In this study, we present the application of the Convolutional Neural Network (CNN) to the forecast of solar flare occurrence. For this, we consider three CNN models (two pretrained models, AlexNet and GoogLeNet, and one newly proposed model). Our inputs are SOHO/Michelson Doppler Imager (from 1996 May to 2010 December) and SDO/Helioseismic and Magnetic Imager (from 2011 January to 2017 June) full-disk magnetograms at 00:00 UT. Model outputs are the “Yes or No” of daily flare occurrence (C, M, and X classes) and they are compared with GOES observations. We train the models using the input data and observations from 1996 to 2008, covering the entire solar cycle 23, and test them using the data sets from 2009 to 2017, covering solar cycle 24. Then we compare the results of the CNN models with those of three previous flare forecast models in view of statistical scores. The major results from this study are as follows. First, we successfully apply CNN to the full-disk solar magnetograms without any preprocessing or feature extraction. Second, the results of our CNN models are slightly better in Heidke skill score and true skill statistics, and considerably better in false alarm ratio (FAR) and critical success index than the previous solar flare forecasting models. Third, our proposed model has better values of all statistical scores except for FAR, than the other two pretrained models. Our results indicate a sufficient possibility that deep learning methods can improve the capability of the solar flare forecast as well as similar types of forecast problems.
Solar magnetic fields play a key role in understanding the nature of the coronal phenomena. Global coronal magnetic fields are usually extrapolated from photospheric fields, for which farside data is taken when it was at the frontside, about two weeks earlier. For the first time we have constructed the extrapolations of global magnetic fields using frontside and artificial intelligence (AI)-generated farside magnetic fields at a near-real time basis. We generate the farside magnetograms from three channel farside observations of Solar Terrestrial Relations Observatory (STEREO) Ahead (A) and Behind (B) by our deep learning model trained with frontside Solar Dynamics Observatory extreme ultraviolet images and magnetograms. For frontside testing data sets, we demonstrate that the generated magnetic field distributions are consistent with the real ones; not only active regions (ARs), but also quiet regions of the Sun. We make global magnetic field synchronic maps in which conventional farside data are replaced by farside ones generated by our model. The synchronic maps show much better not only the appearance of ARs but also the disappearance of others on the solar surface than before. We use these synchronized magnetic data to extrapolate the global coronal fields using Potential Field Source Surface (PFSS) model. We show that our results are much more consistent with coronal observations than those of the conventional method in view of solar active regions and coronal holes. We present several positive prospects of our new methodology for the study of solar corona, heliosphere, and space weather.
In this Letter, we apply deep-learning methods to the image-to-image translation from solar magnetograms to solar ultraviolet (UV) and extreme UV (EUV) images. For this, We consider two convolutional neural network models with different loss functions, one (Model A) is with L1 loss (L 1), and the other (Model B) is with L 1 and cGAN loss (L cGAN). We train the models using pairs of Solar Dynamics Observatory (SDO)/Atmospheric Imaging Assembly (AIA) nine-passband (94, 131, 171, 193, 211, 304, 335, 1600, and 1700 Å) UV/EUV images and their corresponding SDO/Helioseismic and Magnetic Imager (HMI) line-of-sight (LOS) magnetograms from 2011 to 2016. We evaluate the models by comparing pairs of SDO/AIA images and the corresponding ones generated in 2017. Our main results from this study are as follows. First, the models successfully generate SDO/AIA-like solar UV and EUV images from SDO/HMI LOS magnetograms. Second, in view of three metrics (pixel-to-pixel correlation coefficient, relative error, and the percentage of pixels having errors less than 10%), the results from Model A are mostly comparable or slightly better than those from Model B. Third, in view of the rms contrast measure, the generated images by Model A are much more blurred than those by Model B because of L cGAN specialized for generating realistic images.
In this Letter, we generate realistic high-resolution (1024 × 1024 pixels) pseudo-magnetograms from Ca ii K images using a deep learning model based on conditional generative adversarial networks. For this, we consider a model “pix2pixHD” that is specifically devised for high-resolution image translation tasks. We use Ca ii K 393.3 nm images from the Precision Solar Photometric Telescope at the Rome Observatory and line-of-sight magnetograms from the Helioseismic and Magnetic Imager (HMI) at the Solar Dynamics Observatory from 2011 January to 2015 June. 2465 pairs of Ca ii K and HMI are used for training except for January and July data. The remaining 436 pairs are used for an evaluation of the model. Our model shows that the mean correlation coefficient (CC) of total unsigned magnetic flux between AI-generated and real ones is 0.99 and the mean pixel-to-pixel CC after 8 × 8 binning over the full disk is 0.74. We find that the AI-generated absolute magnetic flux densities are highly consistent with real ones, even to the fine scale structures of quiet regions. On the other hand, the mean pixel-to-pixel correlations of magnetic flux densities strongly depend on a region of interest: 0.81 for active regions and 0.24 for quiet regions. Our results suggest a sufficient possibility that we can produce high-resolution solar magnetograms from historical Ca ii data.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.