Examination of physical characteristics of the Carrington event using deep learning
Abstract
We apply an image translation model, which is a popular deep learning method based on conditional Generative Adversarial Networks (cGANs), to the generation from sunspot drawings to the corresponding magnetograms. For this, we train a model using pairs of sunspot data from Debrecen Photoheliographic Data (DPD) and their corresponding SOHO/MDI & SDO/HMI magnetogram from 1996 to 2018 except for every September and October. We evaluate the model by comparing pairs of actual magnetograms and the corresponding AI-generated ones in September and October. Our results show that AI-generated magnetograms unsigned magnetic fluxes are well consistent with those of the original ones. By applying this model to the Carrington sunspot drawing, we successfully produce AI-generated magnetogram and estimate its unsigned magnetic flux. Using several empirical relationships (magnetic flux vs. CME speed, CME speed vs. ICME speed, and ICME speed vs. Dst) in 23 and 24th solar cycle, we conjecture the Dst value of the Carrington event, about 1100nT, which is similar to those from other methods. NOTE: This work was supported by Institute for Information & communications Technology Promotion(IITP) grant funded by the Korea government(MSIP) (2018-0-01422, Study on analysis and prediction technique of solar flares).
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2021
- Bibcode:
- 2021AGUFMNG45B0575L