Identifying Flux Rope Signatures Using a Deep Neural Network

Abstract:

Among the current challenges in Space Weather, one of the main ones is to forecast the internal magnetic configuration within Interplanetary Coronal Mass Ejections (ICMEs). Currently, a monotonic and coherent magnetic configuration observed is associated with the result of a spacecraft crossing a large flux rope with helical magnetic field lines topology. The classification of such an arrangement is essential to predict geomagnetic disturbance. Thus, the classification relies on the assumption that the ICME's internal structure is a well organized magnetic flux rope. This work applies machine learning and a current physical flux rope analytical model to identify and further understand the internal structures of ICMEs. We trained an image recognition artificial neural network with analytical flux rope data, generated from the range of many possible trajectories within a cylindrical (circular and elliptical cross-section) model. The trained network was then evaluated against the observed ICMEs from WIND during 1995-2015.

The results demonstrate that the approach works. We were able to identify flux rope signatures using a pre-established Deep Neural Network handwriting model trained with synthetic data with high accuracy in well-behaved events. The methodology developed in this work can classify 84\% of simple real cases correctly and has a 76\% success rate when extended to a broader set. We have analyzed the discrepancies between manual and machine-learning-based classification to develop the classification further. As a first step towards a generalizable classification and parameterization tool, these results show promise. By improving synthetic fluctuations and adding more complex structures, our model has a strong potential to evolve into a robust tool for identifying flux rope configurations from in situ data.