IBM and NASA have unveiled the most advanced open-source foundational model designed to understand high-resolution solar observational data and predict how solar activity affects Earth and space technology. Surya, named after the Sanskrit word for the Sun, represents a significant advancement in the application of AI to solar image interpretation and space weather research, providing a novel tool to help protect everything from GPS navigation to power grids and telecommunications from the ever-changing nature of the Sun.
The Sun may be 93 million miles away , but its impact on modern life is immediate and growing. Solar flares and coronal mass ejections can destroy satellites, disrupt air travel, cause blackouts, and pose serious radiation risks to astronauts. With humanity’s increasing reliance on space technology and plans for deeper space exploration, accurate prediction of solar weather has become critical.
As humanity’s technological dependence grows, so does our vulnerability to space weather. According to a systemic risk scenario created by Lloyd’s , the global economy could be exposed to losses of $2.4 trillion over a five-year period, with an expected loss of $17 billion from the threat of a hypothetical solar storm. Recent solar events have already demonstrated the risk, disrupting GPS services, forcing flight diversions, and damaging satellites. The effects of solar storms can cause:
- Damage to satellites, spacecraft and/or astronauts stationed outside Earth
- Loss of satellite hardware, damage to solar panels and circuits
- Impact on air travel, due to navigation errors and potential radiation risk to airline crew and passengers
- A reduction in food production as agriculture may be affected by the disruption of GPS navigation
The implications include both academic research and operational preparation. The new model will provide tools to help experts plan for solar storms, which can disrupt Earth’s technological infrastructure.
“Think of this as a weather forecast for space,” said Juan Bernabé-Moreno, Director of IBM Research Europe, UK and Ireland. “Just as we work to prepare for dangerous weather events, we must do the same with solar storms. Surya gives us an unprecedented ability to anticipate what’s coming, and it’s not just a technological achievement, but a fundamental step toward protecting our technological civilization from the star that sustains us.”
Traditional solar weather prediction relies on partial satellite images of the Sun’s surface, historically making accurate forecasts extremely difficult. Surya addresses this limitation by being trained on the largest high-resolution dataset ever curated. This dataset is designed to help researchers better study and evaluate critical space weather prediction tasks. Some examples of these tasks Surya has been tested on include predicting solar flares, solar wind speeds, solar EUV spectra, and the occurrence of active regions on the Sun.
In initial tests, researchers reported achieving a 16% improvement in solar flare classification accuracy, which the researchers consider a very substantial improvement compared to previous methods. In addition to the task of binary solar flare classification, Surya is designed to visually predict solar flares for the first time, providing a high-resolution image of where the flare is predicted to occur up to two hours in advance.
The technical challenges were immense. Surya was trained with nine years of high-resolution solar observation data from NASA’s Solar Dynamics Observatory. These solar images are 10 times larger than typical AI training data, requiring a custom multi-architecture solution to handle the massive scale while maintaining efficiency. The result is a model with unprecedented spatial resolution that can resolve solar features at scales and contexts not previously captured in large-scale AI training workflows.
“We are advancing data-driven science by integrating NASA’s deep scientific expertise into cutting-edge artificial intelligence models,” said Kevin Murphy, director of data science at NASA Headquarters in Washington. “By developing a foundational model based on NASA heliophysics data, we are making it easier to analyze the complexities of the Sun’s behavior with unprecedented speed and accuracy. This model enables a broader understanding of how solar activity impacts the critical systems and technologies we all rely on here on Earth.”
Surya is part of a broader effort at IBM to embrace generative and automated approaches to discover, test, and develop algorithms at scale. Surya is an example of how IBM is positioning AI not just as a tool, but as an engine of scientific discovery. By launching Surya on Hugging Face , IBM and NASA are democratizing access to advanced tools for understanding and forecasting solar weather and scientific exploration. Now researchers around the world can build on this foundation to develop specialized applications for their regions and industries.
This model is part of a broader collaboration between IBM and NASA to use AI technology to explore our planet and solar system. It joins the Prithvi family of foundational models, which includes a geospatial model and a weather model. Last year, IBM and NASA released the Prithvi weather model on Hugging Face for scientists and the broader community to develop short- and long-term weather and climate projections.
