We developed a reliable probabilistic solar flare forecasting model using a deep neural network, named Deep Flare Net-Reliable (DeFN-R). The model can predict the maximum classes of flares that occur in the following 24 h after observing images, along with the event occurrence probability. We detected active regions from 3x10^5 solar images taken during 2010-2015 by Solar Dynamic Observatory and extracted 79 features for each region, which we annotated with flare occurrence labels of X-, M-, and C-classes. The extracted features are the same as used by Nishizuka et al. (2018); for example, line-of-sight/vector magnetograms in the photosphere, brightening in the corona, and the X-ray emissivity 1 and 2 h before an image. We adopted a chronological split of the database into two for training and testing in an operational setting: the dataset in 2010-2014 for training and the one in 2015 for testing. DeFN-R is composed of multilayer perceptrons formed by batch normalizations and skip connections. By tuning optimization methods, DeFN-R was trained to optimize the Brier skill score (BSS). As a result, we achieved BSS = 0.41 for >=C-class flare predictions and 0.30 for >=M-class flare predictions by improving the reliability diagram while keeping the relative operating characteristic curve almost the same. Note that DeFN is optimized for deterministic prediction, which is determined with a normalized threshold of 50%. On the other hand, DeFN-R is optimized for a probability forecast based on the observation event rate, whose probability threshold can be selected according to users' purposes.

We developed a solar flare prediction model using a deep neural network (DNN), named Deep Flare Net (DeFN). The model can calculate the probability of flares occurring in the following 24 h in each active region, which is used to determine the most likely maximum classes of flares via a binary classification (e.g., >=M class versus <M class or >=C class versus <C class). From 3x10^5 observation images taken during 2010-2015 by Solar Dynamic Observatory, we automatically detected sunspots and calculated 79 features for each region, to which flare occurrence labels of X-, M-, and C-class were attached. We adopted the features used in Nishizuka et al. (2017) and added some features for operational prediction: coronal hot brightening at 131 A (T>=10^7 K) and the X-ray and 131 A intensity data 1 and 2 h before an image. For operational evaluation, we divided the database into two for training and testing: the dataset in 2010-2014 for training and the one in 2015 for testing. The DeFN model consists of deep multilayer neural networks, formed by adapting skip connections and batch normalizations. To statistically predict flares, the DeFN model was trained to optimize the skill score, i.e., the true skill statistic (TSS). As a result, we succeeded in predicting flares with TSS=0.80 for >=M-class flares and TSS=0.63 for >=C-class flares. Note that in usual DNN models, the prediction process is a black box. However, in the DeFN model, the features are manually selected, and it is possible to analyze which features are effective for prediction after evaluation.

We analyzed a small flux rope eruption converted into a helical blowout jet in a fan-spine configuration using multi-wavelength observations taken by SDO, which occurred near the limb on 2016 January 9. In our study, first, we estimated the fan-spine magnetic configuration with the potential field calculation and found a sinistral small filament inside it. The filament along with the flux rope erupted upward and interacted with the surrounding fan- spine magnetic configuration, where the flux rope breaks in the middle section. We observed compact brightening, flare ribbons and post-flare loops underneath the erupting filament. The northern section of the flux rope reconnected with the surrounding positive polarity, while the southern section straightened. Next, we observed the untwisting motion of the southern leg, which was transformed into a rotating helical blowout jet. The sign of the helicity of the mini-filament matches the one of the rotating jet. This is consistent with the jet models presented by Adams et al. (2014) and Sterling et al. (2015). We focused on the fine thread structure of the rotating jet and traced three blobs with the speed of 60-120 km/s, while the radial speed of the jet is approx 400 km/s. The untwisting motion of the jet accelerated plasma upward along the collimated outer spine field lines, and it finally evolved into a narrow coronal mass ejection at the height of approx 9 Rsun . On the basis of the detailed analysis, we discussed clear evidence of the scenario of the breaking of the flux rope and the formation of the helical blowout jet in the fan-spine magnetic configuration.

We developed a flare prediction model using machine learning, which is optimized to predict the maximum class of flares occurring in the following 24 h. Machine learning is used to devise algorithms that can learn from and make decisions on a huge amount of data. We used solar observation data during the period 2010-2015, such as vector magnetogram, ultraviolet (UV) emission, and soft X-ray emission taken by the Solar Dynamics Observatory and the Geostationary Operational Environmental Satellite. We detected active regions from the full-disk magnetogram, from which 60 features were extracted with their time differentials, including magnetic neutral lines, the current helicity, the UV brightening, and the flare history. After standardizing the feature database, we fully shuffled and randomly separated it into two for training and testing. To investigate which algorithm is best for flare prediction, we compared three machine learning algorithms: the support vector machine (SVM), k-nearest neighbors (k-NN), and extremely randomized trees (ERT). The prediction score, the true skill statistic (TSS), was higher than 0.9 with a fully shuffled dataset, which is higher than that for human forecasts. It was found that k-NN has the highest performance among the three algorithms. The ranking of the feature importance showed that the previous flare activity is most effective, followed by the length of magnetic neutral lines, the unsigned magnetic flux, the area of UV brightening, and the time differentials of features over 24 h, all of which are strongly correlated with the flux emergence dynamics in an active region.

Hinode observations have revealed intermittent recurrent plasma ejections/jets in the chromosphere. These are interpreted as a result of non-perfectly anti-parallel magnetic reconnection, i.e. component reconnection, between a twisted magnetic flux tube and the pre-existing coronal/chromospheric magnetic field, though the fundamental physics of component reconnection is unrevealed. In this paper, we experimentally reproduced the magnetic configuration and investigated the dynamics of plasma ejections, heating and wave generation triggered by component reconnection in the chromosphere. We set plasma parameters as in the chromosphere (density 10^14 cm^-3, temperature 5-10 eV, i.e. (5-10)x10^4 K, and reconnection magnetic field 200 G) using argon plasma. Our experiment shows bi-directional outflows with the speed of 5 km/s at maximum, ion heating in the downstream area over 30 eV and magnetic fluctuations mainly at 5-10 us period. We succeeded in qualitatively reproducing chromospheric jets, but quantitatively we still have some differences between observations and experiments such as jet velocity, total energy and wave frequency. Some of them can be explained by the scale gap between solar and laboratory plasma, while the others probably by the difference of microscopy and macroscopy, collisionality and the degree of ionization, which have not been achieved in our experiment.

We report observations of slowly drifting pulsating structures (DPS) in the 0.8-4.5 GHz frequency range of the RT4 and RT5 radio spectrographs at Ondrejov observatory, between 2002 and 2012. We found 106 events of drifting pulsating structures, which we classified into 4 cases: (I) single events with a constant frequency drift [12 events], (II) multiple events occurring in the same flare with constant frequency drifts [11 events], (III) single or multiple events with increasing or decreasing frequency drift rates [52 events], and (IV) complex events containing multiple events occurring at the same time in the different frequency range [31 events]. Many DPSs are associated with hard X-ray bursts (15-25 keV) and soft X-ray gradient peaks, as they typically occurred at the beginning of the hard X-ray peaks. This indicates that DPS events are related to the processes of fast energy release and particle acceleration. Furthermore, interpreting DPSs as signatures of plasmoids, we measured their ejection velocity, their width and their height from the DPS spectra, from which we also estimated the reconnection rate and the plasma beta. In this interpretation, constant frequency drift indicates a constant velocity of a plasmoid, and an increasing/decreasing frequency drift indicates a deceleration/acceleration of a plasmoid ejection. The reconnection rate shows a good positive correlation with the plasmoid velocity. Finally we confirmed that some DPS events show plasmoid counterparts in AIA/SDO images.

Shortly after the seminal paper {\sl "Self-Organized Criticality: An explanation of 1/f noise"} by Bak, Tang, and Wiesenfeld (1987), the idea has been applied to solar physics, in {\sl "Avalanches and the Distribution of Solar Flares"} by Lu and Hamilton (1991). In the following years, an inspiring cross-fertilization from complexity theory to solar and astrophysics took place, where the SOC concept was initially applied to solar flares, stellar flares, and magnetospheric substorms, and later extended to the radiation belt, the heliosphere, lunar craters, the asteroid belt, the Saturn ring, pulsar glitches, soft X-ray repeaters, blazars, black-hole objects, cosmic rays, and boson clouds. The application of SOC concepts has been performed by numerical cellular automaton simulations, by analytical calculations of statistical (powerlaw-like) distributions based on physical scaling laws, and by observational tests of theoretically predicted size distributions and waiting time distributions. Attempts have been undertaken to import physical models into the numerical SOC toy models, such as the discretization of magneto-hydrodynamics (MHD) processes. The novel applications stimulated also vigorous debates about the discrimination between SOC models, SOC-like, and non-SOC processes, such as phase transitions, turbulence, random-walk diffusion, percolation, branching processes, network theory, chaos theory, fractality, multi-scale, and other complexity phenomena. We review SOC studies from the last 25 years and highlight new trends, open questions, and future challenges, as discussed during two recent ISSI workshops on this theme.

A broad jet was observed in a weak magnetic field area at the edge of active region NOAA 11106. The peculiar shape and magnetic environment of the broad jet raised the question of whether it was created by the same physical processes of previously studied jets with reconnection occurring high in the corona. We carried out a multi-wavelength analysis using the EUV images from the Atmospheric Imaging Assembly (AIA) and magnetic fields from the Helioseismic and Magnetic Imager (HMI) both on-board the SDO satellite. The jet consisted of many different threads that expanded in around 10 minutes to about 100 Mm in length, with the bright features in later threads moving faster than in the early ones, reaching a maximum speed of about 200 km s^{-1}. Time-slice analysis revealed a striped pattern of dark and bright strands propagating along the jet, along with apparent damped oscillations across the jet. This is suggestive of a (un)twisting motion in the jet, possibly an Alfven wave. A topological analysis of an extrapolated field was performed. Bald patches in field lines, low-altitude flux ropes, diverging flow patterns, and a null point were identified at the basis of the jet. Unlike classical lambda or Eiffel-tower shaped jets that appear to be caused by reconnection in current sheets containing null points, reconnection in regions containing bald patches seems to be crucial in triggering the present jet. There is no observational evidence that the flux ropes detected in the topological analysis were actually being ejected themselves, as occurs in the violent phase of blowout jets; instead, the jet itself may have gained the twist of the flux rope(s) through reconnection. This event may represent a class of jets different from the classical quiescent or blowout jets, but to reach that conclusion, more observational and theoretical work is necessary.

We investigated the dynamic evolution of a 3-dimensional (3D) flux rope eruption and magnetic reconnection process in a solar flare, by simply extending 2-dimensional (2D) resistive magnetohydrodynamic simulation model of solar flares with low $\beta$ plasma to 3D model. We succeeded in reproducing a current sheet and bi-directional reconnection outflows just below the flux rope during the eruption in our 3D simulations. We calculated four cases of a strongly twisted flux rope and a weakly twisted flux rope in 2D and 3D simulations. The time evolution of a weakly twisted flux rope in 3D simulation shows similar behaviors to 2D simulation, while a strongly twisted flux rope in 3D simulation shows clearly different time evolution from 2D simulation except for the initial phase evolution. The ejection speeds of both strongly and weakly twisted flux ropes in 3D simulations are larger than 2D simulations, and the reconnection rates in 3D cases are also larger than 2D cases. This indicates a positive feedback between the ejection speed of a flux rope and the reconnection rate even in the 3D simulation, and we conclude that the plasmoid-induced reconnection model can be applied to 3D. We also found that small scale plasmoids are formed inside a current sheet and make it turbulent. These small scale plasmoid ejections has role in locally increasing reconnection rate intermittently as observed in solar flares, coupled with a global eruption of a flux rope.

We report a detailed examination of the fine structure inside flare ribbons and the temporal evolution of this fine structure during the X2.5 solar flare that occurred on 2004 November 10. We examine elementary bursts of the C IV (1550{\AA}) emission lines seen as local transient brightenings inside the flare ribbons in the ultraviolet (1600{\AA}) images taken with Transition Region and Coronal Explorer, and we call them C IV kernels. This flare was also observed in Ha with the Sartorius 18 cm Refractor telescope at Kwasan observatory, Kyoto University, and in hard X-rays (HXR) with Reuven Ramaty High Energy Solar Spectroscopic Imager. Many C IV kernels, whose sizes were comparable to or less than 2", were found to brighten successively during the evolution of the flare ribbon. The majority of them were well correlated with the Ha kernels in both space and time, while some of them were associated with the HXR emission. These kernels were thought to be caused by the precipitation of nonthermal particles at the footpoints of the reconnecting flare loops. The time profiles of the C IV kernels showed intermittent bursts, whose peak intensity, duration, and time interval were well described by power-law distribution functions. This result is interpreted as evidence for "self-organized criticality" in avalanching behavior in a single flare event, or for fractal current sheets in the impulsive reconnection region.

The Soft X-ray Telescope (SXT) on board Yohkoh revealed that the ejection of X-ray emitting plasmoid is sometimes observed in a solar flare. It was found that the ejected plasmoid is strongly accelerated during a peak in the hard X-ray emission of the flare. In this paper we present an examination of the GOES X 2.3 class flare that occurred at 14.51 UT on 2000 November 24. In the SXT images we found multiple plasmoid ejections with velocities in the range of 250-1500 km/s, which showed blob-like or loop-like structures. Furthermore, we also found that each plasmoid ejection is associated with an impulsive burst of hard X-ray emission. Although some correlation between plasmoid ejection and hard X-ray emission has been discussed previously, our observation shows similar behavior for multiple plasmoid ejection such that each plasmoid ejection occurs during the strong energy release of the solar flare. As a result of temperature-emission measure analysis of such plasmoids, it was revealed that the apparent velocities and kinetic energies of the plasmoid ejections show a correlation with the peak intensities in the hard X-ray emissions.

We propose the particle acceleration model coupled with multiple plasmoid ejections in a solar flare. Unsteady reconnection produces plasmoids in a current sheet and ejects them out to the fast shocks, where particles in a plasmoid are reflected upstream the shock front by magnetic mirror effect. As the plasmoid passes through the shock front, the reflection distance becomes shorter and shorter driving Fermi acceleration, until it becomes proton Larmor radius. The fractal distribution of plasmoids may also have a role in naturally explaining the power-law spectrum in nonthermal emissions.

The heating of the solar chromosphere and corona is a long-standing puzzle in solar physics. Hinode observations show the ubiquitous presence of chromospheric anemone jets outside sunspots in active regions. They are typically 3 to 7 arc seconds = 2000 to 5000 kilometers long and 0.2 to 0.4 arc second = 150 to 300 kilometers wide, and their velocity is 10 to 20 kilometers per second. These small jets have an inverted Y-shape, similar to the shape of x-ray anemone jets in the corona. These features imply that magnetic reconnection similar to that in the corona is occurring at a much smaller spatial scale throughout the chromosphere and suggest that the heating of the solar chromosphere and corona may be related to small-scale ubiquitous reconnection.

Hinode discovered a beautiful giant jet with both cool and hot components at the solar limb on 2007 February 9. Simultaneous observations by the Hinode SOT, XRT, and TRACE 195 satellites revealed that hot (5x10^6 K) and cool (10^4 K) jets were located side by side and that the hot jet preceded the associated cool jet (1-2 minutes). A current-sheet-like structure was seen in optical (Ca IIH), EUV (195A), and soft X-ray emissions, suggesting that magnetic reconnection is occurring in the transition region or upper chromosphere. Alfven waves were also observed with Hinode SOT. These propagated along the jet at velocities of 200 km/s with amplitudes (transverse velocity) of 5-15 km/s and a period of 200 s. We performed two-dimensional MHD simulation of the jets on the basis of the emerging flux-reconnection model, by extending Yokoyama and Shibata's model. We extended the model with a more realistic initial condition (10^6 K corona) and compared our model with multiwavelength observations. The improvement of the coronal temperature and density in the simulation model allowed for the first time the reproduction of the structure and evolution of both the cool and hot jets quantitatively, supporting the magnetic reconnection model. The generation and the propagation of Alfven waves are also reproduced self-consistently in the simulation model.