Predicting Solar Flares Using SDO/HMI Vector Magnetic Data Products and the Random Forest Algorithm

Abstract

Adverse space weather effects can often be traced to solar flares, prediction of which has drawn significant research interests. The Helioseismic and Magnetic Imager (HMI) produces full-disk vector magnetograms with continuous high cadence, while flare prediction efforts utilizing this unprecedented data source are still limited. Here we report results of flare prediction using physical parameters provided by the Space-weather HMI Active Region Patches (SHARP) and related data products. We survey X-ray flares occurred from 2010 May to 2016 December, and categorize their source regions into four classes (B, C, M, and X) according to the maximum GOES magnitude of flares they generated. We then retrieve SHARP related parameters for each selected region at the beginning of its flare date to build a database. Finally, we train a machine-learning algorithm, called random forest (RF), to predict the occurrence of a certain class of flares in a given active region within 24 hours, evaluate the classifier performance using the 10-fold cross validation scheme, and characterize the results using standard performance metrics. Compared to previous works, our experiments indicate that using the HMI parameters and RF is a valid method for flare forecasting with fairly reasonable prediction performance. To our knowledge, this is the first time that RF is used to make multi-class predictions of solar flares. We also find that the total unsigned quantities of vertical current, current helicity, and flux near polarity inversion line are among the most important parameters for classifying flaring regions into different classes.

Predicting Solar Flares Using a Long Short-term Memory Network

“…We next compare our LSTM framework with five closely related machine learning methods including multilayer perceptrons (MLP) (Haykin & Network 2004;Florios et al 2018), Jordan network (JN) (Jordan 1997), support vector machines (SVM) (Qahwaji & Colak 2007;Yuan et al 2010;Bobra & Couvidat 2015;Boucheron et al 2015;Muranushi et al 2015;Florios et al 2018), random forests (RF) (Barnes et al 2016;Liu et al 2017;Florios et al 2018), and a recently published deep learning-based method, Deep Flare Net (DeFN; Nishizuka et al 2018). All these methods including ours (LSTM) can be used as a binary classification model (Nishizuka et al 2018;Jonas et al 2018) or a probabilistic forecasting model (Florios et al 2018).…”

“…Among these 16 features, there are 10 physical parameters, or SDO/HMI magnetic parameters, including TOTUSJH, TOTBSQ, TOTPOT, TOTUSJZ, ABSNJZH, SAVNCPP, USFLUX, AREA ACR, MEANPOT and TOTFX. All these 10 physical parameters except TOTFX are among the 13 magnetic parameters that are also considered important in Bobra & Couvidat (2015) and Liu et al (2017), which used different methods for assessing the importance of features. Thus, our findings are consistent with those reported in the literature.…”

“…The authors considered flares of M1.0 class or arXiv:1905.07095v1 [astro-ph.SR] 17 May 2019 higher, as defined by the peak 1-8Å flux measured by the Geostationary Operational Environmental Satellite system (GOES ). Liu et al (2017) took 13 parameters out of the 25 features and used them to perform multiclass predictions of solar flares. Nishizuka et al (2017) employed both photospheric vector-magnetic field data and chromospheric data to predict prominent flares.…”

“…Machine learning is a subfield of artificial intelligence, which grants computers abilities to learn from the past data and make predictions on unseen future data (Alpaydin 2009). Commonly used machine learning methods for flare prediction include decision trees (Yu et al 2009(Yu et al , 2010, random forests (Barnes et al 2016;Liu et al 2017;Florios et al 2018;Breiman 2001), k-nearest neighbors Huang et al 2013;Winter & Balasubramaniam 2015;Nishizuka et al 2017), support vector machines (Qahwaji & Colak 2007;Yuan et al 2010;Bobra & Couvidat 2015;Boucheron et al 2015;Muranushi et al 2015;Florios et al 2018), ordinal logistic regression (Song et al 2009), the least absolute shrinkage and selection operator (LASSO) (Benvenuto et al 2018;Jonas et al 2018), extremely randomized trees (Nishizuka et al 2017), and neural networks (Qahwaji & Colak 2007;Wang et al 2008;Colak & Qahwaji 2009;Higgins et al 2011;Ahmed et al 2013). Recently, Nishizuka et al (2018) adopted a deep neural network, named Deep Flare Net, for flare prediction.…”

Surveying the reach and maturity of machine learning and artificial intelligence in astronomy

“…Established subfields include: Solar astronomy . Machine learning has been used for classification of solar flares (e.g., Liu, , Liu et al, , and Benvenuto, Piana, Campi, & Massone, via a hybrid method using both supervised and unsupervised methods); clustering (e.g., Yang et al, presented the simulated annealing genetic (SAG) AI method to distinguish between the umbra, penumbra, and solar photosphere through a segmentation approach); and forecasting of coronal mass ejections with a SVM (Liu et al, ), and SVM and multilayer perceptrons (Inceoglu et al, ). A number of systematic comparison studies have been conducted in order to assess which ML methods perform best at forecasting solar events.…”