In the realm of space weather forecasting, a team of researchers from the University of California, San Diego, and Predictive Science Inc. has made significant strides in improving the efficiency and accuracy of solar wind modeling. The team, comprising Reza Mansouri, Dustin Kempton, Pete Riley, and Rafal Angryk, has developed a novel approach using a Spherical Fourier Neural Operator (SFNO) to create a data-driven surrogate model for steady-state solar wind conditions.
The solar wind, a continuous flow of charged particles from the Sun’s corona, plays a crucial role in shaping the heliosphere and impacting space systems near Earth. Variations in the solar wind, such as high-speed streams and coronal mass ejections, can disrupt satellites, power grids, and communications. Accurate modeling of these variations is essential for effective space weather forecasting. However, traditional 3D magnetohydrodynamic (MHD) models are computationally expensive, limiting their usefulness in investigating the impacts of boundary condition uncertainty.
To address this challenge, the researchers developed the SFNO, a flexible and trainable approach that enables efficient real-time forecasting. The SFNO was compared to a previously developed numerical surrogate model called HUX, and the results showed that the SFNO achieved comparable or better performance across several metrics. While HUX retains advantages in physical smoothness, the researchers emphasize the need for improved evaluation criteria rather than viewing this as a flaw in the SFNO.
The practical applications of this research for the energy sector are significant. Accurate and efficient solar wind modeling can enhance space weather forecasting, which is crucial for protecting critical infrastructure such as satellites, power grids, and communication systems. By improving the ability to predict and mitigate the impacts of solar wind variations, the energy industry can ensure the reliability and resilience of its operations.
The research was published in the journal “Space Weather,” a publication of the American Geophysical Union. The source code and additional visual results are available on the project’s GitHub page, providing further insights into the methodology and findings of this innovative study.
This article is based on research available at arXiv.