Subscribe to the gold package and get unlimited access to Shamra Academy
Register a new userReal-time flare prediction based on distinctions between flaring and non-flaring active region spectra
130
0
0.0
(
0
)
Ask ChatGPT about the research
No Arabic abstract
With machine learning entering into the awareness of the heliophysics community, solar flare prediction has become a topic of increased interest. Although machine learning models have advanced with each successive publication, the input data has remained largely fixed on magnetic features. Despite this increased model complexity, results seem to indicate that photospheric magnetic field data alone may not be a wholly sufficient source of data for flare prediction. For the first time we have extended the study of flare prediction to spectral data. In this work, we use Deep Neural Networks to monitor the changes of several features derived from the strong resonant Mg II h&k lines observed by IRIS. The features in descending order of predictive capability are: The triplet emission at 2798.77 $text{AA}$, line core intensity, total continuum emission between the h&k line cores, the k/h ratio, line-width, followed by several other line features such as asymmetry and line center. Regions that are about to flare generate spectra which are distinguishable from non-flaring active region spectra. Our algorithm can correctly identify pre-flare spectra approximately 35 minutes before the start of the flare, with an AUC of 86 % and an accuracy, precision and recall of 80 %. The accuracy and AUC monotonically increases to 90 % and 97 % respectively as we move closer in time to the start of the flare. Our study indicates that spectral data alone can lead to good predictive models and should be considered as an additional source of information alongside photospheric magnetograms.
rate research
Read More
The evolution of magnetic helicity has a close relationship with solar eruptions and is of interest as a predictive diagnostic. In this case study, we analyse the evolution of the normalised emergence, shearing and total magnetic helicity components in the case of three flaring and three non-flaring active regions (ARs) using SHARPs (Spaceweather Helioseismic Magnetic Imager Active Region Patches) vector magnetic field data. The evolution of the three magnetic helicity components is analysed with wavelet transforms, revealing significant common periodicities of the normalised emergence, shearing and total helicity fluxes before flares in the flaring ARs. The three non-flaring ARs do not show such common periodic behaviour. This case study suggests that the presence of significant periodicities in the power spectrum of magnetic helicity components could serve as a valuable precursor for flares.
In a recent work, Kilcik et al. (2017), have detected the temporal and periodic behavior of sunspot counts (SSC) in flaring (i.e. C, M, or X class flares), and non-flaring active regions for the last two solar cycles, covering the period: 1996 - 2016. The main results obtained are: 1) The temporal behavior of monthly means of daily total SSC in flaring and non-flaring active regions are different and these differences are also varying from cycle to cycle; 2) The periodicities detected in SSC of flaring and non-flaring active regions are quite different and these variations are also different from one cycle to another; the highest detected period in the flaring active regions is 113 days, while there are much higher periodicities (327, 312, and 256 days) in non-flaring regions. The detection of typical different periodicities in flaring and non-flaring regions can suggests both important differences and physical interpretation in the magneto-hydrodynamic behavior of the Sun. For this reason in the present paper we show a further periodicity analysis of the sunspot counts in flaring and in non-flaring active regions using the same data source of that used by the above cited authors and applying a powerful wavelet analysis tool which is particularly useful to detect multiscale features of complex unsteady and unevenly sampled time series. In order to futher support the differences and similarities found in the time behavior of SSC in flaring and non-flaring regions, we also computed the behavior of the wavelet entropy, a proper time function which allow us to measure the degree of complexity in the dynamics of the related time series.
We analyzed temporal and periodic behavior of sunspot counts (SSCs) in flaring (C, M, or X class flares), and non-flaring active regions (ARs) for the almost two solar cycles (1996 through 2016). Our main findings are as follows: i) The temporal variation of monthly means of daily total SSCs in flaring and non-flaring ARs are different and these differences are also varying from cycle to cycle; temporal profile of non-flaring ARs are wider than the flaring ones during the solar cycle 23, while they are almost the same during the current cycle 24. The second peak (second maximum) of flaring ARs are strongly dominate during current cycle 24, while this difference is not such a remarkable during cycle 23. The amplitude of SSCs in the non-flaring ARs are comparable during the first and second peaks (maxima) of the current solar cycle, while the first peak is almost not existent in case of the flaring ARs. ii) Periodic variations observed in SSCs of flaring and non-flaring ARs are quite different in both MTM spectrum and wavelet scalograms and these variations are also different from one cycle to another; the largest detected period in the flaring ARs is 113 days, while there are much higher periodicities (327, 312, and 256 days) in non-flaring ARs. There are no meaningful periodicities in MTM spectrum of flaring ARs exceeding 45 days during solar cycle 24, while a 113 days periodicity detected from flaring ARs of solar cycle 23. For the non-flaring ARs the largest period is 72 days during solar cycle 24, while the largest period is 327 days during current cycle.
We analyze coordinated Hinode XRT and EIS observations of a non-flaring active region to investigate the thermal properties of coronal plasma taking advantage of the complementary diagnostics provided by the two instruments. In particular we want to explore the presence of hot plasma in non-flaring regions. Independent temperature analyses from the XRT multi-filter dataset, and the EIS spectra, including the instrument entire wavelength range, provide a cross-check of the different temperature diagnostics techniques applicable to broad-band and spectral data respectively, and insights into cross-calibration of the two instruments. The emission measure distribution, EM(T), we derive from the two datasets have similar width and peak temperature, but show a systematic shift of the absolute values, the EIS EM(T) being smaller than XRT EM(T) by approximately a factor 2. We explore possible causes of this discrepancy, and we discuss the influence of the assumptions for the plasma element abundances. Specifically, we find that the disagreement between the results from the two instruments is significantly mitigated by assuming chemical composition closer to the solar photospheric composition rather than the often adopted coronal composition (Feldman 1992). We find that the data do not provide conclusive evidence on the high temperature (log T[K] >~ 6.5) tail of the plasma temperature distribution, however, suggesting its presence to a level in agreement with recent findings for other non-flaring regions.
We present new constraints on the high-temperature emission measure of a non-flaring solar active region using observations from the recently flown Focusing Optics X-ray Solar Imager sounding rocket payload. FOXSI has performed the first focused hard X-ray (HXR) observation of the Sun in its first successful flight on 2012 November 2. Focusing optics, combined with small strip detectors, enable high-sensitivity observations with respect to previous indirect imagers. This capability, along with the sensitivity of the HXR regime to high-temperature emission, offers the potential to better characterize high-temperature plasma in the corona as predicted by nanoflare heating models. We present a joint analysis of the differential emission measure (DEM) of active region 11602 using coordinated observations by FOXSI, Hinode/XRT and Hinode/EIS. The Hinode-derived DEM predicts significant emission measure between 1 MK and 3 MK, with a peak in the DEM predicted at 2.0-2.5 MK. The combined XRT and EIS DEM also shows emission from a smaller population of plasma above 8 MK. This is contradicted by FOXSI observations that significantly constrain emission above 8 MK. This suggests that the Hinode DEM analysis has larger uncertainties at higher temperatures and that >8 MK plasma above an emission measure of 3x10^44 cm^-3 is excluded in this active region.
Log in to be able to interact and post comments
comments
Fetching comments
Sign in to be able to follow your search criteria
