•  
  •  
 

Bulletin of Chinese Academy of Sciences (Chinese Version)

Keywords

space science, artificial intelligence, research paradigm, space weather

Abstract

Space science is currently confronted with a triple challenge: the explosive growth of observational data, the strongly coupled cross-scale nature of physical processes, and the increasingly urgent national strategic demands. The limitations of traditional research paradigms in analytical efficiency, forecast accuracy, and autonomous capability hinder their effectiveness in meeting critical requirements such as safeguarding on-orbit satellites and ensuring the successful execution of major space missions. This study proposes a three-layer “perception–cognition–decision-making” architecture for intelligent space science. Taking space weather—a domain with strong operational relevance—as a representative case, the four-dimensional paradigm transformation driven by artificial intelligence is systematically examined across key directions, including physics-constrained modeling, explainable causal inference, autonomous agent applications, and satellite–ground collaborative observation. The profound scientific and technical challenges confronting the intelligent transformation and strategic initiatives are further explored for the holistic advancement of space science. This work aims to provide actionable technical pathways and policy frameworks for China to seize the initiative and achieve leadership in the global race of intelligent space science.

First page

1169

Last Page

1181

Language

Chinese

Publisher

Bulletin of Chinese Academy of Sciences

References

[1] Buzulukova N, Tsurutani B. Space weather: From solar origins to risks and hazards evolving in time. Frontiers in Astronomy and Space Sciences, 2022, 9: 1017103.

[2] Hands A D P, Ryden K A, Meredith N P, et al. Radiation effects on satellites during extreme space weather events. Space Weather, 2018, 16(9): 1216-1226.

[3] Elvidge S. Space weather can affect our daily lives—We need a better warning system. Nature, 2024, 630: 271.

[4] Camporeale E. The challenge of machine learning in space weather: Nowcasting and forecasting. Space Weather, 2019, 17(8): 1166-1207.

[5] 白春礼. 人工智能促进科学研究范式变革. 中国科学院院刊, 2025, 40(2): 189-196. Bai C L. Artificial intelligence promotes the transformation of scientific research paradigms. Bulletin of Chinese Academy of Sciences, 2025, 40(2): 189-196. (in Chinese)

[6] Nagashima S, Sugiura K. Deep space weather model: Long-range solar flare prediction from multi-wavelength images. arXiv, 2025: 2508.07847.

[7] 中国科学院学部人工智能咨询研究组. 人工智能赋能科学研究：人工智能学科体系. 北京: 科学出版社, 2026. Academic Divisions of the Chinese Academy of Sciences, AI Consultation Research Group. Artificial Intelligence Empowering Scientific Research: The Discipline System of Artificial Intelligence. Beijing: Science Press, 2026. (in Chinese)

[8] Ghasemi N, Alvarez Justo J, Celesti M, et al. Onboard processing of hyperspectral imagery: Deep learning advancements, methodologies, challenges, and emerging trends. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2025, 18: 4780.

[9] 徐帆江, 周鑫, 赵军锁, 等. 软件定义卫星技术概念及发展. 北京航空航天大学学报, 2023, 49(7): 1543-1552. Xu F J, Zhou X, Zhao J S, et al. Concept and development of software-defined satellite technology. Journal of Beijing University of Aeronautics and Astronautics, 2023, 49(7): 1543-1552. (in Chinese)

[10] Karniadakis G E, Kevrekidis I G, Lu L, et al. Physics-informed machine learning. Nature Reviews Physics, 2021, 3(6): 422-440.

[11] Raissi M, Perdikaris P, Karniadakis G E. Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations. Journal of Computational Physics, 2019, 378: 686-707.

[12] Lu L, Jin P, Pang G, et al. Learning nonlinear operators via DeepONet based on the universal approximation theorem of operators. Nature Machine Intelligence, 2021, 3(3): 218-229.

[13] Jarolim R, Thalmann J K, Veronig A M, et al. Probing the solar coronal magnetic field with physics-informed neural networks. Nature Astronomy, 2023, 7(10): 1171-1179.

[14] 孙泽翼, 王彬, 胡馨月, 等. 深空探测器多智能体强化学习自主任务规划. 深空探测学报(中英文）, 2024, 11(3): 244-255. Sun Z Y, Wang B, Hu X Y, et al. Autonomous mission planning for deep space probes based on multi-agent reinforcement learning. Journal of Deep Space Exploration, 2024, 11(3): 244-255. (in Chinese)

[15] Aminalragia-Giamini S, Raptis S, Anastasiadis A, et al. Solar energetic particle event occurrence prediction using solar flare soft X-ray measurements and machine learning. Journal of Space Weather and Space Climate, 2021, 11: 59.

[16] Asensio Ramos A, Cheung M C M, Chifu I, et al. Machine learning in solar physics. Living Reviews in Solar Physics, 2023, 20(1): 4.

[17] Yang Y, Shen F. Modeling the global distribution of solar wind parameters on the source surface using multiple observations and the artificial neural network technique. Solar Physics, 2019, 294(8): 111.

[18] Roy S, Schmude J, Lal R, et al. Surya: Foundation model for heliophysics. arXiv, 2025: 2508.14112.

[19] Wang T Y, Luo B X, Wang J J, et al. Forecasting of the geomagnetic activity for the next 3 days utilizing neural networks based on parameters related to large-scale structures of the solar corona. Space Weather, 2025, 23(2): e2024SW004090.

[20] 王听雨, 罗冰显, 陈艳红, 等. 基于神经网络的未来3天Kp指数预报建模与可解释AI应用. 空间科学学报, 2024, 44(3): 437-445. Wang T Y, Luo B X, Chen Y H, et al. Modeling next 3-day Kp index forecasting with neural networks and exploring the application of explainable AI. Chinese Journal of Space Science, 2024, 44(3): 437-445. (in Chinese)

[21] Pandey C, Angryk R A, Georgoulis K M, et al. Explainable deep learning-based solar flare prediction with post hoc attention for operational forecasting. Discovery Science, 2023, 14276: 567-581.

[22] Wang J, Liang P, Wang T, et al. Large model driven solar activity AI forecaster: A scalable dual data-model framework. arXiv, 2025: 2508.06892.

[23] Li H, Wang C, Tu C, et al. Machine learning approach for solar wind categorization. Earth and Space Science, 2020, 7(5): e2019EA000997.

[24] 李盛阳, 刘康, 刘云飞, 等. 数智驱动的空间科学实验研究：AI4S范式下的新探索. 中国科学院院刊, 2025, 40(2): 371-379. Li S Y, Liu K, Liu Y F, et al. Data-intelligence-driven space science experimental research: New exploration under the AI4S paradigm. Bulletin of Chinese Academy of Sciences, 2025, 40(2): 371-379. (in Chinese)

[25] Hui Y C, Cheng Y H, Jiang B, et al. A full-lifecycle health monitoring and prognostics method for satellite components based on data distribution features and multi-kernel phase-adaptive Gaussian processes. Aerospace Science and Technology, 2026, 168: 111068.

[26] 王赤. 当AI算法遇上太空算法：对话全国人大代表、中科院院士王赤. 北京周报, 2026-03-23(03). Wang C. When AI algorithms meet space algorithms: A dialogue with Wang Chi, NPC deputy and CAS academician. Beijing Review, 2026-03-23(03). (in Chinese)

[27] Song P, Wang C. Can ChatGPT replace scientists? Science Bulletin, 2023, 68(19): 2128-2131.

[28] Mavromichalaki H, Papailiou M, Livada M, et al. Unusual Forbush decreases and geomagnetic storms on 24 March 2024 and 11 May 2024. Atmosphere, 2024, 15(9): 1033.

[29] 凌宗成, 李勃, 魏广飞, 等. 行星探测特征信息提取与知识挖掘关键技术及应用研究. 数据与计算发展前沿, 2025, 7(4): 3-19. Ling Z C, Li B, Wei G F, et al. Key technologies and applications of feature information extraction and knowledge mining for planetary exploration. Frontiers of Data & Computing, 2025, 7(4): 3-19. (in Chinese)

[30] 于登云, 张哲, 泮斌峰, 等. 深空探测人工智能技术研究与展望. 深空探测学报（中英文）, 2020, 7(1): 11-23. Yu D Y, Zhang Z, Pan B F, et al. Research and prospect of artificial intelligence technology for deep space exploration. Journal of Deep Space Exploration, 2020, 7(1): 11-23. (in Chinese)

[31] 崔平远, 徐瑞, 朱圣英, 等. 深空探测器自主技术发展现状与趋势. 航空学报, 2014, 35(1): 13-28. Cui P Y, Xu R, Zhu S Y, et al. Development status and trends of autonomous technology for deep space probes. Acta Aeronautica et Astronautica Sinica, 2014, 35(1): 13-28. (in Chinese)

Download
Request a Copy

Share

COinS