Solar wind (streams of charged particles from the Sun) can affect satellites, communications, and power systems on Earth. To better understand and predict solar wind behavior, scientists need to identify where on the Sun these winds originate.

In this study, we use artificial intelligence to connect satellite images of the Sun with measurements taken in space. We combine data from NASA’s Solar Dynamics Observatory (SDO) and the Parker Solar Probe, which travels through the solar wind. Using advanced techniques like deep learning and clustering, we train a computer model to recognize patterns in the Sun’s atmosphere that relate to different types of solar wind.

Our method helps label large amounts of data efficiently and provides a scalable way to classify solar wind sources. This can improve space weather forecasting and deepen our understanding of how the Sun affects the space environment around Earth.