Rise and Shine: USU Interdisciplinary Team Receives NSF Grant Toward Predicting Solar Activity

Logan, Utah — NASA’s Apollo missions were a combination of tragedy and triumph, but a chilling hazard of which many are unaware occurred Aug. 4, 1972, halfway between the Apollo 16 crew’s return to Earth and the launch of the Apollo 17 mission.

On that date, scientists detected a major solar energetic particle event — known as an SEP event — that sent high-energy charged particles through near-Earth and cislunar space. Had radiation from such an event struck astronauts during travel to or from the moon or working on the lunar surface, the human explorers could have suffered acute radiation sickness or worse.

As NASA ramps up the Artemis program to return humans to the moon and prepare for further deep-space exploration, astronauts’ exposure to solar radiation remains a top concern.

“Solar energetic particle events are among the most intense solar events, posing serious risks to astronaut safety and space-based infrastructure, including spacecraft and satellites,” says Utah State University computer scientist Shah Muhammad Hamdi. “Therefore, predicting SEP events is a national security priority for protecting personnel and critical space-based assets.”

Hamdi, associate professor in USU’s School of Computing, along with USU colleagues Soukaina Filali Boubrahimi, computer scientist, and Ludger Scherliess, physicist, are recipients of a three-year, $750,000 Research, Innovation, Synergies and Education (RISE) grant from the National Science Foundation for the project “Synthetic Data Generation for Solar Energetic Particle Events by Multimodal Augmentation,” through which the team will create synthetic datasets of SEP events to build a robust machine-learning prediction model.

“Although SEP events pose a continuing threat, their relative rarity limits the availability of training data, making synthetic datasets essential for developing accurate prediction tools,” says Hamdi, who is principal investigator for the project.

Boubrahimi, associate professor in the School of Computing, and a co-PI with Scherliess, professor in USU’s Department of Physics, says the project’s aim is to create widely accessible predictive tools.

“Our goal is to create open-source datasets, along with a robust and accurate online predictive model that can be used by researchers and space operations personnel throughout the world,” says Boubrahimi, who is the recent recipient of a complementary NSF SHINE grant.

Scherliess, a longtime space weather researcher and PI on the NASA-funded AWE Mission, brings expertise in space physics to the project.

“In using machine learning to create synthetic data, we need to mimic reality as closely as possible,” he says. “To do this, we need to bring as much physics into this process as we can. That’s where my experience and expertise as a space physicist comes into the picture. I can help to guide development and validate the data.”

Hamdi says the grant supports a doctoral student in the School of Computing and another in the Department of Physics, who will be actively involved in the project. The project also creates opportunities for master’s level and undergraduate researchers.

“The grant program includes participation in the NSF SHINE Workshop, which will provide students with research presentation experience and opportunities to network with NASA, NOAA, and space industry professionals,” he says. “Our students will also be involved in STEM outreach at USU Physics Day at Lagoon and Science Unwrapped.”

Hamdi adds that data and models from the project will be used in USU computer science and physics courses at the Logan campus, along with classes offered at statewide campuses.

“The NSF funding enables us to provide experiential learning for students, where they will study novel machine learning model design for real-world challenges,” he says. “Advancing our ability to predict solar events and take timely action is critical, as society is now more dependent on satellite technology than at any time in the past.”

Hamdi notes past geomagnetic storm events that caused widespread technological disruption, including the March 1989 storm that led to a nine-hour blackout of Hydro-Québec’s power grid and the May 2024 Gannon storm, which increased atmospheric drag on low-Earth-orbit satellites and prompted large-scale satellite orbit-adjustment maneuvers.

“Understanding and preparing for these events is crucial, as disruptions can cause costly power outages and infrastructure damage, and also threaten the safety and lives of people traveling in the upper atmosphere and outer space,” he says. “Our efforts are aimed at advancing our ability to protect personnel and assets to keep pace with space exploration advances.”