Meridian Project Achievements Report | AI Helps Unveil the "Mysteries" of Ionospheric Changes

Abstract

The research team led by Professor Xue Xianghui from the University of Science and Technology of China, a member of the Meridian Project team, has utilized multiple ground-based ionospheric digital height meters from the Meridian Project, combined with satellite-based observation data and solar activity information, to conduct work on AI modeling and forecasting of ionospheric E-layer outbreaks. This study elucidates the complex relationship between external factors and E-layer outbreaks, revealing the main processes and controlling factors in the climatological evolution of the ionospheric E region. It is significant for exploring the structure and long-term evolution of the Earth's middle and upper atmosphere. The related work has been published in the authoritative journal Atmospheric Chemistry and Physics. The first author is Tian Penghao, and the corresponding authors are Researcher Yu Bingkun and Professor Xue Xianghui. This achievement has been recognized as an Outstanding Achievement of the Meridian Project for 2023.

Solar radiation ionizes the Earth's atmosphere, creating ionospheric irregularities due to uneven plasma distribution. At altitudes of 80 to 130 kilometers, this region experiences partial ionization under solar radiation, generating a large number of electrons and forming what is known as the ionospheric E-layer. Additionally, meteoroids entering this region from outer space produce ablated materials composed of metals such as sodium and iron. These metal components, under complex external conditions, converge to form dense E-region irregularities (outbreak E-layers), which are highly unpredictable in their formation and occurrence, and whose evolutionary mechanisms remain unclear. However, they significantly impact the accuracy and stability of space communications. Therefore, modeling and forecasting ionospheric irregularities is one of the cutting-edge scientific issues in space science.

In their previous work, the research team used empirical models to simulate global ionospheric E-layer outbreaks but lacked the capability to predict small-scale structures. In their latest efforts, they utilized ionospheric digital height meter data from the Meridian Project to preliminarily explore the effects of various driving factors on the evolution of E-region irregularities. Combining this with satellite-based observations, they proposed a more accurate climatological reconstruction model for E-region irregularities (Figure 1), effectively capturing the complex relationship between external variations and E-region irregularities (Figure 2). Comparative validation with multiple domestic ionospheric digital height meters (Beijing, Wuhan, Mohe, Sanya, and Shaoyang) further confirmed the model's effectiveness and superior forecasting capability. Additionally, the research team developed a global forecasting application based on this model (Figure 3), which provides researchers with E-region irregularity forecast information covering a complete solar activity cycle (2002-2025) and is expected to make significant contributions to early warning of extreme space weather events.

Figure 1: Schematic diagram of global ionospheric E-region irregularities reconstructed using artificial intelligence technology

Figure 2: Climatological observations and model predictions of global ionospheric E-region irregularities

Figure 3: The SELF-ANN (Sporadic E Layer Forecast using Artificial Neural Networks) application developed by the research team and the effective forecasting time range

This research provides new insights into the use of AI in space weather modeling. With the continued development of China's Meridian Project Phase II and advancements in deep space exploration, this work offers valuable model support for a deeper understanding and prediction of space environment changes. It also enhances the assessment of how the middle and upper atmosphere affects radio wave propagation, which is crucial for ensuring the stability and safety of long-distance space communications.

The research, titled Ionospheric irregularity reconstruction using multisource data fusion via deep learning, has been published in the prestigious international journal Atmospheric Chemistry and Physics. The study received funding from the Quantum Science and Technology Innovation Program, the Deep Space Exploration Laboratory's Frontier Science Research Program, the CAS Stable Support Basic Research Youth Team Program, the Ministry of Science and Technology Key R&D Program, the National Natural Science Foundation, among others. Additionally, the Supercomputing Center of the University of Science and Technology of China provided simulation support for this research.