Supporting Informed Shutter Operations for SMILE's SXI Using Ensemble Classifiers for Soft Proton High Flux Period Predictions
Space Weather · 2026-07 · 24 (7)
DOI: 10.1029/2025sw004832 ↗Kurz gefasst
Fachartikel in der Zeitschrift Space Weather, online erschienen am 21.07.2026. Verfasst von Simon Mischel und 2 weiteren Autor:innen. Die Arbeit nennt 58 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: European Space Agency, Deutsche Forschungsgemeinschaft.
Abstract (Originalsprache)
Satellite X‐ray imagers are highly vulnerable to contamination by soft protons (∼100 keV), which can trigger spurious charge‐coupled device events and force interruptions in observations. For XMM‐Newton, such contamination reduced usable science time by ∼40%. The forthcoming Solar‐wind Magnetosphere Ionosphere Link Explorer (SMILE) mission will face similar challenges, but uniquely includes an operable shutter on its Soft X‐ray Imager (SXI) to protect the instrument during high‐radiation intervals. Leveraging this capability requires predictions that directly answer a binary operational question: will soft‐proton flux exceed mission‐relevant thresholds? We recast soft‐proton modeling as a probabilistic classification problem with two different preliminary decision thresholds. Using ∼14 years of Cluster/research with adaptive particle imaging detector proton measurements filtered through a SMILE‐specific orbital mask, we trained logistic regression and eXtreme Gradient Boosting (XGBoost) classifiers and combined them in an ensemble framework. The ensemble achieves stable performance, correctly identifying ∼75% of high‐risk and ∼78% of low‐risk cases on unseen data. Synthetic orbital simulations show that SMILE's high‐latitude apogee vantage point is rarely subject to exceedances, supporting largely uninterrupted imaging of the dayside system. Evaluation of radial distance‐dependent shutter strategies further shows that adopting a closure threshold near reduces predicted high‐risk exposure under average conditions to below 2% while maintaining approximately 75% overall SXI uptime. These results establish a mission‐specific forecasting framework that integrates machine learning with operational constraints, providing a transferable approach to protect sensitive space‐based imagers in dynamic radiation environments.
Förderung
- European Space Agency
- Deutsche Forschungsgemeinschaft
Zitieren
APA
Mischel, S., Kronberg, E. A., & Escoubet, C. P. (2026). Supporting Informed Shutter Operations for SMILE's SXI Using Ensemble Classifiers for Soft Proton High Flux Period Predictions. Space Weather, 24(7). https://doi.org/10.1029/2025sw004832
Vancouver
Mischel S, Kronberg EA, Escoubet CP. Supporting Informed Shutter Operations for SMILE's SXI Using Ensemble Classifiers for Soft Proton High Flux Period Predictions. Space Weather. 2026;24(7). doi:10.1029/2025sw004832
BibTeX
@article{Mischel2026,
author = {Mischel, Simon and Kronberg, Elena A. and Escoubet, C. P.},
title = {Supporting Informed Shutter Operations for SMILE's SXI Using Ensemble Classifiers for Soft Proton High Flux Period Predictions},
journal = {Space Weather},
year = {2026},
volume = {24},
number = {7},
doi = {10.1029/2025sw004832}
}
EUVIMED