Prediction of the Magnetic Variation Field Based on Multiple Geomagnetic Components

A new method for the prediction of the geomagnetic variation time series based on multiple geomagnetic components is presented in this paper. The method used the data of multiple geomagnetic components ( H , D , Z ) with different kinds of combination ( H , Z ; H , D ; H , D , Z ) from a series of geomagnetic stations to predict the Z component (both in single-step prediction and multi-steps prediction). The comparison and analysis between the new method and the method based on single Z component are done in the paper. The results indicate that the new prediction method based on components ( Z , H ) has a higher precision than the model based on single Z component.