Monica G. Bobra

The Helioseismic and Magnetic Imager (HMI) Vector Magnetic Field Pipeline: SHARPs – Space-Weather HMI Active Region Patches

A new data product from the Helioseismic and Magnetic Imager (HMI) onboard the Solar Dynamics Observatory (SDO) called Space-weather HMI Active Region Patches (SHARPs) is now available. SDO/HMI is the first space-based instrument to map the full-disk photospheric vector magnetic field with high cadence and continuity. The SHARP data series provide maps in patches that encompass automatically tracked magnetic concentrations for their entire lifetime; map quantities include the photospheric vector magnetic field and its uncertainty, along with Doppler velocity, continuum intensity, and line-of-sight magnetic field. Furthermore, keywords in the SHARP data series provide several parameters that concisely characterize the magnetic-field distribution and its deviation from a potential-field configuration. These indices may be useful for active-region event forecasting and for identifying regions of interest. The indices are calculated per patch and are available on a twelve-minute cadence. Quick-look data are available within approximately three hours of observation; definitive science products are produced approximately five weeks later. SHARP data are available at jsoc.stanford.edu and maps are available in either of two different coordinate systems. This article describes the SHARP data products and presents examples of SHARP data and parameters.

MODELING NONPOTENTIAL MAGNETIC FIELDS IN SOLAR ACTIVE REGIONS

Electric currents are present in the coronae above solar active regions, producing nonpotential magnetic fields that can be approximated as nonlinear force-free fields (NLFFFs). In this paper NLFFF models for two active regions observed in 2002 June are presented. The models are based on magnetograms from SOHO MDI and are constrained by nonpotential structures seen in BBSO Hα images and TRACE EUV images. The models are constructed using the flux rope insertion method. We find that the axial fluxes of the flux ropes are well constrained by the observations. The flux ropes are only weakly twisted, and electric currents flow mainly at the interface between the flux rope and its surroundings. In one case, the flux rope is anchored with both ends in the active region; in the other case, the flux rope extends to the neighboring quiet Sun. We find that the magnetic fields in these active regions are close to an eruptive state: the axial flux in the flux ropes is close to the upper limit for eruption. We also derive estimates for magnetic free energy and helicity in these regions.

SOLAR FLARE PREDICTION USING SDO/HMI VECTOR MAGNETIC FIELD DATA WITH A MACHINE-LEARNING ALGORITHM

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 Observatorys 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.

MAGNETIC HELICITY IN EMERGING SOLAR ACTIVE REGIONS

Using vector magnetic field data from the Helioseismic and Magnetic Imager instrument aboard the Solar Dynamics Observatory, we study magnetic helicity injection into the corona in emerging active regions (ARs) and examine the hemispheric helicity rule. In every region studied, photospheric shearing motion contributes most of the helicity accumulated in the corona. In a sample of 28 emerging ARs, 17 follow the hemisphere rule (61% ± 18% at a 95% confidence interval). Magnetic helicity and twist in 25 ARs (89% ± 11%) have the same sign. The maximum magnetic twist, which depends on the size of an AR, is inferred in a sample of 23 emerging ARs with a bipolar magnetic field configuration.

Fine Thermal Structure of a Coronal Active Region

The determination of the fine thermal structure of the solar corona is fundamental to constraining the coronal heating mechanisms. The Hinode X-ray Telescope collected images of the solar corona in different passbands, thus providing temperature diagnostics through energy ratios. By combining different filters to optimize the signal-to-noise ratio, we observed a coronal active region in five filters, revealing a highly thermally structured corona: very fine structures in the core of the region and on a larger scale further away. We observed continuous thermal distribution along the coronal loops, as well as entangled structures, and variations of thermal structuring along the line of sight.

PREDICTING CORONAL MASS EJECTIONS USING MACHINE LEARNING METHODS

Of all the activity observed on the Sun, two of the most energetic events are flares and coronal mass ejections (CMEs). Usually, solar active regions that produce large flares will also produce a CME, but this is not always true. Despite advances in numerical modeling, it is still unclear which circumstances will produce a CME. Therefore, it is worthwhile to empirically determine which features distinguish flares associated with CMEs from flares that are not. At this time, no extensive study has used physically meaningful features of active regions to distinguish between these two populations. As such, we attempt to do so by using features derived from (1) photospheric vector magnetic field data taken by the Solar Dynamics Observatorys Helioseismic and Magnetic Imager instrument and (2) X-ray flux data from the Geostationary Operational Environmental Satellites X-ray Flux instrument. We build a catalog of active regions that either produced both a flare and a CME (the positive class) or simply a flare (the negative class). We then use machine-learning algorithms to (1) determine which features distinguish these two populations, and (2) forecast whether an active region that produces an M- or X-class flare will also produce a CME. We compute the True Skill Statistic, a forecast verification metric, and find that it is a relatively high value of ~0.8 ± 0.2. We conclude that a combination of six parameters, which are all intensive in nature, will capture most of the relevant information contained in the photospheric magnetic field.

The Helioseismic and Magnetic Imager (HMI) Vector Magnetic Field Pipeline: Magnetohydrodynamics Simulation Module for the Global Solar Corona

Time-dependent three-dimensional magnetohydrodynamics (MHD) simulation modules are implemented at the Joint Science Operation Center (JSOC) of the Solar Dynamics Observatory (SDO). The modules regularly produce three-dimensional data of the time-relaxed minimum-energy state of the solar corona using global solar-surface magnetic-field maps created from Helioseismic and Magnetic Imager (HMI) full-disk magnetogram data. With the assumption of a polytropic gas with specific-heat ratio of 1.05, three types of simulation products are currently generated: i) simulation data with medium spatial resolution using the definitive calibrated synoptic map of the magnetic field with a cadence of one Carrington rotation, ii) data with low spatial resolution using the definitive version of the synchronic frame format of the magnetic field, with a cadence of one day, and iii) low-resolution data using near-real-time (NRT) synchronic format of the magnetic field on a daily basis. The MHD data available in the JSOC database are three-dimensional, covering heliocentric distances from 1.025 to 4.975 solar radii, and contain all eight MHD variables: the plasma density, temperature, and three components of motion velocity, and three components of the magnetic field. This article describes details of the MHD simulations as well as the production of the input magnetic-field maps, and details of the products available at the JSOC database interface. To assess the merits and limits of the model, we show the simulated data in early 2011 and compare with the actual coronal features observed by the Atmospheric Imaging Assembly (AIA) and the near-Earth in-situ data.

Classifying Signatures of Sudden Ionospheric Disturbances

Solar activity, such as flares, produce bursts of high-energy radiation that temporarily enhance the D-region of the ionosphere and attenuate low-frequency radio waves. To track these Sudden Ionospheric Disturbances (SIDs), which disrupt communication signals and perturb satellite orbits, Scherrer et al. (2008) developed an international, ground-based network of around 500 SID monitors that measure the signal strength of low-frequency radio waves. However, these monitors suffer from a host of noise contamination issues that preclude their use for rigorous scientific analysis. As such, we attempt to create an algorithm to automatically identify noisy, contaminated SID data sets from clean ones. To do so, we develop a set of features to characterize times series measurements from SID monitors and use these features, along with a binary classifer called a support vector machine, to automatically assess the quality of the SID data. We compute the True Skill Score, a metric that measures the performance of our classifier, and find that it is ~0.75+/-0.06. We find features characterizing the difference between the daytime and nighttime signal strength of low-frequency radio waves most effectively discern noisy data sets from clean ones.