In this study, we develop an artificial intelligence (AI)-based solar surface flux transport (SFT) model. We predict synoptic maps for the next solar rotation (27.2753 days) using deep learning. Our model takes the latest synoptic maps and their sine-latitude grid data as inputs. Synoptic maps, which represent global magnetic field distributions on the solar surface, have been widely used as initial boundary conditions in the Sun and space-weather prediction models. Here we train and evaluate our deep-learning model, based on the Pix2PixCC architecture, using data sets of Solar Dynamics Observatory/Helioseismic and Magnetic Imager, Solar and Heliospheric Observatory/Michelson Doppler Imager, and National Solar Observatory/Global Oscillation Network Group synoptic maps with a resolution of 360 by 180 (longitude and sine latitude) from 1996 to 2023. We present results of our model and compare them with those from the persistent model and the conventional SFT model, including the effects of differential rotation, meridional flow, and diffusion on the solar surface. The average pixel-to-pixel correlation coefficient between the target and our AI-generated data, after 10 by 10 binning with a 10° resolution in longitude, is 0.71. This result is qualitatively similar to the results of the conventional SFT model (0.65–0.68) and better than the results of the persistent model (0.56). Our model successfully generates magnetic features, such as the diffusion of solar active regions and the motions of supergranules. Using synthetic input data with bipolar structures, we confirm that our model successfully reproduces differential rotation and meridional flow. Finally, we discuss the advantages and limitations of our model in view of magnetic field evolution and its potential applications. 

 Here we greatly improve artificial intelligence (AI)–generated solar farside magnetograms using data sets from the Solar Terrestrial Relations Observatory (STEREO) and Solar Dynamics Observatory (SDO). We modify our previous deep-learning model and configuration of input data sets to generate more realistic magnetograms than before. First, our model, which is called Pix2PixCC, uses updated objective functions, which include correlation coefficients (CCs) between the real and generated data. Second, we construct input data sets of our model: solar farside STEREO extreme-ultraviolet (EUV) observations together with nearest frontside SDO data pairs of EUV observations and magnetograms. We expect that the frontside data pairs provide historic information on magnetic field polarity distributions. We demonstrate that magnetic field distributions generated by our model are more consistent with the real ones than previously, in consideration of several metrics. The averaged pixel-to-pixel CC for full disk, active regions, and quiet regions between real and AI-generated magnetograms with 8 × 8 binning are 0.88, 0.91, and 0.70, respectively. Total unsigned magnetic flux and net magnetic flux of the AI-generated magnetograms are consistent with those of real ones for the test data sets. It is interesting to note that our farside magnetograms produce polar field strengths and magnetic field polarities consistent with those of nearby frontside magnetograms for solar cycles 24 and 25. Now we can monitor the temporal evolution of active regions using solar farside magnetograms by the model together with the frontside ones. Our AI-generated solar farside magnetograms are now publicly available at the Korean Data Center for SDO (http://sdo.kasi.re.kr). 

 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 study, we develop deep learning models to forecast the 6 hr interplanetary magnetic field (IMF) Bz component for southward cases. The models are based on a bidirectional long short-term memory method, and input parameters are solar wind data (V, N, T) and IMF components (Bt, Bx, By, Bz). The data are obtained from OMNI, whose period is from 2000 to 2022. We use the preceding 12 hr of data as input and the subsequent 6 hr of Bz data as target. To focus on strong geomagnetic conditions, we consider periods where Bz values drop below the negative standard deviation (approximately ‑3 nT) for at least 6 hr. The models are trained and validated using a 12-fold cross-validation process, with each model trained over 8 months of data and tested over 4 months. The ensemble model, which averages 12-fold model results, achieves an RMSE ranging from 1.75 (30 minutes prediction) to 2.55 nT (6 hr prediction), significantly outperforming two baseline methods: multilayer perceptron and multiple linear regression. Our model can capture both decreasing and increasing phases of Bz, showing reliable performance across varying geomagnetic conditions. Our results suggest a sufficient possibility for predicting Bz under noticeable southward conditions. We expect that our model can be used for subsequent space weather predictions, such as global magnetohydrodynamic simulations in the magnetosphere. 

In this study, we develop a physics-informed neural operator (PINO) model that learns the solution operator from 2D photospheric vector magnetic fields to 3D nonlinear force-free fields (NLFFFs). We train our PINO model using physics loss from NLFFF partial differential equations, as well as data loss from target NLFFFs. We validate our method using an analytical NLFFF model. We then train and evaluate our PINO model with 2327 numerical NLFFFs of 211 active regions from the Institute for Space-Earth Environmental Research database. The results show that our trained PINO model can generate an NLFFF within 1 s for any active region on a single consumer GPU, making real-time extrapolation of NLFFFs possible. Our artificial intelligence (AI)-generated NLFFFs are qualitatively and quantitatively quite similar to the target NLFFFs for 30 active regions. The magnetic energy of the AI-generated NLFFFs of active region 11158 follows a similar trend to the target NLFFFs as well as other conventional methods.

In this study, we address the question of whether we can properly determine differential emission measures (DEMs) using Solar Orbiter/Extreme Ultraviolet Imager (EUI)/Full Sun Imager (FSI) and AI-generated extreme UV (EUV) data. The FSI observes only two full-disk EUV channels (174 and 304 Å), which is insufficient for accurately determining DEMs and can lead to significant uncertainties. To solve this problem, we trained and tested deep learning models based on Pix2PixCC using the Solar Dynamics Observatory (SDO)/Atmospheric Imaging Assembly (AIA) dataset. The models successfully generated five-channel (94, 131, 193, 211, and 335 Å) EUV data from 171 and 304 Å EUV observations with high correlation coefficients. Then we applied the trained models to the Solar Orbiter/EUI/FSI dataset and generated the five-channel data that the FSI cannot observe. We used the regularized inversion method to compare the DEMs from the SDO/AIA dataset with those from the Solar Orbiter/EUI/FSI dataset, which includes AI-generated data. We demonstrate that, when SDO and Solar Orbiter are at the inferior conjunction, the main peaks and widths of both DEMs are consistent with each other at the same coronal structures. Our study suggests that deep learning can make it possible to properly determine DEMs using Solar Orbiter/EUI/FSI and AI-generated EUV data. 

In this study, we examine the relationships between the National Oceanic and Atmospheric Administration (NOAA) space weather scale frequencies and the maximum monthly sunspot number in each solar cycle: 1975 to 2020 for radio blackouts (R scales) and solar radiation storms (S scales), 1932 to 2020 for geomagnetic storms (G scales). Our main results are as follows. First, we find that NOAA space weather scale frequencies have strong solar cycle dependencies. Second, we propose new linear relationships between the frequency of certain scales (R1 to R4, and G1 to G4) and the maximum monthly sunspot number. T-test results show that R1 to R3 and G1 to G4 relationships are statistically meaningful, but marginal for R4. Third, our results significantly reduce the root-mean-square error (RMSE) between observed and suggested frequencies compared to the NOAA's current frequencies. For example, in the case of solar cycle 24, our new prediction (74) for R3 scale is much more consistent with the observational frequency (74) than the NOAA prediction (175), and our prediction (85) for G3 scale is much closer to the observation (40) than the NOAA prediction (200). Our work may provide a useful guideline for advancing the space weather scales. 

The Fe/O enhancements exhibit significant variations in gradual solar energetic particle (SEP) events. Several causes have been suggested including transport effects in the interplanetary space and flare contribution. In this study, we investigate the relationship between the integrated Fe/O ratios of 27 gradual SEP events, locations of associated solar flares, and positions along the heliospheric current sheet (HCS) between 2010 and 2014. We employ synchronic potential field source surface (PFSS) extrapolations at 2.5R⊙, derived in near real-time using Artificial Intelligence (AI)-generated far side and Helioseismic and Magnetic Imager (HMI) magnetograms, referred to as AIHMI-PFSS extrapolations. We examine low-energy (∼0.5 MeV/nucleon) Fe and O ion measurements obtained from Suprathermal Ion Telescope on Solar Terrestrial Relations Observatories and Ultra Low Energy Isotope Spectrometer on Advanced Composition Explorer. We found a moderate anticorrelation between the Fe/O ratios and the absolute longitudinal separation angles from the source regions to the spacecraft magnetic footpoints. Furthermore, we investigate the variations in Fe/O ratios with respect to the separation angle, grouped by the same and opposite polarity sectors of the SEP source regions. We found that the mean and median Fe/O values are higher in the same polarity group compared to the opposite polarity group, with the largest contrast at separation angles between 25° and 50°, where the values are approximately 3 times larger. The results imply that the enhanced Fe/O ratios in the examined gradual SEP events are likely associated with direct source regions, while the HCS affects particle transport. 

We address the question of whether the solar limb flare prediction can be properly made by EUV intensity, which has less projection effects than solar white light and magnetogram data. We develop empirical and multilayer perceptron (MLP) models to forecast the probability of a major solar limb flare within a day. We use Solar Dynamics Observatory (SDO)/Atmospheric Imaging Assembly (AIA) 94 and 131 Å that have high correlations and large slopes with X-ray flare fluxes from 2010 to 2022. We select 240 flares stronger than or equal to the M1.0 class and located near the limb region (60° or more in heliographic longitude). For input data, we use the limb intensity as the sum of SDO/AIA intensities in the limb region and the total intensity of the whole image. We compare the model performances using metrics such as the receiver operating characteristic—area under the curve. Our major results are as follows. First, we can forecast major solar limb flare occurrences with only SDO/AIA 94 and/or 131 Å intensities. Second, our models show better probability prediction than the climatological model. Third, both empirical (AUC = 0.85) and MLP (AUC = 0.84) models have similar performances, which are much better than a random forecast (AUC = 0.50). Finally, it is interesting to note that our model can forecast the flaring probability of all 52 events during the test period, while the models in the NASA/CCMC flare scoreboard can forecast only 22 events. From the above results, we can answer that the solar limb flare prediction using EUV intensity can be properly made. 

In this Letter, we examine the relationship between 3D magnetic free energy (MFE) and flaring activity using 83 major solar flares (M-class and X-class) in nine solar active regions (ARs). For this, we use 998 nonlinear force-free field extrapolations compiled by the "Institute for Space-Earth Environmental Research Database" at Nagoya University. These ARs produced at least three major flares with distinct rising and peak phases of 3D MFE. For each phase, the solar flare occurrence rate (FOR) is calculated as a ratio of the number of flares to the duration. The major results from this study are summarized as follows. First, there is no clear linear correlation (CC = 0.15) between 3D MFE and flare peak flux. Second, the FOR (3.4 day−1) of the rising phase is a little higher than that (3.1 day−1) of the peak phase, depending on AR. Third, for several flares, there are noticeable decreases in 3D MFE, which correspond to the typical energy of a major flare (about 1032 erg). Fourth, it is interesting to note that there are noticeable enhancements in FORs at several local maxima of 3D MFE, which may be associated with flux emergence and/or shearing motions. Fifth, the flare index rates, which are defined as the summation of flaring activity divided by the duration, of rising and peak phases are 151 day−1 and 314 day−1, respectively. Our results imply that the traditional and simple "storage and release" model does not apply to flare activities, and the random perturbation may be important for triggering flares. 

The Sun-Earth Lagrange point L4, which is called a parking space of space, is considered one of the unique places where solar activity and the heliospheric environment can be observed continuously and comprehensively. The L4 mission affords a clear and wide-angle view of the Sun-Earth line for the study of Sun-Earth connections from remote-sensing observations. The L4 mission will significantly contribute to advancing heliophysics science, improving space weather forecasting capability, extending space weather studies far beyond near-Earth space, and reducing risk from solar radiation hazards on human missions to the Moon and Mars. Our paper outlines the importance of L4 observations by using remote-sensing instruments and advocates comprehensive and coordinated observations of the heliosphere at multi-points including other planned L1 and L5 missions. We mainly discuss scientific perspectives on three topics in view of remote sensing observations: (1) solar magnetic field structure and evolution, (2) source regions of geoeffective solar energetic particles (SEPs), and (3) stereoscopic views of solar corona and coronal mass ejections (CMEs). 

For the first time, we generate solar coronal parameters (density, magnetic field, radial velocity, and temperature) on a near-real-time basis by deep learning. For this, we apply the Pix2PixCC deep-learning model to threedimensional (3D) distributions of these parameters: synoptic maps of the photospheric magnetic field as an input and the magnetohydrodynamic algorithm outside a sphere (MAS) results as an output. To generate the 3D structure of the solar coronal parameters from 1 to 30 solar radii, we train and evaluate 152 distinct deep-learning models. For each parameter, we consider the data of 169 Carrington rotations from 2010 June to 2023 February: 132 for training and 37 for testing. The key findings of our study are as follows: First, our deep-learning models successfully reconstruct the 3D distributions of coronal parameters from 1 to 30 solar radii with an average correlation coefficient of 0.98. Second, during the solar active and quiet periods, the AI-generated data exhibits consistency with the target MAS simulation data. Third, our deep-learning models for each parameter took a remarkably short time (about 16 s for each parameter) to generate the results with an NVIDIA Titan XP GPU. As the MAS simulation is a regularization model, we may significantly reduce the simulation time by using our results as an initial configuration to obtain an equilibrium condition. We hope that the generated 3D solar coronal parameters can be used for the near-real-time forecasting of heliospheric propagation of solar eruptions.

We study the source regions of six solar energetic particle (SEP) events accelerated near or behind the limbs of the Sun. We use AI-generated farside magnetograms at a near real-time basis developed by Jeong et al. and AIHMI-PFSS extrapolations up to 2.5Rsun computed using the input of the synchronic data combining AI-generated farside and HMI magnetograms. By comparing the AIHMI, HMI, Global Oscillations Network Group (GONG) synoptic magnetograms, and Air force Data Assimilative Photospheric flux Transport synchronic magnetograms, as well as the PFSS extrapolations, we find interesting differences between them in view of SEP source regions and magnetic field configurations. First, the structures and sizes of the source active regions (ARs) are changed. The total unsigned magnetic field fluxes of the ARs are mostly stronger in the AIHMI than in the HMI and GONG magnetograms. Second, newly emerging ARs are observed in the SEP source regions in the AIHMI magnetograms for two events. Third, the alterations in the magnetic flux, the emergence, and the dissipation of ARs lead to modifications in the locations of the global polarity inversion lines (PILs). The EUV wave propagation is typically observed to be oriented nearly perpendicular with respect to the local PIL, suggesting that the AIHMI-PFSS extrapolations around the source region are more realistic. This study shows that the continuous farside evolution of AR magnetic fields, which is accomplished by our AI synchronic magnetograms, can lead to an improved understanding of SEP source ARs.

In this study, we forecast solar wind speed for the next 3 days with a 6 hr cadence using a deep-learning model. For this we use Solar Dynamics Observatory/Atmospheric Imaging Assembly 211 and 193 Å images together with solar wind speeds for the last 5 days as input data. The total period of the data is from 2010 May to 2020 December. We divide them into a training set (January–August), validation set (September), and test set (October–December), to consider the solar cycle effect. The deep-learning model consists of two networks: a convolutional layer–based network for images and a dense layer–based network for solar wind speeds. Our main results are as follows. First, our model successfully predicts the solar wind speed for the next 3 days. The rms error (RMSE) of our model is from 37.4 km/s (for the 6 hr prediction) to 68.2 km/s (for the 72 hr prediction), and the correlation coefficient is from 0.92 to 0.67. These results are much better than those of previous studies. Second, the model can predict sudden increase of solar wind speeds caused by large equatorial coronal holes. Third, solar wind speeds predicted by our model are more consistent with observations than those by the Wang–Sheely–Arge–ENLIL model, especially in high-speed-stream regions. It is also noted that our model cannot predict solar wind speed enhancement by coronal mass ejections. Our study demonstrates the effectiveness of deep learning for solar wind speed prediction, with potential applications in space weather forecasting.

Accurate and reliable disaster forecasting is vital for saving lives and property. Hence, effective disaster management is necessary to reduce the impact of natural disasters and to accelerate recovery and reconstruction. Typhoons are one of the major disasters related to heavy rainfall in Korea. As a typhoon develops in the far ocean, satellite observations are the only means to monitor them. Our study uses satellite observations to propose a deep-learning-based disaster monitoring model for short-term typhoon rainfall forecasting. For this, we consider two deep learning models: a video frame prediction model, Warp and Refine Network (WR-Net), to predict future satellite observations and an image-to-image translation model, geostationary rainfall product (GeorAIn) (based on the Pix2PixCC model), to generate rainfall maps from predicted satellite images. Typhoon Hinnamnor, the worst typhoon case in 2022 in Korea, is selected as a target case for model verification. The results show that the predicted satellite images can capture the structures and patterns of the typhoon. The rainfall maps generated from the GeorAIn model using predicted satellite images show a correlation coefficient of 0.81 for 3-hr and 0.56 for 7-hr predictions. The proposed disaster monitoring model can provide us with practical implications for disaster alerting systems and can be extended to flood-monitoring systems.

This study is the first attempt to generate a three-dimensional (3D) coronal electron density distribution based on the pix2pixHD model, whose computing time is much shorter than that of the magnetohydrodynamic (MHD) simulation. For this, we consider photospheric solar magnetic fields as input, and electron density distribution simulated with the MHD Algorithm outside a Sphere (MAS) at a given solar radius is taken as output. We consider 155 pairs of Carrington rotations as inputs and outputs from 2010 June to 2022 April for training and testing. We train 152 deep-learning models for 152 solar radii, which are taken up to 30 solar radii. The artificial intelligence (AI) generated 3D electron densities from this study are quite consistent with the simulated ones from lower radii to higher radii, with an average correlation coefficient 0.97. The computing time of testing data sets up to 30 solar radii of 152 deep-learning models is about 45.2 s using the NVIDIA TITAN XP graphics-processing unit, which is much less than the typical simulation time of MAS. We find that the synthetic coronagraphic images estimated from the deep-learning models are similar to the Solar Heliospheric Observatory (SOHO)/Large Angle and Spectroscopic Coronagraph C3 coronagraph data, especially during the solar minimum period. The AI-generated coronal density distribution from this study can be used for space weather models on a near-real-time basis.

 In this study, we present the application of deep reinforcement learning to the forecasting of major solar flares. For this, we consider full-disk magnetograms at 00:00 UT from the Solar and Heliospheric Observatory/Michelson Doppler Imager (1996–2010) and the Solar Dynamics Observatory/Helioseismic and Magnetic Imager (2011–2019), as well as Geostationary Operational Environmental Satellite X-ray flare data. We apply Deep Q-Network (DQN) and Double DQN, which are popular deep reinforcement learning methods, to predict "Yes or No" for daily M- and X-class flare occurrence. The reward functions, consisting of four rewards for true positive, false positive, false negative, and true negative, are used for our models. The major results of this study are as follows. First, our deep-learning models successfully predict major solar flares with good skill scores, such as HSS, F1, TSS, and ApSS. Second, the performance of our models depends on the reward function, learning method, and target agent update time. Third, the performance of our deep-learning models is noticeably better than that of a convolutional neural network (CNN) model with the same structure: 0.38 (CNN) to 0.44 (ours) for HSS, 0.47 to 0.52 for F1, 0.53 to 0.59 for TSS, and 0.09 to 0.12 for ApSS. 

 In this study, we suggest a pixel-to-pixel image translation method among similar types of filtergrams such as solar extreme-ultraviolet (EUV) images. For this, we consider a deep-learning model based on a fully connected network in which all pixels of solar EUV images are independent of one another. We use six-EUV-channel data from the Atmospheric Imaging Assembly (AIA) on board the Solar Dynamics Observatory (SDO), of which three channels (17.1, 19.3, and 21.1 nm) are used as the input data and the remaining three channels (9.4, 13.1, and 33.5 nm) as the target data. We apply our model to representative solar structures (coronal loops inside of the solar disk and above the limb, coronal bright point, and coronal hole) in SDO/AIA data and then determine differential emission measures (DEMs). Our results from this study are as follows. First, our model generates three EUV channels (9.4, 13.1, and 33.5 nm) with average correlation coefficient values of 0.78, 0.89, and 0.85, respectively. Second, our model generates the solar EUV data with no boundary effects and clearer identification of small structures when compared to a convolutional neural network-based deep-learning model. Third, the estimated DEMs from AI-generated data by our model are consistent with those using only SDO/AIA channel data. Fourth, for a region in the coronal hole, the estimated DEMs from AI-generated data by our model are more consistent with those from the 50 frames stacked SDO/AIA data than those from the single-frame SDO/AIA data. 

 In this study, we generate He i 1083 nm images from Solar Dynamic Observatory (SDO)/Atmospheric Imaging Assembly (AIA) images using a novel deep learning method (pix2pixHD) based on conditional Generative Adversarial Networks (cGAN). He i 1083 nm images from National Solar Observatory (NSO)/Synoptic Optical Long-term Investigations of the Sun (SOLIS) are used as target data. We make three models: single-input SDO/AIA 19.3 nm image for Model I, single-input 30.4 nm image for Model II, and double-input (19.3 and 30.4 nm) images for Model III. We use data from 2010 October to 2015 July except for June and December for training and the remaining one for test. Major results of our study are as follows. First, the models successfully generate He i 1083 nm images with high correlations. Second, Model III shows better results than those with one input image in terms of metrics such as correlation coefficient (CC) and root mean square error (RMSE). CC and RMSE between real and synthetic ones for model III with 4 by 4 binnings are 0.88 and 9.49, respectively. Third, synthetic images show well observational features such as active regions, filaments, and coronal holes. This work is meaningful in that our model can produce He i 1083 nm images with higher cadence without data gaps, which would be useful for studying the time evolution of the chromosphere and transition region. 

 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.