Improving the Arrival Time Prediction of Coronal Mass Ejections using Magnetohydrodynamic Ensemble Modeling, Heliospheric Imager data and Machine Learning

Coronal mass ejections (CMEs) are responsible for extreme space weather which has many undesirable consequences to our several space-based activities. The arrival time prediction of CMEs is an area of active research. Many methods with varying levels of complexity have been developed to predict CME arrival. However, the mean absolute error in the predictions have remained above 12 hours even with the best methods. In this work, we develop a method for CME arrival time prediction that uses magnetohydrodynamic simulations of a data constrained flux rope-based CME model which is introduced in a data driven solar wind background. We found that for 6 CMEs studied in this work, the mean absolute error in arrival time was 8 hours. We further improved the arrival time predictions by using ensemble modeling and comparing the ensembles with STEREO A and B heliospheric imager data by creating synthetic J-maps from our simulations. A machine learning method called lasso regression was used for this comparison. Our mean absolute error was reduced to 4.1 hours after using this method. This is a significant improvement in the CME arrival time prediction. Thus, our work highlights the importance of using machine learning techniques in combination of other models for improving space weather predictions.