Machine Learning Approach for Solar Wind Categorization

Abstract

Abstract Solar wind classification is conducive to understanding the ongoing physical processes at the Sun and in solar wind evolution in interplanetary space, and, furthermore, it is helpful for early warning of space weather events. With rapid developments in the field of artificial intelligence, machine learning approaches are increasingly being used for pattern recognition. In this study, an approach from machine learning perspectives is developed to automatically classify the solar wind at 1 AU into four types: coronal‐hole‐origin plasma, streamer‐belt‐origin plasma, sector‐reversal‐region plasma, and ejecta. By exhaustive enumeration, an eight‐dimensional scheme (BT, NP, TP, VP, Nαp, Texp/TP, Sp, and Mf) is found to perform the best among 8,191 combinations of 13 solar wind parameters. Ten popular supervised machine learning models, namely, k‐nearest neighbors (KNN), Support Vector Machines with linear and radial basic function kernels, Decision Tree, Random Forest, Adaptive Boosting, Neural Network, Gaussian Naive Bayes, Quadratic Discriminant Analysis, and eXtreme Gradient Boosting, are applied to the labeled solar wind data sets. Among them, KNN classifier obtains the highest overall classification accuracy, 92.8%. Although the accuracy can be improved by 1.5% when O7+/O6+ information is additionally considered, our scheme without composition measurements is still good enough for solar wind classification. In addition, two application examples indicate that solar wind classification is helpful for the risk evaluation of predicted magnetic storms and surface charging of geosynchronous spacecraft.

Keywords

• solar wind classification
• machine learning
• automatical method
• k‐nearest neighbors
• space weather early warning