Team Title:
AI for Ionosphere Physics and Propagation 

Team Lead:
Haiyang Fu haiyang_fu@fudan.edu.cn

Keywords (Impact):
Climate, Electric power systems, GICs, Satellite/debris drag, Navigation and/or Communications, (Aero)space assets functions, Human exploration

Keywords (Other):
Enter my own keyword (impact):[webform_submission:values:enter_my_own_keyword_for_impact]

Keywords (Activity Type):
Understanding , Requirements , Modeling, Forecasting , Data Utilization , Information Architecture , Assessment , New Instrumentation A, Roadmap, Training

Introduction:
The research group integrates artificial intelligence with space science to advance ionospheric physics and radio wave propagation under complex and extreme solar conditions, with a particular focus on causal science of solar-ionosphere coupling. Adopting a genuine interdisciplinary approach, we fuse the latest tools from physics, mathematics, and AI to overcome fundamental bottlenecks. The ionosphere critically affects national communication, navigation, and remote sensing systems, yet its plasma parameters exhibit cross-scale correlations, nonlinearity, multiphysics coupling, and strong stochasticity—challenges that conventional models cannot adequately address. By synergizing large-scale observational data (GNSS TEC, ionosondes, satellite payloads) with advanced AI techniques (deep learning, GANs, physics-informed neural networks) and causal inference methods, our research advances fundamental understanding of solar-ionosphere interactions and provides critical technical support for mitigating space weather impacts on critical technological systems.

Objectives:
Our overarching mission is to advance AI for Space Science by establishing a new paradigm for ionospheric physics and radio wave propagation. To achieve this, we have defined four interconnected objectives:

Interdisciplinary Integration. We aim to pioneer a genuine convergence of space physics, applied mathematics, causal inference, and artificial intelligence. By breaking down traditional disciplinary boundaries, we will develop hybrid methodologies that seamlessly fuse physical principles, mathematical rigor, and data-driven learning, setting a new standard for AI-driven space science research.

Scientific Understanding. We seek to fundamentally advance causal knowledge of solar-ionosphere coupling, particularly under complex and extreme solar conditions. By integrating physics-informed neural networks with causal discovery algorithms, we aim to uncover the governing mechanisms behind ionospheric variability—from quiet periods to severe geomagnetic storms—moving beyond black-box predictions toward interpretable, explainable space weather science.

Technological Development. We target the creation of next-generation AI-powered operational tools for ionospheric modeling and propagation forecasting. These tools will synergize large-scale observational data (GNSS TEC, ionosondes, satellite payloads) with advanced techniques including deep learning, generative adversarial networks, and hybrid physical-AI models. Key deliverables include intelligent information extraction systems, real-time forecasting platforms, and simulation environments for complex electromagnetic scenarios.

Talent Cultivation. We are committed to training a new generation of space scientists who are equally fluent in physics, mathematics, and AI. Through interdisciplinary graduate programs, hands-on AI-for-Science workshops, and collaborative projects with academic, industrial, and operational partners, we will cultivate researchers capable of leading the emerging field of AI for Space Science.

Action Topics:
Understand and forecast the global state of the ionosphere, Improve the ionospheric forecasting window through the definition and forecast of ionospheric drivers in the sun, the solar wind, the magnetosphere and the lower atmosphere, Assessment of modeling capability of plasma density and temperature profiles

Cluster with overlapping topics:
G2B: Ionosphere variability

Link to external website: