Solar coronal white light image generation from SDO/AIA EUV (171,193 & 211) images using deep learning

Low coronal white light observations are very important in understanding low coronal features of the Sun but rarely made. We generate MLSO KCoronagraph-like white light images from SDO/AIA EUV images using a deep learning model based on conditional generative adversarial networks. We train (from 1.11 to 1.25 solar radii) the model using pairs of MLSO KCoronagraph images and their corresponding SDO/AIA EUV (171, 193 & 211) images from 2014 to 2019 (January to September). For this we make seven (single channels and combination of multiple channels) deep learning models for image translations. We evaluate the models by comparing the pairs of target white-light images and the corresponding AI-generated ones in October and November. Our results from this study are summarized as follows. First, the multiple channel AIA 193 & 211 model is the best among seven models in-view of metrics such as correlation coefficient and normalized root mean square error. Second, major low coronal features like helmet streamers and polar coronal holes are well identified in the AI-generated ones by this model, and their positions and sizes are consistent with those of the target ones. Third, from AI-generated images we successfully identified a few interesting phenomena: jets and CMEs. We hope that our model can provide us with complementary data to study the low coronal features in white light during non-observable cases (night-time or poor atmospheric conditions or instrumental maintenance). NOTE: This work was supported by Institute for Information & communications Technology Promotion (IITP) grant funded by the Korea government (MSIP) (2018-0-01488, study on analysis and prediction technique of solar flares).