Automatic Detection of Coronal Holes and Study their Properties

Coronal hole is the outer most part of the sun’s atmosphere where it is a region with low plasma density and open magnetic field lines in solar corona to understand its solar wind generation and space weather phenomena, this makes them to look dark region on the sun. And the causes done by sun’s radiation solar wind effect on Earth’s atmosphere and also on ground and it mainly causes some communication and appearing of aurora on North-pole and South-pole. By using pretrained machine learning model it uses neural network to detect the coronal and by using traditional method intensity- thresholding technique to detect the coronal hole using images taken from Solar Dynamics Observatory’s Atmospheric Imaging Assembly (SDO/AIA). It includes high resolution images for segmentation and parameter extraction using morphological evolution of coronal holes driven by solar magnetic. Extracting the parameters like centroid, area, latitude-longitude and intensity where analyzed after contouring the coronal hole revealing correlation with solar cycle variations in geomagnetic disturbances. Comparing the coronal hole with sunspots, it is eleven years process for analyzing the coronal hole and its properties. It helps to study about evolution of coronal hole and coronal mass ejection. Key Words: numpy, skimage, morphology, CNN, CHRONNOS, data analysis