Automated Solar Feature Detection for Space Weather Applications

摘要: The solar surface and atmosphere are highly dynamic plasma environments, which evolve over a wide range of temporal spatial scales. Large-scale eruptions, such as coronal mass ejections, can be accelerated to millions kilometres per hour in matter minutes, making their automated detection characterisation challenging. Additionally, there numerous faint features, holes dimmings, important for space weather monitoring forecasting, but low intensity sometimes transient nature makes them problematic detect using traditional image processing techniques. These difficulties compounded by advances ground- space- based instrumentation, have increased the volume data that physicists confronted with on minute-by-minute basis; NASA's Solar Dynamics Observatory example is returning many thousands images (~1.5 TB/day). This chapter reviews recent application techniques active regions, holes, filaments, CMEs, dimmings purposes prediction.