Predicting CMEs using ELEvoHI with STEREO-HI beacon data

TL;DR

This study evaluates the ELEvoHI model's ability to predict CME arrival times and speeds using both high-quality science data and real-time beacon data from STEREO-A, introducing a Python tool for data handling.

Contribution

It demonstrates the effectiveness of ELEvoHI with beacon data for real-time CME prediction and provides a new Python tool for data processing.

Findings

Science data yields more accurate predictions than beacon data.

ELEvoHI achieves a mean absolute error of ~9 hours for arrival time with science data.

Beacon data predictions are less accurate but still viable in real-time scenarios.

Abstract

Being able to accurately predict the arrival of coronal mass ejections (CMEs) at Earth has been a long-standing problem in space weather research and operations. In this study, we use the ELlipse Evolution model based on Heliospheric Images (ELEvoHI) to predict the arrival time and speed of 10 CME events that were observed by HI on the STEREO-A spacecraft between 2010 and 2020. Additionally, we introduce a Python tool for downloading and preparing STEREO-HI data, as well as tracking CMEs. In contrast to most previous studies, we use not only science data, which has a relatively high spatial and temporal resolution, but also low-quality beacon data, which is - in contrast to science data - provided in real-time by the STEREO-A spacecraft. We do not use data from the STEREO-B spacecraft. We get a mean absolute error of 8.81 $\pm$ 3.18 h / 59 $\pm$ 31 kms$^{-1}$ for arrival time/speed predictions using science data and 11.36 $\pm$ 8.69 h / 106 $\pm$ 61 kms$^{-1}$ for beacon data. We find that using science data generally leads to more accurate predictions, but using beacon data with the ELEvoHI model is certainly a viable choice in the absence of higher resolution real-time data. We propose that these differences could be minimized if not eliminated altogether if higher quality real-time data was available, either by enhancing the quality of the already available data or coming from a new mission carrying a HI instrument on-board.