Can you predict solar flares

Related: The sun's wrath: Worst solar storms in history. Scientists have long suspected that an effect known as magnetic reconnection underlies solar flares. This effect takes place when two magnetic regions with differently oriented field lines meet. When this happens, their magnetic field lines can break and reconnect with each other, explosively converting magnetic energy to heat and kinetic energy.

In the new study, researchers in Japan suggested magnetic reconnection can lead sheared magnetic loops to form unstable double-arc magnetic loops, which somewhat resemble the letter "m.

Assuming that double-arc instabilities do trigger solar explosions, the scientists developed a model to predict when large solar flares might occur based on routine magnetic observations of the sun. The model can also identify where these flares might happen and how much energy they might release.

The researchers tested their model using data on the largest, so-called "X-class" flares, collected using NASA's Solar Dynamics Observatory from to Want to read more? Register to unlock all the content on the site. E-mail Address. Sam Jarman is a science writer based in the UK. The Astrophysical Journal Letters Rapidly publish short notices of significant astrophysical research. Read previous Diagnostic imaging Research update Novel X-ray technology depicts real-time airflow through the lungs.

Read next States of matter Research update Experiments pin down conditions that make hot water freeze before cold. Discover more from Physics World. Kusano likens it to analyzing a snow-covered mountain slope to find the point that requires the least disturbance to trigger an avalanche.

The team was also able to identify a critical threshold for a parameter in their model, called kappa, that signals whether a flare is imminent — as well as predict roughly how strong the eventual flare will be based on the strength of the overlying magnetic field.

The team tested their so-called "kappa scheme" on roughly large groups of sunspots from to , seven of which resulted in large flares. The model successfully predicted six of the seven flares, giving early warning time of several hours to 24 hours. In each of those cases, the flare occurred at or near the site predicted by their model. Of the sunspot groups that didn't produce flares, the model predicted only three false positives.

Kusano's team's paper "seems an impressive piece of work to me," Bin Chen, a solar physicist at the New Jersey Institute of Technology in Newark, tells Astronomy. It's too early to definitively know whether the method will be more reliable than existing methods when it's up and running in real-time, says Veronig. But the fact that it's physics-based means that even the cases where it fails are interesting, because they may indicate that a different magnetic configuration is present.

Kusano says he aims to incorporate the model into an operational forecast within one or two years. However, he says the main challenge is getting access to enough computational power, as the models require supercomputers to run. While Kusano's team provides insight into the initial trigger of solar flares, this week, Chen and his colleagues reported the most detailed look yet at what happens during a flare, as it gathers and unleashes energy throughout the Sun's corona.

The work, published in Nature Astronomy on July 27 , is the first time researchers have been able to measure the strength of the magnetic field of the flare's "central engine," the region where the runaway reconnection takes place.

Surprisingly, Chen says the measured profile "matches beautifully" with the standard solar flare model proposed decades ago.

They also find that at the bottom of this region, the magnetic field is shaped like a bottle — a structure that naturally traps and accelerates charged particles. This is likely responsible for accelerating electrons from the flare to nearly the speed of light.

The two studies are generally consistent with each other, Chen and Kusano say. The double arc configuration — which Kusano thinks is responsible for triggering most large flares — could naturally give rise to a bottle shape as the flare unfolds. And Chen's first-of-its-kind direct observations of coronal magnetic fields could in turn improve the simulations that Kusano uses to extrapolate them. Still, Chen cautions, it's impossible to tell what triggers flares based on their data alone.

Veronig says the double arc idea is "for sure one possibility," but notes that there could be others as well. To put it another way: "Flares come in many flavors," says Veronig. Receive news, sky-event information, observing tips, and more from Astronomy's weekly email newsletter. View our Privacy Policy. This procedure will be more suitable for an observation dataset with a longer time period. We developed an operational flare prediction model using DNNs, which was based on a research version of the DeFN model, for operational forecasts.

DeFN has been continuously used for operational forecasting since January , and we evaluated its performance using the forecast and actual flare occurrences between January and June We found that operational DeFN achieved an accuracy of 0. Operational DeFN has the advantages of a large TSS, good discrimination performance, and the low probability of missed detection of observed flares.

This is why, it is useful for operations that require that no flares are missed, such as human activities in space and critical operations of satellites. On the other hand, it tends to over-forecast and the false alarm ratio FAR increases. Because the number of true negatives is very large in an imbalanced problem such as solar flare prediction, TSS is less sensitive to false positives than to false negatives. When we compared the evaluation results, we observed no significant difference between the pretrained and operational results.

This means that, at least during January —June , the difference between NRT and definitive series science data did not greatly affect the forecasts. We found a TSS of 0. This suggests that the chronological split is more suitable for the training and validation of the operational model than shuffle and split CV. Here, we discuss how to train and evaluate machine learning models for operational forecasting.

For an exact comparison, it is desirable to use the same datasets among participants. If this is not possible, there are three points that require attention. Observation database: The ratio of positive-to-negative events should not be artificially changed, and datasets should not be selected artificially. Data should be the climatological event rate and kept natural.

This is because some metrics are affected by controlling the positive-to-negative event ratio of datasets, especially HSS, which will result in a difference from the operational evaluations. For operational evaluations, it is also desirable to include ARs near the limb, although they are excluded in most papers, because the values of magnetograms are unreliable owing to the projection effect. Datasets for Training and Testing: We recommend that a chronological split or time-series CV is used for training and evaluation of operational models.

Although K-fold CV using random shuffling is common in solar flare predictions, it has a problem for a time-series dataset divided into two for training and testing when the time variation is very small, e. If the two neighboring datasets, which are very similar, are divided into both training and testing sets, the model becomes biased to overpredict flares. It might be true that a K-fold CV on data split by active region can also prevent data from a single active region being used in training and testing.

Therefore, in the point of view of generalization performance, a time-series CV is stricter and more suitable for operational evaluation. Selection of metrics: The ranking of models is easily affected by the selection of the metric. Depending on the purpose, users should select their preferred model by looking at the contingency tables and skill scores of each model. After understanding that each skill score can evaluate one aspect of performance, verification methods should be discussed in the space weather community see also Pagano et al.

In this paper, we showed contingency tables of our prediction results. We evaluated our prediction results as a deterministic forecasting model.

The ROC curve and the reliability diagram, which are shown in Barnes et al. We demonstrated the performance of a machine learning model in an operational flare forecasting scenario. The same methods and discussion of prediction using machine learning algorithms can be applied to other forecasting models of space weather in the magnetosphere and ionosphere.

Our future aim is to extend our model to predicting CMEs and social impacts on Earth by extending our database to include geoeffective phenomena and technological infrastructures. Nat Med 25 Article Google Scholar. Astrophys J — Weather Forecast — Astrophys J Bhattacharjee S, Alshehhi R, Dhuri DB, Hanasoge SM Supervised convolutional neural networks for classification of flaring and nonflaring active regions using line-of-sight magnetograms.

Astrophysical J Bishop CM Pattern recognition and machine learning. Springer, New York, p Google Scholar. Astropphys J Lett L Astropphys J Solar Phys 9 — Space Weather — Living Rev Solar Phys Monthly Notices R Astron Soc — Solar Phys — Colak T, Qahwaji R Automated solar activity prediction: a hybrid computer platform using machine learning and solar imaging for automated prediction of solar flares. Space Weather 7 6 :S Space Weather S J Space Weather Space Clim Solar Phys Falconer D, Barghouty AF, Khazanov I, Moore R A tool for empirical forecasting of major flares, coronal mass ejections, and solar particle events from a proxy of active-region free magnetic energy.

Space Weather 9:S Fletcher L et al An observational overview of solar flares. Space Sci Rev — Data processing and first results. Hanssen AW, Kuipers WJA On the relationship between the frequency of rain and various meteorological parameters: with reference to the problem ob objective forecasting Mededelingen en verhandelingen, 81, Royal Netherlans Meteorological Institute 65pp.

Results for line-of-sight magnetograms. Nat Commun Space Weather eSW Book Google Scholar. Nat Astron — Science — Space Weather Space Clim. Leka KD et al A comparison of flare forecasting methods. Astropphys J S Space Weather 13 11 —