About the project:
Solar flares is a sudden flash of increased brightness on the Sun, usually observed near its surface and in close proximity to a sunspot group. It can strongly influence the space weather condition around the Earth via generating streams of highly energetic particles in the solar wind that may impact Earth's magnetosphere.
Predictions of solar flares is critical for safeguarding our technological infrastructure. Extreme space storms -- those that could significantly degrade critical infrastructure -- could disable large portions of the electrical power grid, resulting in cascading failures that would affect key services such as water supply, health care, and transportation, and cost trillions of dollars.
In this research, we utilized Sun's image data and Long-Short-Term-Memory (LSTM) model for predicting this type of rare astrophysics event.
Papers coming from this project:
Wang, X., Chen, Y., Toth, G., Manchester, W., Gombosi, T., Hero, A., Jiao, Z., Sun, H., Jin, M., Liu, Y. (2019) Predicting Solar Flares with Machine Learning: Investigating Solar Cycle Dependence. Published on AGU Fall meeting. Preprint: https://arxiv.org/abs/1912.00502
Jiao, Z., Sun, H., Wang, X., Manchester, W., Gombosi, T., Hero, A., Chen, Y. (2019) Solar Flare Intensity Prediction with Machine Learning Models. Preprint: http://arxiv.org/abs/1912.06120 (Submitted for review)
Sun, H., Manchester, W., Jiao, Z., Wang, X., Chen, Y. (2019) Interpreting LSTM prediction on Solar Flare Eruption with Time-series Clustering. Preprint: https://arxiv.org/abs/1912.12360
More details can be found on my personal website: https://sites.google.com/umich.edu/husun