skip to main content

Title: Operational prediction of solar flares using a transformer-based framework

Solar flares are explosions on the Sun. They happen when energy stored in magnetic fields around solar active regions (ARs) is suddenly released. Solar flares and accompanied coronal mass ejections are sources of space weather, which negatively affects a variety of technologies at or near Earth, ranging from blocking high-frequency radio waves used for radio communication to degrading power grid operations. Monitoring and providing early and accurate prediction of solar flares is therefore crucial for preparedness and disaster risk management. In this article, we present a transformer-based framework, named SolarFlareNet, for predicting whether an AR would produce a$$\gamma$$γ-class flare within the next 24 to 72 h. We consider three$$\gamma$$γclasses, namely the$$\ge$$M5.0 class, the$$\ge$$M class and the$$\ge$$C class, and build three transformers separately, each corresponding to a$$\gamma$$γclass. Each transformer is used to make predictions of its corresponding$$\gamma$$γ-class flares. The crux of our approach is to model data samples in an AR as time series and to use transformers to capture the temporal dynamics of the data samples. Each data sample consists of magnetic parameters taken from Space-weather HMI Active Region Patches (SHARP) and related data products. We survey flare events that occurred from May 2010 to December 2022 using the Geostationary Operational Environmental Satellite X-ray flare catalogs provided by the National Centers for Environmental Information (NCEI), and build a database of flares with identified ARs in the NCEI flare catalogs. This flare database is used to construct labels of the data samples suitable for machine learning. We further extend the deterministic approach to a calibration-based probabilistic forecasting method. The SolarFlareNet system is fully operational and is capable of making near real-time predictions of solar flares on the Web.

more » « less
Author(s) / Creator(s):
; ; ;
Publisher / Repository:
Nature Publishing Group
Date Published:
Journal Name:
Scientific Reports
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. Abstract

    Let us fix a primepand a homogeneous system ofmlinear equations$$a_{j,1}x_1+\dots +a_{j,k}x_k=0$$aj,1x1++aj,kxk=0for$$j=1,\dots ,m$$j=1,,mwith coefficients$$a_{j,i}\in \mathbb {F}_p$$aj,iFp. Suppose that$$k\ge 3m$$k3m, that$$a_{j,1}+\dots +a_{j,k}=0$$aj,1++aj,k=0for$$j=1,\dots ,m$$j=1,,mand that every$$m\times m$$m×mminor of the$$m\times k$$m×kmatrix$$(a_{j,i})_{j,i}$$(aj,i)j,iis non-singular. Then we prove that for any (large)n, any subset$$A\subseteq \mathbb {F}_p^n$$AFpnof size$$|A|> C\cdot \Gamma ^n$$|A|>C·Γncontains a solution$$(x_1,\dots ,x_k)\in A^k$$(x1,,xk)Akto the given system of equations such that the vectors$$x_1,\dots ,x_k\in A$$x1,,xkAare all distinct. Here,Cand$$\Gamma $$Γare constants only depending onp,mandksuch that$$\Gamma Γ<p. The crucial point here is the condition for the vectors$$x_1,\dots ,x_k$$x1,,xkin the solution$$(x_1,\dots ,x_k)\in A^k$$(x1,,xk)Akto be distinct. If we relax this condition and only demand that$$x_1,\dots ,x_k$$x1,,xkare not all equal, then the statement would follow easily from Tao’s slice rank polynomial method. However, handling the distinctness condition is much harder, and requires a new approach. While all previous combinatorial applications of the slice rank polynomial method have relied on the slice rank of diagonal tensors, we use a slice rank argument for a non-diagonal tensor in combination with combinatorial and probabilistic arguments.

    more » « less
  2. Abstract

    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.

    more » « less
  3. Abstract

    Based on the recent development of the framework of Volterra rough paths (Harang and Tindel in Stoch Process Appl 142:34–78, 2021), we consider here the probabilistic construction of the Volterra rough path associated to the fractional Brownian motion with$$H>\frac{1}{2}$$H>12and for the standard Brownian motion. The Volterra kernelk(ts) is allowed to be singular, and behaving similar to$$|t-s|^{-\gamma }$$|t-s|-γfor some$$\gamma \ge 0$$γ0. The construction is done in both the Stratonovich and Itô senses. It is based on a modified Garsia–Rodemich–Romsey lemma which is of interest in its own right, as well as tools from Malliavin calculus. A discussion of challenges and potential extensions is provided.

    more » « less
  4. Abstract

    We performed a differential emission measure (DEM) analysis of candle-flame-shaped flares observed with theAtmospheric Imaging Assemblyonboard theSolar Dynamic Observatory. The DEM profile of flaring plasmas generally exhibits a double peak distribution in temperature, with a cold component around$\log T\approx6.2$logT6.2and a hot component around$\log T\approx7.0$logT7.0. Attributing the cold component mainly to the coronal background, we propose a mean temperature weighted by the hot DEM component as a better representation of flaring plasma than the conventionally defined mean temperature, which is weighted by the whole DEM profile. Based on this corrected mean temperature, the majority of the flares studied, including a confined flare with a double candle-flame shape sharing the same cusp-shaped structure, resemble the famous Tsuneta flare in temperature distribution,i.e., the cusp-shaped structure has systematically higher temperatures than the rounded flare arcade underneath. However, the M7.7 flare on 19 July 2012 poses a very intriguing violation of this paradigm: the temperature decreases with altitude from the tip of the cusp toward the top of the arcade; the hottest region is slightly above the X-ray loop-top source that is co-spatial with the emission-measure-enhanced region at the top of the arcade. This signifies that a different heating mechanism from the slow-mode shocks attached to the reconnection site operates in the cusp region during the flare decay phase.

    more » « less
  5. Abstract

    We study thin films with residual strain by analyzing the$$\Gamma -$$Γ-limit of non-Euclidean elastic energy functionals as the material’s thickness tends to 0. We begin by extending prior results (Bhattacharya et al. in Arch Ration Mech Anal 228: 143–181, 2016); (Agostiniani et al. in ESAIM Control Opt Calculus Var 25: 24, 2019); (Lewicka and Lucic in Commun Pure Appl Math 73: 1880–1932, 2018); (Schmidt in J de Mathématiques Pures et Appliquées 88: 107–122, 2007) , to a wider class of films, whose prestrain depends on both the midplate and the transversal variables. The ansatz for our$$\Gamma -$$Γ-convergence result uses a specific type of wrinkling, which is built on exotic solutions to the Monge-Ampere equation, constructed via convex integration (Lewicka and Pakzad in Anal PDE 10: 695–727, 2017). We show that the expression for our$$\Gamma -$$Γ-limit has a natural interpretation in terms of the orthogonal projection of the residual strain onto a suitable subspace. We also show that some type of wrinkling phenomenon is necessary to match the lower bound of the$$\Gamma -$$Γ-limit in certain circumstances. These results all assume a prestrain of the same order as the thickness; we also discuss why it is natural to focus on that regime by considering what can happen when the prestrain is larger.

    more » « less