Solar Events Tracking
Solar Events Interpolation
One of the main challenges that solar physicists face in their research is the scarcity of accurate event trajectory metadata. Even though solar images are captured relatively frequently, after the initiation if SDO’s FFT modules, the detections are not run on all solar images and thus report solar events metadata less frequently. One way to mitigate the scarcity problem is to interpolate the missing trajectory data with a given cadence based on the originally reported solar events in the HEK.
A number of interpolation strategies were designed depending on the event type to be interpolated. The simplest interpolation method is MBR-interpolation that was designed for event types that are reported using their minimum bounding rectangles such as emerging flux and flares. FI-interpolation is another interpolation method that includes the unique physical characteristics of the filament event type to make it more specialized. For most of the event types, CP-interpolation is adopted which uses centroid shape signature along with dynamic time warping alignment to deal with complex geometries. In addition to interpolating trajectory data, extrapolation is also used to estimate the shape of the geometries that do not belong to any trajectory.