An official website of the United States government
Abstract We present a transformer model, named DeepHalo, to predict the occurrence of halo coronal mass ejections (CMEs). Our model takes as input an active region (AR) and a profile, where the profile contains a time series of data samples in the AR that are collected 24 hr before the beginning of a day, and predicts whether the AR would produce a halo CME during that day. Each data sample contains physical parameters, or features, derived from photospheric vector magnetic field data taken by the Helioseismic and Magnetic Imager on board the Solar Dynamics Observatory. We survey and match CME events in the Space Weather Database Of Notification, Knowledge, Information and the Large Angle and Spectrometric Coronagraph CME Catalog, and we compile a list of CMEs, including halo CMEs and nonhalo CMEs, associated with ARs in the period between 2010 November and 2023 August. We use the information gathered above to build the labels (positive vs. negative) of the data samples and profiles at hand, where the labels are needed for machine learning. Experimental results show that DeepHalo with a true skill statistic (TSS) score of 0.907 outperforms a closely related long short-term memory network with a TSS score of 0.821. To our knowledge, this is the first time that the transformer model has been used for halo CME prediction.
more »
« less
Flare to CME Association Integration
For integrating CME data to their solar sources, we perform a confidence-based scoring process that involves spatial and temporal data integration. For each GOES >C1.0 flare, we identify the likely CME candidate(s) with a temporal search using the start and peak times of flares and first detection time of CMEs. In this step, we check if the CMEs’ first detection time is between 30 minutes before the flare’s start time and 60 minutes after the flare’s peak time. Then, for each potential CME candidate, if any, we generate a confidence score between 1 to 5 (from lowest to highest) based on checking four additional criteria where each criteria gives an extra confidence point for the association: 1. Determine a potential one-to-one mapping between a given flare and a CME, in the case of only a single CME that satisfies the temporal search. 2. Check if the flare’s principal angle is in the same solar-disk quadrant as the CME’s principal angle, which is assumed as a 8 degree margin for boundary conditions. 3. Check if the difference between the flare’s principal angle and the CME’s principal angle (i.e., difference angle) is less than the CME’s observed width. 4. Check if the difference angle is less than 60 degrees threshold where such threshold is designated for wide CMEs (i.e., Halo and partial Halo) whose width almost always fulfill the width-based criterion in the previous criteria and aims to provide a more strict level of confidence. At the end of the integration procedure, CMEs are associated with flares using the maximum confidence score and therefore are connected to their most likely solar source. If there are multiple CMEs with the same high score, only those with the lowest difference angle between flare and the CME will be connected. Due to the fact that this process is not manually verified, we expect to generate certain ‘noisy’ data points. However, for overall model building, these data points should have a minimal statistical significance and hence a minimal impact.
more »
« less
- Award ID(s):
- 2104004
- PAR ID:
- 10435344
- Publisher / Repository:
- Harvard Dataverse
- Date Published:
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
We study the evolution of solar eruptive events by investigating the temporal relationships among magnetic reconnection, flare energy release, and the acceleration of coronal mass ejections (CMEs). Leveraging the optimal viewing geometry of the Solar TErrestrial RElations Observatory (STEREO) relative to the Solar Dynamics Observatory (SDO) and the Reuven Ramaty High-Energy Solar Spectroscopic Imager (RHESSI) during 2010–2013, we identify 12 events with sufficient spatial and temporal coverage for a detailed examination. STEREO and SDO data are used to measure the CME kinematics and the reconnection rate, respectively, and hard X-ray (HXR) measurements from RHESSI provide a signature of the flare energy release. This analysis expands upon previous solar eruptive event timing studies by examining the fast-varying features, or “bursts,” in the HXR and reconnection rate profiles, which represent episodes of energy release. Through a time lag correlation analysis, we find that HXR bursts occur throughout the main CME acceleration phase for most events, with the HXR bursts lagging the acceleration by 2 ± 9 minutes for fast CMEs. Additionally, we identify a nearly one-to-one correspondence between bursts in the HXR and reconnection rate profiles, with HXRs lagging the reconnection rate by 1.4 ± 2.8 minutes. The studied events fall into two categories: events with a single dominant HXR burst and events with a train of multiple HXR bursts. Events with multiple HXR bursts, indicative of intermittent reconnection and/or particle acceleration, are found to correspond with faster CMEs.more » « less
-
Abstract Decameter hectometric (DH; 1–14 MHz) type IV radio bursts are produced by flare-accelerated electrons trapped in postflare loops or the moving magnetic structures associated with the coronal mass ejections (CMEs). From a space weather perspective, it is important to systematically compile these bursts, explore their spectrotemporal characteristics, and study the associated CMEs. We present a comprehensive catalog of DH type IV bursts observed by the Radio and Plasma Wave Investigation instruments on board the Wind and Solar TErrestrial RElations Observatory spacecraft covering the period of white-light CME observations by the Large Angle and Spectrometric Coronagraph on board the Solar and Heliospheric Observatory mission between 1996 November and 2023 May. The catalog has 139 bursts, of which 73% are associated with a fast (>900 km s−1) and wide (>60°) CME, with a mean CME speed of 1301 km s−1. All DH type IV bursts are white-light CME-associated, with 78% of the events associated with halo CMEs. The CME source latitudes are within ±45°. Seventy-seven events had multiple-vantage-point observations from different spacecraft, letting us explore the impact of the line of sight on the dynamic spectra. For 48 of the 77 events, there were good data from at least two spacecraft. We find that, unless occulted by nearby plasma structures, a type IV burst is best viewed when observed within a ±60° line of sight. Also, bursts with a duration above 120 minutes have source longitudes within ±60°. Our inferences confirm the inherent directivity in the type IV emission. Additionally, the catalog forms a Sun-as-a-star DH type IV burst database.more » « less
-
Abstract Solar flares have been linked to some of the most significant space weather hazards at Earth. These hazards, including radio blackouts and energetic particle events, can start just minutes after the flare onset. Therefore, it is of great importance to identify and predict flare events. In this paper we introduce the Detection and EUV Flare Tracking (DEFT) tool, which allows us to identify flare signatures and their precursors using high spatial and temporal resolution extreme-ultraviolet (EUV) solar observations. The unique advantage of DEFT is its ability to identify small but significant EUV intensity changes that may lead to solar eruptions. Furthermore, the tool can identify the location of the disturbances and distinguish events occurring at the same time in multiple locations. The algorithm analyzes high temporal cadence observations obtained from the Solar Ultraviolet Imager instrument aboard the GOES-R satellite. In a study of 61 flares of various magnitudes observed in 2017, the “main” EUV flare signatures (those closest in time to the X-ray start time) were identified on average 6 minutes early. The “precursor” EUV signatures (second-closest EUV signatures to the X-ray start time) appeared on average 14 minutes early. Our next goal is to develop an operational version of DEFT and to simulate and test its real-time use. A fully operational DEFT has the potential to significantly improve space weather forecast times.more » « less
-
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
