X-class Flares Research Articles

Forecasting Solar Flares by Data Assimilation in Sandpile Models

The prediction of solar flares is still a significant challenge in space weather research, with no techniques currently capable of producing reliable forecasts performing significantly above climatology. In this paper, we present a flare forecasting technique using data assimilation coupled with computationally inexpensive cellular automata called sandpile models. Our data assimilation algorithm uses the simulated annealing method to find an optimal initial condition that reproduces well an energy-release time series. We present and empirically analyze the predictive capabilities of three sandpile models, namely the Lu and Hamilton model (LH) and two deterministically-driven models (D). Despite their stochastic elements, we show that deterministically-driven models display temporal correlations between simulated events, a needed condition for data assimilation. We present our new data assimilation algorithm and demonstrate its success in assimilating synthetic observations produced by the avalanche models themselves. We then apply our method to GOES X-Ray time series for 11 active regions having generated multiple X-class flares in the course of their lifetime. We demonstrate that for such large flares, our data assimilation scheme substantially increases the success of ``All-Clear'' forecasts, as compared to model climatology.

Call and Response: A Time-resolved Study of Chromospheric Evaporation in a Large Solar Flare

We studied an X1.6 solar flare produced by NOAA Active Region 12602 on 2014 October 22. The entirety of this event was covered by RHESSI, IRIS, and Hinode/EIS, allowing analysis of the chromospheric response to a nonthermal electron driver. We derived the energy contained in nonthermal electrons via RHESSI spectral fitting and linked the time-dependent parameters of this call to the response in Doppler velocity, density, and nonthermal width across a broad temperature range. The total energy injected was 4.8 × 1030 erg and lasted 352 s. This energy drove explosive chromospheric evaporation, with a delineation in both Doppler and nonthermal velocities at the flow reversal temperature, between 1.35 and 1.82 MK. The time of peak electron injection (14:06 UT) corresponded to the time of highest velocities. At this time, we found 200 km s−1 blueshifts in the core of Fe xxiv, which is typically assumed to be at rest. Shortly before this time, the nonthermal electron population had the shallowest spectral index (≈6), corresponding to the peak nonthermal velocity in Si iv and Fe xxi. Nonthermal velocities in Fe xiv, formed near the flow reversal temperature, were low and not correlated with density or Doppler velocity. Nonthermal velocities in ions with similar temperatures were observed to increase and correlate with Doppler velocities, implying unresolved flows surrounding the flow reversal point. This study provides a comprehensive, time-resolved set of chromospheric diagnostics for a large X-class flare, along with a time-resolved energy injection profile, ideal for further modeling studies.

Sunspot Rotation in High- and Low-Flaring Active Regions

Sunspot rotations are closely linked with flaring activity. They are thought to contribute to the accumulation of helicity in magnetic flux tubes and to triggering magnetic reconnection in large solar flares. This link to solar flares has led to sunspot rotations being used as a parameter in solar flare prediction methods, but analysis for long-period observations of rotations in the literature is scarce. In this study, the rotation profiles of sunspots in a selection of six active regions are studied over time periods of 5 – 10 days to measure how sunspot rotation varies as active regions develop. The active regions are divided into two categories: high-flaring groups, which produced at least one X-class flare, and low-flaring regions that had little flaring activity. Comparison of the rotation profiles in these regions showed that young complex sunspot groups exhibit faster angular velocities and more frequent changes in rotation than older single-spot groups and, although the most rotating groups were also the most flare-productive, sudden changes in rotation were found to not definitively indicate an imminent eruption.

In Situ Observation of Solar-flare-induced Proton Cyclotron Waves Upstream from Mars

Proton cyclotron waves (PCWs) upstream from Mars are usually interpreted as waves generated by ion/ion instabilities due to the interaction between the solar wind plasma and the pickup protons, originating from the extended hydrogen (H) exosphere of Mars. Their generation mainly depends on the solar wind properties and the relative density of the newborn protons with respect to the background solar wind. Under stable solar wind conditions, a higher solar irradiance leads to both increased exospheric H density and ionization rate of H atoms, and therefore a higher relative density, which tends to increase the linear wave growth rate. Here we show that the solar irradiance is likely to contribute significantly to PCW generation. Specifically, we present observations from the Mars Atmosphere and Volatile EvolutioN (MAVEN) spacecraft indicating that, around the peak of the X8.2 flare on 2017 September 10, the increased solar irradiance gave rise to higher pickup H+ fluxes, which in turn excited PCWs. This result has implications for inferring the loss of hydrogen to space in early Martian history with more intense and frequent X-class flares as well as their contributions to the total loss.

Predicting Solar Flares Using CNN and LSTM on Two Solar Cycles of Active Region Data

We consider the flare prediction problem that distinguishes flare-imminent active regions that produce an M- or X-class flare in the succeeding 24 hr, from quiet active regions that do not produce any flares within ±24 hr. Using line-of-sight magnetograms and parameters of active regions in two data products covering Solar Cycles 23 and 24, we train and evaluate two deep learning algorithms—a convolutional neural network (CNN) and a long short-term memory (LSTM)—and their stacking ensembles. The decisions of CNN are explained using visual attribution methods. We have the following three main findings. (1) LSTM trained on data from two solar cycles achieves significantly higher true skill scores (TSSs) than that trained on data from a single solar cycle with a confidence level of at least 0.95. (2) On data from Solar Cycle 23, a stacking ensemble that combines predictions from LSTM and CNN using the TSS criterion achieves a significantly higher TSS than the “select-best” strategy with a confidence level of at least 0.95. (3) A visual attribution method called “integrated gradients” is able to attribute the CNN’s predictions of flares to the emerging magnetic flux in the active region. It also reveals a limitation of CNNs as flare prediction methods using line-of-sight magnetograms: it treats the polarity artifact of line-of-sight magnetograms as positive evidence of flares.

Sunspot numbers and proton events in solar cycles 19 to 24

In this study we compare the mean annual sunspot numbers for the six most recent solar cycles (19 to 24) with the number of solar energetic proton events occurring in those years. Though most of the individual cycles exhibit quite strong correlations, when they are combined into a generic solar cycle the result is very significant (ρ = 0.98). For cycles 21 to 24 (when spacecraft observations have been regularly available), we also investigate the variation in the source location of the solar energetic proton events in relation to four peak flux thresholds in orders of magnitude from ≥10 to ≥10,000 pfu. For helio-latitudes within ±40° (the range within which active regions usually occur) there is negligible variation in the helio-longitude of the source regardless of the peak flux threshold. However, the effect on the helio-longitude of varying the peak flux threshold is very significant: the higher the threshold the closer is the median source longitude to 0°, the variation being almost uniform from ≥10 pfu (32°W) to ≥10,000 pfu (2°W). Finally, of the 10 events in cycles 19 to 24 with peak flux ≥10,000 pfu, all occurred between years 4 and 8 after solar minimum, all were the result of M- or X-class flares with associated halo CMEs and shock fronts, and the three most intense events (≥40,000 pfu) were all related to X-class flares which occurred in the longitude range 10–28°E. These results support and extend previous work by the cited authors, and have implications for solar-terrestrial relations and the effects of space weather within geospace.

Predicting Solar Energetic Particles Using SDO/HMI Vector Magnetic Data Products and a Bidirectional LSTM Network

Solar energetic particles (SEPs) are an essential source of space radiation, and are hazardous for humans in space, spacecraft, and technology in general. In this paper, we propose a deep-learning method, specifically a bidirectional long short-term memory (biLSTM) network, to predict if an active region (AR) would produce an SEP event given that (i) the AR will produce an M- or X-class flare and a coronal mass ejection (CME) associated with the flare, or (ii) the AR will produce an M- or X-class flare regardless of whether or not the flare is associated with a CME. The data samples used in this study are collected from the Geostationary Operational Environmental Satellite's X-ray flare catalogs provided by the National Centers for Environmental Information. We select M- and X-class flares with identified ARs in the catalogs for the period between 2010 and 2021, and find the associations of flares, CMEs, and SEPs in the Space Weather Database of Notifications, Knowledge, Information during the same period. Each data sample contains physical parameters collected from the Helioseismic and Magnetic Imager on board the Solar Dynamics Observatory. Experimental results based on different performance metrics demonstrate that the proposed biLSTM network is better than related machine-learning algorithms for the two SEP prediction tasks studied here. We also discuss extensions of our approach for probabilistic forecasting and calibration with empirical evaluation.

Relationship Between the Evolution of Solar Magnetic Field and Flares in Different Active Regions

Two comparative active regions, NOAA AR 11283 and AR 11654 are studied. Both active regions have big sunspot numbers, and large areas. However, AR 11283 erupted two X-class flares, AR 11654 erupted little flares. Through investigating the evolution of magnetic field, the active region with complex magnetic configuration has more possibility to create big flares. NOAA is National Oceanic and Atmospheric Administration

Statistical Analysis of Circular-ribbon Flares

Circular-ribbon flares (CFs) are a special type of solar flares owing to their particular magnetic topology. In this paper, we conducted a comprehensive statistical analysis of 134 CFs from 2011 September to 2017 June, including 4 B-class, 82 C-class, 40 M-class, and 8 X-class flares. The flares were observed by the Atmospheric Imaging Assembly on board the Solar Dynamics Observatory spacecraft. The physical properties of CFs are derived, including the location, area (A CF), equivalent radius (r CF) assuming a semispherical fan dome, lifetime (τ CF), and peak soft X-ray (SXR) flux in 1–8 Å. It is found that all CFs are located in active regions, with the latitudes between −30° and 30°. The distributions of areas and lifetimes could be fitted with a lognormal function. There is a positive correlation between the lifetime and area. The peak SXR flux in 1–8 Å is well in accord with a power-law distribution with an index of −1.42. For the 134 CFs, 57% of them are accompanied by remote brightenings or ribbons. A positive correlation exists between the total length (L RB) and average distance (D RB) of remote brightenings. About 47% and 51% of the 134 CFs are related to type III radio bursts and jets, respectively. The association rates are independent of flare energies. About 38% of CFs are related to minifilament eruptions, and the association rates increase with flare classes. Only 28% of CFs are related to coronal mass ejections (CMEs), meaning that a majority of them are confined rather than eruptive events. There is a positive correlation between the CME speed and peak SXR flux in 1–8 Å, and faster CMEs tend to be wider.

Three-dimensional Propagation of the Global Extreme-ultraviolet Wave Associated with a Solar Eruption on 2021 October 28

Abstract We present a case study for the global extreme-ultraviolet (EUV) wave and its chromospheric counterpart the Moreton-Ramsey Wave associated with the second X-class flare in Solar Cycle 25 and a halo coronal mass ejection (CME). The EUV wave was observed in the Hα and EUV passbands with different characteristic temperatures. In the 171 Å and 193/195 Å images, the wave propagates circularly with an initial velocity of 600–720 km s−1 and a deceleration of 110–320 m s−2. The local coronal plasma is heated from log(T/K) ≈ 5.9 to log(T/K) ≈ 6.2 during the passage of the wave front. The Hα and 304 Å images also reveal signatures of wave propagation with a velocity of 310–540 km s−1. With multiwavelength and dual-perspective observations, we found that the wave front likely propagates forwardly inclined to the solar surface with a tilt angle of ∼53°.2. Our results suggest that this EUV wave is a fast-mode magnetohydrodynamic wave or shock driven by the expansion of the associated CME, whose wave front is likely a dome-shaped structure that could impact the upper chromosphere, transition region, and corona.

Active region and flare ribbon properties associated with X-class flares and CMEs of solar cycle 24

Active region and flare ribbon properties associated with X-class flares and CMEs of solar cycle 24

Magnetic Winding as an Indicator of Flare Activity in Solar Active Regions

Abstract Magnetic helicity is a measure of the entanglement of magnetic field lines used to characterize the complexity of solar active region (AR) magnetic fields. Previous attempts to use helicity-based indicators to predict solar eruptive/flaring events have shown promise but not been universally successful. Here we investigate the use of a quantity associated with the magnetic helicity, the magnetic winding, as a means to predict flaring activity. This quantity represents the fundamental entanglement of magnetic field lines and is independent of the magnetic field strength. We use vector magnetogram data derived from the Helioseismic Magnetic Imager (HMI) to calculate the evolution and distribution of the magnetic winding flux associated with five different ARs, three of them with little flaring activity/nonflaring (AR 11318, AR 12119, AR 12285) and two highly active with X-class flares (AR 11158, AR 12673). We decompose these quantities into “current-carrying” and “potential” parts. It is shown that the ARs that show flaring/eruptive activity have significant contributions to the winding input from the current-carrying part of the field. A significant and rapid input of current-carrying winding is found to be a precursor of flaring/eruptive activity, and, in conjunction with the helicity, sharp inputs of both quantities are found to precede individual flaring events by several hours. This suggests that the emergence/submergence of topologically complex current-carrying field is an important element for the ignition of AR flaring.

The development of PIL extraction application in active region of SDO/HMI magnetogram

The Polarity Inversion Line (PIL) on the active region (AR) is a parameter that correlates with the occurrence of the solar flare. This research has developed application to extract the PIL length in the AR. Image processing techniques were used to detect and classify AR from the SDO/HMI Magnetogram image of 1024x1024 pixels. The detection of AR in magnetogram images has its challenges because the ARs are constructed by the white (black) regions represent the magnetic positive (negative) polarities. To improve detection accuracy, the application used The Solar Region Summary (SRS), compiled by SWPC, and the centroid position of sunspot on SDO/HMI Continuum as additional information. The results of AR detection had high accuracy because SRS stores the centroid position of the flaring active region at peak time with the Stonyhurst heliographic on the solar disk which is converted to Cartesian coordinates. The application test used 1450 images and obtained the PIL length on the peak time of the X-class and M-class flares from 2011-2017. The application detected the high distribution of PIL length in the range of 120 to 580 Mm for X class flare, and 120 to 390 Mm for M-class flare. This research provides new data that has the potential to improve the machine learning-based solar flare prediction models.

On the Differences in the Periodic Behavior of Magnetic Helicity Flux in Flaring Active Regions with and without X-class Events

Abstract Observational precursors of large solar flares provide a basis for future operational systems for forecasting. Here, we study the evolution of the normalized emergence (EM), shearing (SH), and total (T) magnetic helicity flux components for 14 flaring (with at least one X-class flare) and 14 nonflaring (<M5-class flares) active regions (ARs) using the Space-weather Helioseismic Magnetic Imager Active Region Patches vector magnetic field data. Each of the selected ARs contain a δ-type spot. The three helicity components of these ARs were analyzed using wavelet analysis. Localized peaks of the wavelet power spectrum (WPS) were identified and statistically investigated. We find that (i) the probability density function of the identified WPS peaks for all the EM/SH/T profiles can be fitted with a set of Gaussian functions centered at distinct periods between ∼3 and 20 hr. (ii) There is a noticeable difference in the distribution of periods found in the EM profiles between the flaring and nonflaring ARs, while no significant difference is found in the SH and T profiles. (iii) In flaring ARs, the distributions of the shorter EM/SH/T periods (<10 hr) split up into two groups after flares, while the longer periods (>10 hr) do not change. (iv) When the EM periodicity does not contain harmonics, the ARs do not host a large energetic flare. (v) Finally, significant power at long periods (∼20 hr) in the T and EM components may serve as a precursor for large energetic flares.

A statistical analysis of X- class solar flares and their association with the super active regions during solar cycles 23- 24

The A statistical distribution function of X-class flares and their association with super-active regions (SARs) during two recent solar cycles 23-24 have been analyzed. The obtained results show that, during solar cycle 23, 57.60 % of the total number of Xclass flares coincides with a set of X1.0 - X1.9 flares, 15.20 % to X2.0 - X2.9, 12.80 % to X3.0 - X4.9, 9.60 % to X5.0 - X9.9, and 4.80 % to ≥ X10 flares. During solar cycle 24, 61.20 % of the total number of X-class flares coincides with a set of X1.0 - X1.9 flares, 22.40 % to X2.0 - X2.9, 8.16 % to both the X3.0 - X4.9 flares, and X5.0 - X9.9 flares. None of the ≥ X10 flares. All X-class flares are found to be in a scattered manner in solar cycles 23 and 24. Among the cycles, solar cycle 24 has been observed relatively very weak cycle and the total number of X-class flares diminished in comparison to previous solar cycles. During solar cycle 23, 29.16% of the X1.0 - X1.9 flares, 36.84 % of the X2.0 - X2.9 flares, 43.75 % of the X3.0 - X4.9 flares, 83.33 % of both the X5.0 - X9.9 flares, and 100.00 % of the ≥ X10 flares were produced by SARs. While during solar cycle 24, 20.00 % of the X1.0 - X1.9 flares, 36.36 % of the X2.0 - X2.9 flares, 25.00 % of the X3.0 - X4.9 flares, 75.00 % of the X5.0 - X9.9 flares, and 0.00 % of the ≥ X10 flares were produced by SARs.

Low-dimensional Convolutional Neural Network for Solar Flares GOES Time-series Classification

Abstract Space weather phenomena such as solar flares have a massive destructive power when they reach a certain magnitude. Here, we explore the deep-learning approach in order to build a solar flare-forecasting model, while examining its limitations and feature-extraction ability based on the available Geostationary Operational Environmental Satellite (GOES) X-ray time-series data. We present a multilayer 1D convolutional neural network to forecast the solar flare event probability occurrence of M- and X-class flares at 1, 3, 6, 12, 24, 48, 72, and 96 hr time frames. The forecasting models were trained and evaluated in two different scenarios: (1) random selection and (2) chronological selection, which were compared afterward in terms of common score metrics. Additionally, we also compared our results to state-of-the-art flare-forecasting models. The results indicates that (1) when X-ray time-series data are used alone, the suggested model achieves higher score results for X-class flares and similar scores for M-class as in previous studies. (2) The two different scenarios obtain opposite results for the X- and M-class flares. (3) The suggested model combined with solely X-ray time-series fails to distinguish between M- and X-class magnitude solar flare events. Furthermore, based on the suggested method, the achieved scores, obtained solely from X-ray time-series measurements, indicate that substantial information regarding the solar activity and physical processes are encapsulated in the data, and augmenting additional data sets, both spatial and temporal, may lead to better predictions, while gaining a comprehensive physical interpretation regarding solar activity. All source codes are available at https://github.com/vladlanda.

The Precursor Phase of an X-class Flare: Magnetic Reconnection, Powering and Non-thermal Electrons

In this paper, we report three interesting phenomena that occurred during the precursor phase of the X1.6 class flare on 2014 September 10. (1) The magnetic reconnection initiating the flare occurs between one of the two J-shaped magnetic flux ropes that constitute a sigmoidal structure and the overlying sheared magnetic arcade that runs across the sigmoid over its middle part. The reconnection formed an erupting structure that ultimately leads to flare onset. Another J-shaped magnetic flux rope remains unaffected during the whole eruption. The phenomenon is revealed by the observation made by the Atmospheric Imaging Assembly on board the Solar Dynamic Observatory (SDO) at 94 and 131 Å. (2) Being simultaneously with starting time of the precursor, photospheric vertical electric current (VEC) around the footpoint region of the overlying magnetic arcade underwent an obvious increase, as observed by the Helioseismic and Magnetic Imager (HMI) on board SDO. By only taking into account the VEC with current density over 3σ value (1σ: 10 mA m−2), we are able to pick out precursor-associated VEC increase starting from nearly the level of zero. We regard it as a kind of powering process for the magnetic reconnection between the two magnetic loops. (3) With high-resolution narrow-band Helium 10830 Å images taken by Goode Solar Telescope at Big Bear Solar Observatory (BBSO), we observe a narrow absorption (dark) front that runs along the erupting magnetic structure (or the erupting hot channel) and moves in the direction of the eruption during the precursor phase. Assuming the excitation mechanism of Helium atoms along the absorption front by non-thermal electrons, the phenomenon shows that the interaction between the erupted hot channel and the overlying (or surrounding) magnetic field has yielded electron acceleration.

A Study of Solar Flare Effects on the Geomagnetic Field Components during Solar Cycles 23 and 24

In this paper, we investigated the impact of solar flares on the horizontal (H), eastward (Y) and vertical (Z) components of the geomagnetic field during solar cycles 23 and 24 (SC23/24) using data of magnetometer measurements on the sunlit side of the Earth. We examined the relation between sunspot number and solar flare occurrence of various classes during both cycles. During SC23/24, we obtained correlation coefficient of 0.93/0.97, 0.96/0.96 and 0.60/0.56 for C-class, M-class and X-class flare, respectively. The three components of the geomagnetic field reached a peak a few minutes after the solar flare occurrence. Generally, the magnetic crochet of the H component was negative between the mid-latitudes and Low-latitudes in both hemispheres and positive at low latitudes. By contrast, the analysis of the latitudinal variation of the Y and Z components showed that unlike the H component, their patterns of variations were not coherent in latitude. The peak amplitude of solar flare effect (sfe) on the various geomagnetic components depended on many factors including the local time at the observing station, the solar zenith angle, the position of the station with respect to the magnetic equator, the position of solar flare on the sun and the intensity of the flare. Thus, these peaks were stronger for the stations around the magnetic equator and very low when the geomagnetic field components were close to their nighttime values. Both cycles presented similar monthly variations with the highest sfe value (ΔHsfe = 48.82 nT for cycle 23 and ΔHsfe = 24.68 nT for cycle 24) registered in September and lowest in June for cycle 23 (ΔHsfe = 8.69 nT) and July for cycle 24 (ΔHsfe = 10.69 nT). Furthermore, the sfe was generally higher in cycle 23 than in cycle 24.

Comparative case study of two methods to assess the eruptive potential of selected active regions

Solar eruptive events, like flares and coronal mass ejections, are characterized by the rapid release of energy that can give rise to emission of radiation across the entire electromagnetic spectrum and to an abrupt significant increase in the kinetic energy of particles. These energetic phenomena can have important effects on the space weather conditions and therefore it is necessary to understand their origin, in particular, what is the eruptive potential of an active region (AR). In these case studies, we compare two distinct methods that were used in previous works to investigate the variations of some characteristic physical parameters during the pre-flare states of flaring ARs. These methods consider: i) the magnetic flux evolution and magnetic helicity accumulation, and ii) the fractal and multi-fractal properties of flux concentrations in ARs. Our comparative analysisis based on time series of photospheric data obtained bythe Solar Dynamics Observatory between March 2011 and June 2013. We selected two distinct samples of ARs: one is distinguished by the occurrence of more energetic M- and X-class flare events, that may have a rapid effect on not just the near-Earth space, but also on the terrestrial environment; the second is characterized by no-flares or having just a few C- and B-class flares. We foundthat the two tested methods complement each other in their ability to assess the eruptive potentials of ARs and could be employed to identify ARs prone to flaring activity. Based on the presented case study, we suggest that using a combination of different methods may aid to identify more reliably the eruptive potentials of ARs and help to better understand the pre-flare states.

Fine-grained Solar Flare Forecasting Based on the Hybrid Convolutional Neural Networks*

Abstract Improving the performance of solar flare forecasting is a hot topic in the solar physics research field. Deep learning has been considered a promising approach to perform solar flare forecasting in recent years. We first used the generative adversarial networks (GAN) technique augmenting sample data to balance samples with different flare classes. We then proposed a hybrid convolutional neural network (CNN) model (M) for forecasting flare eruption in a solar cycle. Based on this model, we further investigated the effects of the rising and declining phases for flare forecasting. Two CNN models, i.e., M rp and M dp, were presented to forecast solar flare eruptions in the rising phase and declining phase of solar cycle 24, respectively. A series of testing results proved the following. (1) Sample balance is critical for the stability of the CNN model. The augmented data generated by GAN effectively improved the stability of the forecast model. (2) For C-class, M-class, and X-class flare forecasting using Solar Dynamics Observatory line-of-sight magnetograms, the means of the true skill statistics (TSS) scores of M are 0.646, 0.653, and 0.762, which improved by 20.1%, 22.3%, and 38.0% compared with previous studies. (3) It is valuable to separately model the flare forecasts in the rising and declining phases of a solar cycle. Compared with model M, the means of the TSS scores for No-flare, C-class, M-class, and X-class flare forecasting of the M rp improved by 5.9%, 9.4%, 17.9%, and 13.1%, and those of the M dp improved by 1.5%, 2.6%, 11.5%, and 12.2%.