Beacon2Science: Enhancing STEREO/HI beacon data with machine learning for efficient CME tracking
2025
Read Beacon2Science: Enhancing STEREO/HI beacon data with machine learning for efficient CME tracking (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-15288.txt.
Introduction
Justin Le Louëdec and co-authors upgrade in 2025 the quick-look images from the STEREO space cameras. These beacon images arrive fast, which suits early storm warnings, but they are grainy and gappy. Solar storms, called coronal mass ejections, can damage satellites and power gear, so tracking them early matters.
Their pipeline, Beacon2Science, first cleans the beacon images, then fills in missing frames to match the slower, sharper science images. Storm tracks drawn on the upgraded images landed much closer to the science tracks, halving the angle error. The method readies beacon data for future missions that will watch the Sun for storm warnings.
Catalog record
| YEAR | 2025 |
|---|---|
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 49 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2503-15288
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Beacon2Science: Enhancing STEREO/HI beacon data with machine learning for efficient CME tracking Abstract Observing and forecasting coronal mass ejections (CME) in real-time is crucial due to the strong geomagnetic storms they can generate that can have a potentially damaging effect, for example, on satellites and electrical devices. With its near-real-time availability, STEREO/HI beacon data is the perfect candidate for early forecasting of CMEs. However, previous work concluded that CME arrival prediction based on beacon data could not achieve the same accuracy as with high-resolution science data due to data gaps and lower quality. We present our novel pipeline entitled “Beacon2Science”, bridging the gap between beacon and science data to improve CME tracking. Through this pipeline, we first enhance the quality (signal-to-noise ratio and spatial resolution) of beacon data. We then increase the time resolution of enhanced beacon images through learned interpolation to match science data’s 40-minute resolution. We maximize information coherence between consecutive frames with adapted model architecture and loss functions through the different steps. The improved beacon images are comparable to science data, showing better CME visibility than the original beacon data. Furthermore, we compare CMEs tracked in beacon, enhanced beacon, and science images. The tracks extracted from enhanced beacon data are closer to those from science images, with a mean average error of ∼ 0.5 ∘ \sim 0.5^{\circ} of elongation compared to 1 ∘ 1^{\circ} with original beacon data. The work presented in this paper paves the way for its application to forthcoming missions such as Vigil and PUNCH. † † journal: Space Weather † † authors: J. Le Louëdec, M. Bauer, T. Amerstorfer, J.A. Davies, † † corresponding: Justin Le Louëdec, [email protected] Austrian Space Weather Office, GeoSphere Austria, Reininghausstraße 3, Graz, 8020, Austria Institute of Physics, University of Graz, Universitätsplatz 5, Graz, 8010, Austria RAL Space, STFC Rutherford Appleton Laboratory, Didcot, UK keypoints Near-real-time CME arrival time prediction with HI data is challenging due to the low resolution and high noise of beacon data. We propose the new “Beacon2Science” machine-learning pipeline to upsample, denoise, and increase the temporal resolution of beacon data. The improved images show better CME visibility and better tracking of the CME leading edge tracks closer to those made with science data. Plain Language Summary Coronal mass ejections (CMEs) are large eruptions of plasma from the Sun, which can disrupt satellites, power grids, and communication systems when reaching Earth. To better study and predict their arrival, the STEREO mission was sent to capture pictures of CMEs in visible light using a heliospheric imager. These images are transmitted to Earth at high quality with a delay of 3 to 4 days (so-called science data, which is not helpful for real-time prediction of a CME), or at very low quality in near-real-time called beacon data. To use this beacon data, we use novel machine-learning techniques to increase the spatial and temporal resolution, reduce the noise, and create intermediate images for smoother visualization. With improved data quality and visibility of CMEs, we track them through space toward Earth nearly as accurately as with science data. This research is important for studying CMEs, as it can be applied to improve real-time predictions of their arrival at Earth. Furthermore, we can apply our method to future space weather missions. One of them is Vigil, a planned ESA satellite that will monitor CMEs from a unique position in space (L5) and help to protect human technology and infrastructure on Earth. 1 Introduction Coronal mass ejections (CMEs) are plasma and magnetic flux eruptions from the Sun that propagate outward through the heliosphere. The interaction of the fast-expelled material with planets and their magnetospheres leads to various phenomena commonly described as space weather. While these events can be harmless, more “geo-effective” events can induce intense and destructive geomagnetic storms with substantial impacts, such as satellite orbit changes, electronic device perturbations on the planet’s surface and radio transmission disturbances. Such impacts are getting even more critical with the increasing reliance on technology and continuous power supply to various infrastructures, thus increasing the importance of real-time CME predictions Feynman & Gabriel (2000) ; Cannon (2013) ; Eastwood et al. (2018) ; Oughton et al. (2018) ; Riley et al. (2018) . Coronagraph data, i.e. white-light images from around the Sun up to 30 solar radii, are the primary data source for modeling the evolution of CMEs. However, arrival predictions based on CME characteristics extracted from these data often come with high uncertainties due to the imprecise initial parameters Singh et al. (2022) ; Verbeke et al. (2023) , even more so when limited numbers of vantage points are used as described by Kay & Palmerio (2024) . Even though coronagraph data is essential in operational space weather forecasting as it provides early initial CME parameters, integrating additional data, such as from heliospheric imagers (HI), particularly when they provide a side-view on CMEs, is essential for improving forecasting accuracy and mitigating uncertainties. Since the end of 2006, HI1 and HI2 Eyles et al. (2009) onboard the Solar TErrestrial RElations Observatory ( Kaiser et al. (2008) , STEREO;) have continuously monitored heliocentric distances between 12–215 solar radii (line-of-sight). STEREO provides science-grade data with a few days’ latency and near-real-time beacon data of lower quality. STEREO consisted of twin-spacecrafts STEREO-A(head) and STEREO-B(ehind), with STEREO-A orbiting inside the Earth’s orbit around the Sun, while STEREO-B orbiting outside Earth’s orbit such that it lags behind STEREO-A. This configuration attains an angular separation to allow for observing space weather phenomena from two distinct vantage points. Contact with STEREO-B was lost in 2014 and data was unavailable/highly limited from STEREO-A for about a year around conjunction. But STEREO-A offers close to 18 years of observations of Earth-directed space weather events. HI cameras are also found onboard of Parker Solar Probe ( Fox et al. (2016) , PSP;) with the Wide-Field Imager for Solar Probe ( Vourlidas et al. (2016) , WISPR;) and on Solar Orbiter ( Müller et al. (2020) , SolO;) with the Solar Orbiter Heliospheric Imager ( Howard et al. (2020) , SolOHI;) . However, these are only available for specific events and periods and provide no real-time data. HI data has been utilized to develop forecasting models, notably the ELlipse Evolution model based on HI ( Rollett et al. (2016) ; Amerstorfer et al. (2018) , ELEvoHI;) , which can be run such that it relies solely on the utilization of HI data without the necessity of integrating CME parameters measured from coronagraph data. However, one of the main current challenges with integrating heliospheric imager data for CME arrival time and speed prediction is the absence of acceptable quality data in real-time Tucker-Hood et al. (2015) ; Bauer et al. (2021) . While STEREO/HI data is used to improve the forecasting accuracy for well-observed CMEs using its science data, a posteriori, the beacon mode, due to low broadcast priority, suffers from four main drawbacks: lower spatial resolution, lower temporal resolution, significant data gaps, and for HI1, the usage of lossy compression. Future missions, such as Vigil, planned for launch in 2031, aim to provide continuous observations between the Sun and Earth from the fifth Lagrangian point (L5) of the Sun-Earth system, 60^{\circ} behind Earth in its orbit. This