Solar Flare Prediction Using SDO/HMI Vector Magnetic Field Data with a Machine-Learning Algorithm

Abstract

We attempt to forecast M-and X-class solar flares using a machine-learning algorithm, called Support Vector Machine (SVM), and four years of data from the Solar Dynamics Observatory's Helioseismic and Magnetic Imager, the first instrument to continuously map the full-disk photospheric vector magnetic field from space. Most flare forecasting efforts described in the literature use either line-of-sight magnetograms or a relatively small number of ground-based vector magnetograms. This is the first time a large dataset of vector magnetograms has been used to forecast solar flares. We build a catalog of flaring and non-flaring active regions sampled from a database of 2,071 active regions, comprised of 1.5 million active region patches of vector magnetic field data, and characterize each active region by 25 parameters. We then train and test the machine-learning algorithm and we estimate its performances using forecast verification metrics with an emphasis on the True Skill Statistic (TSS). We obtain relatively high TSS scores and overall predictive abilities. We surmise that this is partly due to fine-tuning the SVM for this purpose and also to an advantageous set of features that can only be calculated from vector magnetic field data. We also apply a feature selection algorithm to determine which of our 25 features are useful for discriminating between flaring and non-flaring active regions and conclude that only a handful are needed for good predictive abilities.

Accepted for publication in the Astrophysical Journal

References in corpus (5)

• The Helioseismic and Magnetic Imager (HMI) Vector Magnetic Field Pipeline: SHARPs -- Space-weather HMI Active Region Patches
• The Helioseismic and Magnetic Imager (HMI) Vector Magnetic Field Pipeline: Overview and Performance
• Toward Reliable Benchmarking of Solar Flare Forecasting Methods
• Global Forces in Eruptive Solar Flares: The Lorentz force Acting on the Solar Atmosphere and the Solar Interior
• What is the relationship between photospheric flow fields and solar flares?

Cited by in corpus (117)

• The Challenge of Machine Learning in Space Weather Nowcasting and Forecasting
• Flare-productive active regions
• Why Is the Great Solar Active Region 12192 Flare-Rich But CME-Poor?
• Solar Flare Prediction Model with Three Machine-Learning Algorithms Using Ultraviolet Brightening and Vector Magnetogram
• Forecasting Solar Flares Using Magnetogram-based Predictors and Machine Learning
• Deep Flare Net (DeFN) model for solar flare prediction
• Predicting Solar Flares Using a Long Short-Term Memory Network
• Predicting Solar Flares Using SDO/HMI Vector Magnetic Data Product and Random Forest Algorithm
• Predicting Coronal Mass Ejections Using Machine Learning Methods
• A Comparison of Flare Forecasting Methods. II. Benchmarks, Metrics and Performance Results for Operational Solar Flare Forecasting Systems
• Identifying Solar Flare Precursors Using Time Series of SDO/HMI Images and SHARP Parameters
• Relative magnetic helicity as a diagnostic of solar eruptivity
• Flare Prediction Using Photospheric and Coronal Image Data
• A Machine Learning Dataset Prepared From the NASA Solar Dynamics Observatory Mission
• A New Tool for CME Arrival Time Prediction Using Machine Learning Algorithms: CAT-PUMA
• A hybrid supervised/unsupervised machine learning approach to solar flare prediction
• Predicting Solar Flares Using CNN and LSTM on Two Solar Cycles of Active Region Data
• How to Train Your Flare Prediction Model: Revisiting Robust Sampling of Rare Events
• Predicting solar flares with machine learning: investigating solar cycle dependence
• Numerical Simulations of Flare-productive Active Regions: delta-sunspots, Sheared Polarity Inversion Lines, Energy Storage, and Predictions
• Feature ranking of active region source properties in solar flare forecasting and the uncompromised stochasticity of flare occurrence
• Solar Flare Intensity Prediction with Machine Learning Models
• Prediction of Solar Flares Using Unique Signatures of Magnetic Field Images
• Why is a flare-rich active region CME-poor?
• Non-neutralized electric currents in solar active regions and flare productivity
• Using Machine Learning Methods to Forecast If Solar Flares Will Be Associated with CMEs and SEPs
• UFCORIN: A Fully Automated Predictor of Solar Flares in GOES X-Ray Flux
• A blueprint of state-of-the-art techniques for detecting quasi-periodic pulsations in solar and stellar flares
• Operational solar flare prediction model using Deep Flare Net
• Relationships Between Characteristics of the Line-of-Sight Magnetic Field and Solar Flare Forecasts
• Real-time flare prediction based on distinctions between flaring and non-flaring active region spectra
• Statistical study of magnetic non-potential measures in confined and eruptive flares
• The Vertical Current Approximation Nonlinear Force-Free Field Code - Description, Performance Tests, and Measurements of Magnetic Energies Dissipated in Solar Flares
• Stratification of physical parameters in a C-class solar flare using multi-line observations
• Solar Flare Prediction Using Magnetic Field Diagnostics Above the Photosphere
• A new method to quantify and reduce projection error in whole-solar-active-region parameters measured from vector magnetograms
• Implementation paradigm for supervised flare forecasting studies: a deep learning application with video data
• Towards Coupling Full-disk and Active Region-based Flare Prediction for Operational Space Weather Forecasting
• The nonpotentiality of coronae of solar active regions, the dynamics of the surface magnetic field, and the potential for large flares
• Fine-grained Solar Flare Forecasting Based on the Hybrid Convolutional Neural Networks
• Which Photospheric Characteristics are Most Relevant to Active-Region Coronal Mass Ejections?
• Toward a Next Generation Particle Precipitation Model: Mesoscale Prediction Through Machine Learning (a Case Study and Framework for Progress)
• Reliable Probability Forecast of Solar Flares: Deep Flare Net-Reliable (DeFN-R)
• Electric Current Neutralization in Solar Active Regions and Its Relation to Eruptive Activity
• Solar Flares Complex Networks
• Supervised convolutional neural networks for classification of flaring and nonflaring active regions using line-of-sight magnetograms
• Incorporating Physical Knowledge into Machine Learning for Planetary Space Physics
• Toward Improved Understanding of Magnetic Fields Participating in Solar Flares: Statistical Analysis of Magnetic Field within Flare Ribbons
• Decreasing False Alarm Rates in ML-based Solar Flare Prediction using SDO/HMI Data
• Measuring the magnetic origins of solar flares, CMEs and Space Weather
• Flare Forecasting Algorithms Based on High-Gradient Polarity Inversion Lines in Active Regions
• All-Clear Flare Prediction Using Interval-based Time Series Classifiers
• Using an Artificial Neural Network to Classify Multi-component Emission Line Fits
• A Framework for Designing and Evaluating Solar Flare Forecasting Systems
• Non-LTE inversions of a confined X2.2 flare: I. Vector magnetic field in the photosphere and chromosphere
• Statistical Relation between Solar Flares and Coronal Mass Ejections with Respect to Sigmoidal Structures in Active Regions
• Novel data analysis techniques in coronal seismology
• Self-consistent propagation of flux ropes in realistic coronal simulations
• DeepSun: Machine-Learning-as-a-Service for Solar Flare Prediction
• Identifying preflare spectral features using explainable artificial intelligence
• Solar Flare Index Prediction Using SDO/HMI Vector Magnetic Data Products with Statistical and Machine Learning Methods
• Solar Flares Forecasting Using Time Series and Extreme Gradient Boosting Ensembles
• Distinguishing between Flaring and Non-Flaring Active Regions: A Machine Learning Perspective
• Machine Learning Approaches to Solar-Flare Forecasting: Is Complex Better?
• Leveraging the Mathematics of Shape for Solar Magnetic Eruption Prediction
• Properties of Flare-Imminent versus Flare-Quiet Active Regions from the Chromosphere through the Corona II: NonParametric Discriminant Analysis Results from the NWRA Classification Infrastructure (NCI)
• Low Dimensional Convolutional Neural Network For Solar Flares GOES Time Series Classification
• The characteristics of flare- and CME-productive solar active regions
• Photospheric Shear Flows in Solar Active Regions and Their Relation to Flare Occurrence
• Explaining Full-disk Deep Learning Model for Solar Flare Prediction using Attribution Methods
• On the Performance of Multi-Instrument Solar Flare Observations During Solar Cycle 24
• Multi-Channel Auto-Calibration for the Atmospheric Imaging Assembly using Machine Learning
• Solar Active Region Magnetogram Image Dataset for Studies of Space Weather
• Unsupervised classification of simulated magnetospheric regions
• Automatic detection of small-scale EUV brightenings observed by the Solar Orbiter/EUI
• Improving the Arrival Time Estimates of Coronal Mass Ejections by Using Magnetohydrodynamic Ensemble Modeling, Heliospheric Imager data, and Machine Learning
• Machine learning reveals systematic accumulation of electric current in lead-up to solar flares
• Properties of Flare-Imminent versus Flare-Quiet Active Regions from the Chromosphere through the Corona I: Introduction of the AIA Active Region Patches (AARPs)
• Machine learning as a flaring storm warning machine: Was a warning machine for the September 2017 solar flaring storm possible?
• Global Energetics of Solar Flares. XI. Flare Magnitude Predictions of the GOES-Class
• Performance of Major Flare Watches from the Max Millennium Program (2001-2010)
• Global Energetics of Solar Flares. IX. Refined Magnetic Modeling
• Deep learning reconstruction of sunspot vector magnetic fields for forecasting solar storms
• What Are the Causes of Super Activity of Solar Active Regions?
• Lorentz Force Evolution Reveals the Energy Buildup Processes during Recurrent Eruptive Solar Flares
• Relationship between Successive Flares in the Same Active Region and Space-Weather HMI Active Region Patch (SHARP) Parameters
• Flux emergence and generation of flare-productive active regions
• Evaluation of Applicability of a Flare Trigger Model based on Comparison of Geometric Structures
• On the possibility to probe the flare productivity of an active region on the early stage of emergence
• Synchronization of Small-scale Magnetic Features, Blinkers, and Coronal Bright Points
• Application and interpretation of deep learning for identifying pre-emergence magnetic-field patterns
• Flare Productivity of Major Flaring Solar Active Regions: A Time-series Study of Photospheric Magnetic Properties
• A Deep Learning Approach to Generating Photospheric Vector Magnetograms of Solar Active Regions for SOHO/MDI Using SDO/HMI and BBSO Data
• Spectropolarimetric Inversion in Four Dimensions with Deep Learning (SPIn4D): I. Overview, Magnetohydrodynamic Modeling, and Stokes Profile Synthesis
• The STONE curve: A ROC-derived model performance assessment tool
• Inferring Line-of-Sight Velocities and Doppler Widths from Stokes Profiles of GST/NIRIS Using Stacked Deep Neural Networks
• A New Space Weather Tool for Identifying Eruptive Active Regions
• Transfer Learning in Spatial-Temporal Forecasting of the Solar Magnetic Field
• Using Multiple Instance Learning for Explainable Solar Flare Prediction
• Long-Term Evolution of Magnetic Fields in Flaring Active Region NOAA 12673
• Time Series Analysis of Photospheric Magnetic Parameters of Flare-quiet versus Flaring Active Regions: Scaling Properties of Fluctuations
• Data Handling and Assimilation for Solar Event Prediction
• Analysing AIA Flare Observations using Convolutional Neural Networks
• Exploring the potential of neural networks to predict statistics of solar wind turbulence
• Classifying Signatures of Sudden Ionospheric Disturbances
• The Random Hivemind: An Ensemble Deep Learner Application to Solar Energetic Particle Prediction Problem
• Automatic detection of Ellerman bombs using Deep Learning
• Solar Flare Prediction Using Long Short-term Memory (LSTM) and Decomposition-LSTM with Sliding Window Pattern Recognition
• Comparative case study of two methods to assess the eruptive potential of selected active regions
• Evaluating Pointing Strategies for Future Solar Flare Missions
• Bypassing the static input size of neural networks in flare forecasting by using spatial pyramid pooling
• Which Upstream Solar Wind Conditions Matter Most in Predicting Bz within Coronal Mass Ejections
• Correlation between active regions' spectra at high radio frequencies and solar flare occurrences
• A Comparative Analysis of Machine-learning Models for Solar Flare Forecasting: Identifying High-performing Active Region Flare Indicators
• ALEXIS: Recreating the X-Ray Emission from the Full Solar Disk as a Linear Combination of Discrete Regions in the Extreme Ultraviolet and Soft X-Rays
• Towards Hybrid Embedded Feature Selection and Classification Approach with Slim-TSF
• Image patch analysis of sunspots and active regions. II. Clustering via matrix factorization