Automatic Tracking of Active Regions and Detection of Solar Flares in Solar EUV Images

Solar catalogs are frequently handmade by experts using a manual approach or semi-automated approach. The appearance of new tools is very useful because the work is automated. Nowadays it is impossible to produce solar catalogs using these methods, because of the emergence of new spacecraft that provide a huge amount of information. In this article an automated system for detecting and tracking active regions and solar flares throughout their evolution using the Extreme UV Imaging Telescope (EIT) on the Solar and Heliospheric Observatory (SOHO) spacecraft is presented. The system is quite complex and consists of different phases: i) acquisition and preprocessing; ii) segmentation of regions of interest; iii) clustering of these regions to form candidate active regions which can become active regions; iv) tracking of active regions; v) detection of solar flares. This article describes all phases, but focuses on the phases of tracking and detection of active regions and solar flares. The system relies on consecutive solar images using a rotation law to track the active regions. Also, graphs of the evolution of a region and solar evolution are presented to detect solar flares. The procedure developed has been tested on 3500 full-disk solar images (corresponding to 35 days) taken from the spacecraft. More than 75 % of the active regions are tracked and more than 85 % of the solar flares are detected.     

Poisson Recoding Of Solar Images For Enhanced Compression

The amount of useful scientific data that a space-borne telescope produces is often limited by the available telemetry of the platform. General purpose image compression schemes are usually used to compress the image either lossy or losslessly. These schemes do not take into account the fact that pixel values of typical solar images are only known to within a certain uncertainty range. We present a preprocessing method to enhance the performance (compression ratio) of any subsequent image compression scheme. The method uses estimates of the photon shot and thermal noises to compute a recoding look-up table that maps the initial data into uncertainty intervals. The recoding method is lossy in a mathematical sense but lossless in a physical sense, since the image alterations are guaranteed to be smaller than the Poisson noise. The performance of any further compression algorithm is enhanced while achieving a known guaranteed maximum absolute error for each pixel in the case of lossless compression. This revised version was published online in October 2005 with corrections to one of the authors' names.     

Coronal Temperature Maps from Solar EUV Images: A Blind Source Separation Approach

Multi-wavelength solar images in the extreme ultraviolet (EUV) are routinely used for analysing solar features such as coronal holes, filaments, and flares. However, images taken in different bands often look remarkably similar, as each band receives contributions coming from regions with a range of different temperatures. This has motivated the search for empirical techniques that may unmix these contributions and concentrate salient morphological features of the corona in a smaller set of less redundant source images. Blind Source Separation (BSS) does precisely this. Here we show how this novel concept also provides new insight into the physics of the solar corona, using observations made by SDO/AIA. The source images are extracted using a Bayesian positive source-separation technique. We show how observations made in six spectral bands, corresponding to optically thin emissions, can be reconstructed by a linear combination of three sources. These sources have a narrower temperature response and allow for considerable data reduction, since the pertinent information from all six bands can be condensed into a single composite picture. In addition, they give access to empirical temperature maps of the corona. The limitations of the BSS technique and some applications are briefly discussed.     

A Comparative Study of Clustering Methods for Active Region Detection in Solar EUV Images

The increase in the amount of solar data provided by new satellites makes it necessary to develop methods to automate the detection of solar features. Here we present a method for automatically detecting active regions in solar extreme ultraviolet (EUV) images using a series of steps. Initially, the bright regions in the image are segmented using seeded region growing. In a second phase these bright regions are clustered into active regions. Partition-based clustering (both hard and fuzzy) and hierarchical clustering are compared in this work. The aim of the clustering phase is to associate a group to each segmented region in order to reduce the total number of active regions. This facilitates the documentation or subsequent monitoring of these regions. We use two indicators to validate the partitioning: i) the number of detected clusters approximates the number of active regions reported by the National Oceanic and Atmospheric Administration (NOAA) and ii) the area that defines each cluster overlaps with the area of an active region of NOAA. Experiments have been performed on over 6000 images from SOHO/EIT (195 Å). The best results were obtained using hierarchical clustering. The method detects a set of active regions in an image of the solar corona that successfully matches the number of NOAA regions. We will use these regions to perform real-time monitoring and flare detection.     

Segmentation of Loops from Coronal EUV Images

We present a procedure to extract bright loop features from solar EUV images. In terms of image intensities, these features are elongated ridge-like intensity maxima. To discriminate the maxima, we need information about the spatial derivatives of the image intensity. Commonly, the derivative estimates are strongly affected by image noise. We therefore use a regularized estimation of the derivative, which is then used to interpolate a discrete vector field of ridge points; these “ridgels” are positioned on the ridge center and have the intrinsic orientation of the local ridge direction. A scheme is proposed to connect ridgels to smooth, spline-represented curves that fit the observed loops. Finally, a half-automated user interface allows one to merge or split curves or eliminate or select loop fits obtained from this procedure. In this paper we apply our tool to one of the first EUV images observed by the SECCHI instrument onboard the recently launched STEREO spacecraft. We compare the extracted loops with projected field lines computed from near-simultaneous magnetograms measured by the SOHO/MDI Doppler imager. The field lines were calculated by using a linear force-free field model. This comparison allows one to verify faint and spurious loop connections produced by our segmentation tool and it also helps to prove the quality of the magnetic-field model where well-identified loop structures comply with field-line projections. We also discuss further potential applications of our tool such as loop oscillations and stereoscopy.     

Solar EUV flux variations near activity maxima

Data on solar emission variations in the extreme ultraviolet range <1300 (EUV-range) performed on board the Prognoz satellites and the Phobos spacecraft by the thermoluminescent method are presented. Flux variations from the 11-years cycle are factors of 2–2.5, and that by the 27-days cycle do not exceed 30%.     

Solar EUV Spectrum Calculated for Quiet Sun Conditions

We present spectral synthesis calculations of the solar extreme UV (EUV) in spherical symmetry carried out with the ‘Solar Modeling in 3D’ code. The calculations are based on one-dimensional atmospheric structures that represent a temporal and spatial mean of the chromosphere, transition region, and corona. The synthetic irradiance spectra are compared with the recent calibration spectrum taken with the EUV Variability Experiment during the Whole Heliospheric Interval. The good agreement between the synthetic and observed quiet Sun spectrum shows that the employed atmospheric structures are suitable for irradiance calculations. The validation of the quiet Sun spectrum for the present solar minimum is the first step toward the modeling of the EUV variations.     

Spatial and Temporal Noise in Solar EUV Observations

Solar telescopes will never be able to resolve the smallest events at their intrinsic physical scales. Pixel signals recorded by SOHO/(CDS, EIT, SUMER), STEREO/SECCHI/ EUVI, TRACE, SDO/AIA, and even by the future Solar Orbiter EUI/HRI contain an inherent “spatial noise” since they represent an average of the solar signal present at subpixel scales. In this paper, we aim at investigating this spatial noise, and hopefully at extracting information from subpixel scales. Two paths are explored. We first combine a regularity analysis of a sequence of EIT images with an estimation of the relationship between mean and standard deviation, and we formulate a scenario for the evolution of the local signal-to-noise ratio (SNR) as the pixel size becomes smaller. Second, we use an elementary forward modeling to examine the relationship between nanoflare characteristics (such as area, duration, and intensity) and the global mean and standard deviation. We use theoretical distributions of nanoflare parameters as input to the forward model. A fine-grid image is generated as a random superposition of those pseudo-nanoflares. Coarser resolution images (simulating images acquired by a telescope) are obtained by rebinning and are used to compute the mean and standard deviation to be analyzed. Our results show that the local SNR decays more slowly in regions exhibiting irregularities than in smooth regions.     

Photometric and Thermal Cross-calibration of Solar EUV Instruments

We present an assessment of the accuracy of the calibration measurements and atomic physics models that go into calculating the SDO/AIA response as a function of wavelength and temperature. The wavelength response is tested by convolving SDO/EVE and Hinode/EIS spectral data with the AIA effective area functions and by comparing the predictions with AIA observations. For most channels, the AIA intensities summed over the disk agree with the corresponding measurements derived from the current version (V2) of the EVE data to within the estimated 25 % calibration error. This agreement indicates that the AIA effective areas are generally stable in time. The AIA 304 Å channel, however, does show degradation by a factor of almost 3 from May 2010 through September 2011, when the throughput apparently reached a minimum. We also found some inconsistencies in the 335 Å passband, possibly due to higher-order contamination of the EVE data. The intensities in the AIA 193 Å channel agree to within the uncertainties with the corresponding measurements from EIS full CCD observations. Analysis of high-resolution X-ray spectra of the solar-like corona of Procyon and of EVE spectra allowed us to investigate the accuracy and completeness of the CHIANTI database in the AIA shorter wavelength passbands. We found that in the 94 Å channel, the spectral model significantly underestimates the plasma emission owing to a multitude of missing lines. We derived an empirical correction for the AIA temperature responses by performing differential emission measure (DEM) inversion on a broad set of EVE spectra and adjusting the AIA response functions so that the count rates predicted by the full-disk DEMs match the observations.     

Solar EUV line identifications from delayed beam-foil spectra

A number of electric dipole transitions from low-lying long-lived levels in Si-, P- and S-like ions of Fe and neighbouring elements have been identified by their longevity in delayed EUV spectra after foil excitation of fast ion beams. Some of these new identifications pertain to previously unidentified lines in solar flare spectra.     

Solar flares in the EUV observed from OSO-5

Solar flares in three broad EUV spectral bands have been observed from OSO-5 with a grating spectrophotometer. Results are given for three large flares of March 12, March 21 and April 21, 1969. In general the time dependence of flare intensity in each band is characterized by a slowly varying component with impulsive bursts superimposed. Bands 2 (465–630 Å) and 3 (760–1030 Å) are quite similar in their time variations, but band 1 (280–370 Å) shows less impulsive structure, and declines more slowly. Absolute EUV intensities for the flares are estimated, and a comparison made with the 2800 mc s^–1 radio emission. A flare model is proposed to account for the EUV time variations during a large flare.     

Detection of Quasi-Periodic Pulsations in Solar EUV Time Series

Quasi-periodic pulsations (QPPs) are intrinsically connected to the mechanism of solar flares. They are regularly observed in the impulsive phase of flares since the 1970s. In the past years, the studies of QPPs regained interest with the advent of a new generation of soft X-ray/extreme ultraviolet radiometers that pave the way for statistical surveys. Since the amplitude of QPPs in these wavelengths is rather small, detecting them implies that the overall trend of the time series needs to be removed before applying any Fourier or wavelet transform. This detrending process is known to produce artificial detection of periods that must then be distinguished from real ones. In this paper, we propose a set of criteria to help identify real periods and discard artifacts. We apply these criteria to data taken by the Extreme Ultraviolet Variability Experiment (EVE)/ESP onboard the Solar Dynamics Observatory (SDO) and the Large Yield Radiometer (LYRA) onboard the PRoject for On-Board Autonomy 2 (PROBA2) to search for QPPs in flares stronger than M5.0 that occurred during Solar Cycle 24.     

Intensity Distribution and Contrast of the Solar EUV Network

Intensity distributions of the EUV network and the cell interior in the solar atmosphere have been obtained in fourteen emission lines from Solar and Heliospheric Observatory (SOHO)/Coronal Diagnostic Spectrometer (CDS) observations. The formation temperature of the observed lines is in the range log T=4.90 – 6.06 (T in Kelvin), and hence they represent increasing heights in the solar atmosphere from the upper chromosphere and the transition region to the low corona. Intensity distributions of the cell interior have been found to be different in the quiet Sun and the coronal hole even at the lower transition region, which is at variance with some earlier results. The intensity contrast of the network with respect to the cell interior has been obtained for each line, and differences in the quiet Sun and the coronal hole have been examined. The network contrast, in general, is lower for the coronal hole as compared to the quiet Sun, but becomes equal to it in the upper transition region. The maximum contrast for both the regions is at about log T=5.3. Also obtained are the relative contributions of the network and the cell interior to the total intensity. The implications of the results for models of the transition region are briefly mentioned.     

Joint EUV/Radio Observations of a Solar Filament

In this paper we compare simultaneous extreme ultraviolet (EUV) line intensity and microwave observations of a filament on the disk. The EUV line intensities were observed by the CDS and SUMER instruments on board SOHO and the radio data by the Very Large Array and the Nobeyama radioheliograph. The main results of this study are the following: (1) The Lyman continuum absorption is responsible for the lower intensity observed above the filament in the EUV lines formed in the transition region (TR) at short wavelengths. In the TR lines at long wavelengths the filament is not visible. This indicates that the proper emission of the TR at the filament top is negligible. (2) The lower intensity of coronal lines and at radio wave lengths is due to the lack of coronal emission: the radio data supply the height of the prominence, while EUV coronal lines supply the missing hot matter emission measure (EM). (3) Our observations support a prominence model of cool threads embedded in the hot coronal plasma, with a sheath-like TR around them. From the missing EM we deduce the TR thickness and from the neutral hydrogen column density, derived from the Lyman continuum and Hei absorption, we estimate the hydrogen density in the cool threads.     

Solar rotation as determined from OSO-4 EUV spectroheliograms

Spectroheliograms obtained in extreme ultraviolet (EUV) lines and the Lyman continuum are used to determine the rotation rate of the solar chromosphere, transition region, and corona. A cross-correlation analysis of the observations indicates the presence of differential rotation through the chromosphere and transition region. The rotation rate does not vary with height. The average sidereal rotation rate is given by (deg day^–1) = 13.46 - 2.99 sin² B where B is the solar latitude. This rate agrees with spectroscopic determinations of the photospheric rotation rate, but is slower by 1 deg day^–1) = 13.46 - 2.99 sin² than rates determined from the apparent motion of photospheric magnetic fields and from the brightest points of active regions observed in the EUV. The corona does not clearly show differential rotation as do the chromosphere and transition region.     

Pioneer Venus Orbiter Measurements of Solar EUV Flux During Solar Cycles 21 and 22

The Pioneer Venus Orbiter (PVO) had on board the electron temperature probe experiment which measured temperature and concentration of electrons in the ionosphere of Venus. When the probe was outside the Venus ionosphere and was in the solar wind, the probe current was entirely due to solar photons striking the probe surface. This probe thus measured integrated solar EUV flux (Ipe) over a 13-year period from January 1979 to December 1991, thereby covering the declining phase of solar cycle 21 and the rising phase of solar cycle 22. In this paper, we examine the behavior of Ipe translated to the solar longitude of Earth (to be called EIpe) during the two solar cycles. We find that total EUV flux changed by about 60% during solar cycle 21 and by about 100% in solar cycle 22. We also compare this flux with other solar activity indicators such as F_10.7 , Lα, and the solar magnetic field. We find that while the daily values of EIpe are highly correlated with F_10.7 (correlation coefficient 0.87), there is a large scatter in EIpe for any value of this Earth-based index. A comparison of EIpe with SME and UARS SOLSTICE Lα measurements taken during the same period shows that EIpe tracks Lα quite faithfully with a correlation coefficient of 0.94. Similar comparison with the solar magnetic field (Bs) shows that EIpe correlates better with Bs than with F_10.7 . We also compare EIpe with total solar irradiance measured during the same period.     

Hard X-ray Source Distributions on EUV Bright Kernels in a Solar Flare

We explore the hard X-ray source distributions of an C1.1 flare occurred on 14 December 2007. Both Hinode/EIS and RHESSI observations are used. One of EIS rasters perfectly covers the double hard X-ray footpoints, where the EUV emission appears strong from the cool line of He ii (log T=4.7) to the hot line of Fe xvi (log T=6.4). We analyze RHESSI X-ray images at different energies and different times before the hard X-ray maximum. The results show a similar topology for the time-dependent source distribution (i.e. at 14:14:35 UT) as that for energy-dependent source distribution (i.e. at a given energy band of 6 – 9 keV) overlapped on EUV bright kernels, which seems to be consistent with the evaporation model.     

Solar rotation as measured in EUV chromospheric and coronal lines

Active regions were followed across the disk on OSO 4 spectroheliograms in the Lyman continuum (LC) and in Mg x 625. These observations indicate differential rotation with latitude, but not with height in the atmosphere. The measured equatorial sidereal rotation velocity is 14.7° ±0.2° per day in both chromospheric LC and coronal Mg x, where the quoted error is the standard deviation of a least-squares fit to the data.     

Center-to-Limb Variability of Hot Coronal EUV Emissions During Solar Flares

It is generally accepted that densities of quiet-Sun and active region plasma are sufficiently low to justify the optically thin approximation, and this is commonly used in the analysis of line emissions from plasma in the solar corona. However, the densities of solar flare loops are substantially higher, compromising the optically thin approximation. This study begins with a radiative transfer model that uses typical solar flare densities and geometries to show that hot coronal emission lines are not generally optically thin. Furthermore, the model demonstrates that the observed line intensity should exhibit center-to-limb variability (CTLV), with flares observed near the limb being dimmer than those occurring near disk center. The model predictions are validated with an analysis of over 200 flares observed by the EUV Variability Experiment (EVE) on the Solar Dynamics Observatory (SDO), which uses six lines, with peak formation temperatures between 8.9 and 15.8 MK, to show that limb flares are systematically dimmer than disk-center flares. The data are then used to show that the electron column density along the line of sight typically increases by \(1.76 \times 10^{19}~\mbox{cm}^{-2}\) for limb flares over the disk-center flare value. It is shown that the CTLV of hot coronal emissions reduces the amount of ionizing radiation propagating into the solar system, and it changes the relative intensities of lines and bands commonly used for spectral analysis.