Kakioka, Kanoya, Alma-Ata, Hermanius, San Juan, Tucson, Honolulu with and without data from the OMNI satellite system on the parameters of the solar wind and interplanetary magnetic field, it is shown that the technique of artificial neural networks can to successfully fill in the gaps and failures in the records of individual observatories of the global network of magnetic observation stations. The created artificial neural network tool can be used for scientific and applied problems of geomagnetic information recovery.