Improving the Arrival Time Estimates of Coronal Mass Ejections by Using Magnetohydrodynamic Ensemble Modeling, Heliospheric Imager Data, and Machine Learning

Abstract

The arrival time prediction of coronal mass ejections (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 (MAE) of predictions remains above 12 hr, even with the increasing complexity of methods. In this work we develop a new method for CME arrival time prediction that uses magnetohydrodynamic simulations involving data-constrained flux-rope-based CMEs, which are introduced in a data-driven solar wind background. We found that for six CMEs studied in this work the MAE in arrival time was ∼8 hr. We further improved our arrival time predictions by using ensemble modeling and comparing the ensemble solutions with STEREO-A and STEREO-B heliospheric imager data. This was done by using our simulations to create synthetic J-maps. A machine-learning (ML) method called the lasso regression was used for this comparison. Using this approach, we could reduce the MAE to ∼4 hr. Another ML method based on the neural networks (NNs) made it possible to reduce the MAE to ∼5 hr for the cases when HI data from both STEREO-A and STEREO-B were available. NNs are capable of providing similar MAE when only the STEREO-A data are used. Our methods also resulted in very encouraging values of standard deviation (precision) of arrival time. The methods discussed in this paper demonstrate significant improvements in the CME arrival time predictions. Our work highlights the importance of using ML techniques in combination with data-constrained magnetohydrodynamic modeling to improve space weather predictions.

Space Weather with Quantified Uncertainties: Improving Space Weather Predictions with Data-Driven Models of the Solar Atmosphere and Inner Heliosphere

“…The functionality to extract data along spacecraft trajectories has been incorporated into HelioCubed. Furthermore, the Constant Turn Flux Rope CME model [3,17] has been assimilated into HelioCubed. Left panel of figure 5 shows our preliminary results for data-driven solar wind simulations performed using HelioCubed.…”

“…In [17], we developed an innovative technique for predicting CME arrival times, which utilizes magnetohydrodynamic simulations with data-informed flux-rope-based CMEs introduced into a data-driven solar wind background. For the six CMEs we analyzed, the Mean Absolute Error (MAE) in arrival time was roughly 8 hours.…”

Updating Measures of CME Arrival Time Errors

“…In addition, more recent developments include machinelearning (ML) models (e.g., Alobaid et al, 2022;Liu et al, 2018;Y. Wang et al, 2019), and it is worth remarking that there exists a class of models that may be considered "hybrid," as they bring together aspects from multiple categories (e.g., Barnard & Owens, 2022;Kay et al, 2022;Singh et al, 2023).…”

Forecasting Geoffective Events from Solar Wind Data and Evaluating the Most Predictive Features through Machine Learning Approaches

“…Recently, machine learning and deep learning techniques are increasingly being employed in SW applications (Camporeale et al 2018), especially in the prediction of solar flares (Bobra & Couvidat 2015;Barnes et al 2016;Campi et al 2019;Georgoulis et al 2021Georgoulis et al , 2024Guastavino et al 2022aGuastavino et al , 2023a, the onset of CMEs, and their arrival time on Earth (Guastavino et al 2023b;Singh et al 2023). Such studies are based on remote observations of the Sun and its atmosphere and, more specifically, on the identification of CMEs and the extraction of their morphological/dynamic properties (such as angular width, velocity, and acceleration) from time sequences of coronagraphic white-light images, such as those from the Large Angle Spectroscopic Coronagraph, as in Pricopi et al (2022) and Vourlidas et al (2019).…”