Local Correlation Tracking Method Research Articles

Weekly earthquake prediction in a region of China based on an intensive precursor network AETA

Once a majority of earthquakes occur without prediction, it is very likely to have a huge impact on human society. To solve the worldwide challenging problem of earthquake prediction, our laboratory has developed a set of sensory systems to monitor the abnormal activity of geological signals before an earthquake happens in China. At present, more than 300 stations have been deployed, and the observation time has exceeded 4 years. Based on the various geological activities collected, a local correlation tracking method is used to capture signal anomalies before an earthquake, and then the ROC curve is used for the evaluation of the predictive accuracy. The method is applied in the Sichuan-Yunnan area weekly, verifying the forecast within a 91-week time frame and a 30-week time frame. The method proposed in this article has earthquake prediction ability with a rate of over 70%. It promotes and contributes to helping people avoid the fear of unpredictable earthquakes.

Imaging the Sun’s Near-surface Flows Using Mode-coupling Analysis

Abstract The technique of normal-mode coupling is a powerful tool with which to seismically image non-axisymmetric phenomena in the Sun. Here we apply mode coupling in the Cartesian approximation to probe steady, near-surface flows in the Sun. Using Doppler cubes obtained from the Helioseismic and Magnetic Imager on board the Solar Dynamics Observatory, we perform inversions on mode-coupling measurements to show that the resulting divergence and radial vorticity maps at supergranular length scales (∼30 Mm) near the surface compare extremely well with those obtained using the local correlation tracking method. We find that the Pearson correlation coefficient is ≥0.9 for divergence flows, while ≥0.8 is obtained for the radial vorticity.

Velocity Distribution Associated With EUV Disturbances Caused by Eruptive MFR

Extreme ultraviolet (EUV) disturbances are ubiquitous during eruptive phenomena like solar flare and Coronal Mass Ejection (CME). In this work, we have performed a three-dimensional (3D) magnetohydrodynamic numerical simulation of CME with an analytic magnetic fluxrope (MFR) to study the complex velocity distribution associated with EUV disturbances. When the MFR erupts upward, a fast shock (FS) appears as a 3D dome, followed by outward moving plasma. In the center of the eruptive source region, an expanding CME bubble and a current sheet continuously grow, both of which are filled by inward moving plasma. At the flanks of the CME bubble, a complex velocity distribution forms because of the dynamical interaction between inward and outward plasma, leading to the formation of slow shock (SS) and velocity separatrix (VS). We note two types of vortices near the VS, not mentioned in the preceding EUV disturbance simulations. In first type of vortex, the plasma converges toward the vortex center, and in the second type, the plasma spreads out from the center. The forward modeling method has been used to create the synthetic SDO/AIA images, in which the eruptive MFR and the FS appear as bright structures. Furthermore, we also deduce the plasma velocity field by utilizing the Fourier local correlation tracking method on the synthetic images. However, we do not observe the VS, the SS, and the two types of vortices in this deduced velocity field.

Thermodynamic Evolution of Solar Flare Supra-arcade Downflows

Abstract Solar flares are rapid energy release phenomena that appear as bright ribbons in the chromosphere and high temperature loops in the corona, respectively. Supra-arcade Downflows (SADs) are plasma voids that first come out above the flare loops and then move quickly toward the flare loop top during the decay phase of the flare. In our work, we study 20 SADs appearing in three flares. By differential emission measure (DEM) analysis, we calculate the DEM weighted average temperature and emission measure of the front region and the main body of SADs. It is found that the temperatures of the SAD front and body tend to increase during the course of SADs flowing downwards. The relationship between the pressure and temperature fits well with the adiabatic equation for both the SAD front and body, suggesting that the heating of SADs is mainly caused by adiabatic compression. Moreover, we also estimate the velocities of SADs via the Fourier Local Correlation Tracking method and find that increase of the temperature of the SAD front presents a correlation with the decrease of the SAD kinetic energy, while the SAD body does not, implying that the viscous process may also heat the SAD front in spite of a limited role.

Hydrodynamic Properties of the Sun’s Giant Cellular Flows

Abstract Measurements of the large cellular flows on the Sun were made by local correlation tracking of features (supergranules) seen in full-disk Doppler images obtained by the Helioseismic and Magnetic Imager (HMI) instrument on the NASA Solar Dynamics Observatory (SDO) satellite. Several improvements made to the local correlation tracking method allowed for more precise measurements of these flows. Measurements were made hourly over the nearly ten years of the mission-to-date. A four-hour time-lag between images was determined to give the best results as a compromise between increased feature displacement and decreased feature evolution. The hourly measurements were averaged over the 34 days that it takes to observe all longitudes at all latitudes to produce daily maps of the latitudinal and longitudinal velocities. Analyses of these flow maps reveal many interesting characteristics of these large cellular flows. While flows at all latitudes are largely in the form of vortices with left-handed helicity in the north and right-handed helicity in the south, there are key distinctions between the low-latitude and high-latitude cells. The low-latitude cells have roughly circular shapes, lifetimes of about one month, rotate nearly rigidly, do not drift in latitude, and do not exhibit any correlation between longitudinal and latitudinal flow. The high-latitude cells have long extensions that spiral inward toward the poles and can wrap nearly completely around the Sun. They have lifetimes of several months, rotate differentially with latitude, drift poleward at speeds approaching 2 m s−1, and have a strong correlation between prograde and equatorward flows. Spherical harmonic spectral analyses of maps of the divergence and curl of the flows confirm that the flows are dominated by the curl component with rms velocities of about 12 m s−1 at wavenumber ℓ = 10. Fourier transforms in time over 1024 daily records of the spherical harmonic spectra indicate two notable components—an m = ±ℓ feature representing the low-latitude component and an m = ±1 feature representing the high-latitude component. The dispersion relation for the low-latitude component is well represented by that derived for Rossby waves or r-modes. The high-latitude component has a constant temporal frequency for all ℓ indicating features advected by differential rotation at rates representative of the base of the convection zone high latitudes. The poleward motions of these features further suggest that the high-latitude meridional flow at the base of the convection zone is poleward—not equatorward.

High-resolution Spectroscopic Imaging of Counter-streaming Motions in Solar Active Region Magnetic Loops

Abstract We carried out high-resolution spectroscopic imaging in He i 10830 Å and H α for a set of active region (NOAA 12569) magnetic loops of different sizes (classified into short and long loops) with the Goode Solar Telescope at the Big Bear Solar Observatory on 2016 July 18. The long loops take the form of an chromospheric arch filament system, yet their extreme ultraviolet (EUV) counterparts are observed by the Atmospheric Imaging Assembly on board the Solar Dynamics Observatory. Animations of blue- and red-wing images give counter-streaming motions; i.e., chromospheric absorption features in blue- and red-wing images move in opposite directions at different strands. The moving pattern is detected with the local correlation tracking method and confirmed by Doppler shifts. We speculate that, combined with the results of wavelet analysis that gives obvious 4 minute oscillation along trailing polarity, counter-streaming motions for short loops could be powered by p-mode leakage. However, for counter-streaming motions in long loops, we show that unidirectional mass flows in two opposite directions are accompanied with simultaneous weak EUV brightenings. Heating processes, probably by magnetic reconnection at footpoints, may have occurred. In addition, plasma flows along the magnetic loops, tracked with absorption features in He i 10830 Å, are found to be ejected from and drained out into inter-granule lane areas at different ends of the loop system.

Photospheric plasma and magnetic field dynamics during the formation of solar AR 11190

Context.The Sun features on its surface typical flow patterns called the granulation, mesogranulation, and supergranulation. These patterns arise due to convective flows transporting energy from the interior of the Sun to its surface. The other well known elements structuring the solar photosphere are magnetic fields arranged from single, isolated, small-scale flux tubes to large and extended regions visible as sunspots and active regions.Aims.In this paper we will shed light on the interaction between the convective flows in large-scale cells as well as the large-scale magnetic fields in active regions, and investigate in detail the statistical distribution of flow velocities during the evolution and formation of National Oceanic and Atmospheric Administration active region 11190.Methods.To do so, we employed local correlation tracking methods on data obtained by the Solar Dynamics Observatory in the continuum as well as on processed line-of-sight magnetograms.Results.We find that the flow fields in an active region can be modelled by a two-component distribution. One component is very stable, follows a Rayleigh distribution, and can be assigned to the background flows, whilst the other component is variable in strength and velocity range and can be attributed to the flux emergence visible both in the continuum maps as well as magnetograms. Generally, the plasma flows, as seen by the distribution of the magnitude of the velocity, follow a Rayleigh distribution even through the time of formation of active regions. However, at certain moments of large-scale fast flux emergence, a second component featuring higher velocities is formed in the velocity magnitudes distribution.Conclusions.The plasma flows are generally highly correlated to the motion of magnetic elements and vice versa except during the times of fast magnetic flux emergence as observed by rising magnetic elements. At these times, the magnetic fields are found to move faster than the corresponding plasma.

Chromospheric impact of an exploding solar granule

Observations of multi-wavelength and therefore height-dependent information following events throughout the solar atmosphere and unambiguously assigning a relation between these rapidly evolving layers are rare and difficult to obtain. Yet, they are crucial for our understanding of the physical processes that couple the different regimes in the solar atmosphere. We characterize the exploding granule event with simultaneous observations of Hinode spectroplarimetric data in the solar photosphere and Hinode broadband CaIIH images combined with Interface Region Imaging Spectrograph (IRIS) slit spectra. We follow the evolution of an exploding granule and its connectivity throughout the atmosphere and analyze the dynamics of a magnetic element that has been affected by the abnormal granule. In addition to magnetic flux maps we use a local correlation tracking method to infer the horizontal velocity flows in the photosphere and apply a wavelet analysis on several IRIS chromospheric emission features such as MgIIk2v and MgIIk3 to detect oscillatory phenomena indicating wave propagation. During the vigorous expansion of the abnormal granule we detect radially outward horizontal flows, causing, together with the horizontal flows from the surrounding granules, the magnetic elements in the bordering intergranular lanes to be squeezed and elongated. In reaction to the squeezing, we detect a chromospheric intensity and velocity oscillation pulse which we identify as an upward traveling hot shock front propagating clearly through the IRIS spectral line diagnostics of MgIIh&k. Conclusion: Exploding granules can trigger upward-propagating shock fronts that dissipate in the chromosphere.

Data compression for local correlation tracking of solar granulation

Context. Several upcoming and proposed space missions, such as Solar Orbiter, will be limited in telemetry and thus require data compression. Aims. We test the impact of data compression on local correlation tracking (LCT) of time-series of continuum intensity images. We evaluate the effect of several lossy compression methods (quantization, JPEG compression, and a reduced number of continuum images) on measurements of solar differential rotation with LCT. Methods. We apply the different compression methods to tracked and remapped continuum intensity maps obtained by the Helioseismic and Magnetic Imager (HMI) onboard the Solar Dynamics Observatory. We derive 2D vector velocities using the local correlation tracking code FLCT and determine the additional bias and noise introduced by compression to differential rotation. Results. We find that probing differential rotation with LCT is very robust to lossy data compression when using quantization. Our results are severely affected by systematic errors of the LCT method and the HMI instrument. The sensitivity of LCT to systematic errors is a concern for Solar Orbiter.

HORIZONTAL FLOWS IN ACTIVE REGIONS FROM RING-DIAGRAM AND LOCAL CORRELATION TRACKING METHODS

ABSTRACT Continuous high-cadence and high spatial resolution Dopplergrams allow us to study subsurface dynamics that may be further extended to explore precursors of visible solar activity on the surface. Since the p-mode power is absorbed in the regions of high magnetic field, the inferences in these regions are often presumed to have large uncertainties. In this paper, using the Dopplergrams from space-borne Helioseismic Magnetic Imager, we compare horizontal flows in a shear layer below the surface and the photospheric layer in and around active regions. The photospheric flows are calculated using the local correlation tracking (LCT) method, while the ring-diagram technique of helioseismology is used to infer flows in the subphotospheric shear layer. We find a strong positive correlation between flows from both methods near the surface. This implies that despite the absorption of acoustic power in the regions of strong magnetic field, the flows inferred from the helioseismology are comparable to those from the surface measurements. However, the magnitudes are significantly different; the flows from the LCT method are smaller by a factor of 2 than the helioseismic measurements. Also, the median difference between the direction of corresponding vectors is 49°.

SUDDEN PHOTOSPHERIC MOTION AND SUNSPOT ROTATION ASSOCIATED WITH THE X2.2 FLARE ON 2011 FEBRUARY 15

The Helioseismic and Magnetic Imager provides 45 s cadence intensity images and 720 s cadence vector magnetograms. These unprecedented high-cadence and high-resolution data give us a unique opportunity to study the change of photospheric flows and sunspot rotations associated with flares. By using the differential affine velocity estimator method and the Fourier local correlation tracking method separately, we calculate velocity and vorticity of photospheric flows in the flaring NOAA AR 11158, and investigate their temporal evolution around the X2.2 flare on 2011 February 15. It is found that the shear flow around the flaring magnetic polarity inversion line exhibits a sudden decrease, and both of the two main sunspots undergo a sudden change in rotational motion during the impulsive phase of the flare. These results are discussed in the context of the Lorentz-force change that was proposed by Hudson et al. (2008) and Fisher et al. (2012). This mechanism can explain the connections between the rapid and irreversible photospheric vector magnetic field change and the observed short-term motions associated with the flare. In particular, the torque provided by the horizontal Lorentz force change agrees with what is required for the measured angular acceleration.

Comparisons of CME Morphological Characteristics Derived from Five 3D Reconstruction Methods

We compare different methods to reconstruct the three-dimensional (3D) CME morphology. The explored methods include geometric localisation, mask fitting, forward modeling, polarisation ratio and local correlation tracking plus triangulation. The five methods are applied to the same CME event, which occurred on August 7 2010. Their corresponding results are presented and compared, especially in their propagation direction and spatial extent in 3D. We find that mask fitting and geometric localisation method produce consistent results. Reconstructions including three-view observations are more precise than reconstructions done with only two views. Compared to the forward modeling method, in which a-priori shape of the CME geometry is assumed, mask fitting has more flexibility. Polarisation ratio method makes use of the Thomson scattering geometry. We find spatially the 3D CME derived from mask fitting lies mostly in the overlap region obtained with the polarisation method from COR2 A and B. In addition, mask fitting can help resolve the front/back ambiguity inherent in the polarisation ratio method. However, local correlation tracking plus triangulation did not show a consistent result with the other four methods. For reconstructions of a diffuse CME, when the separation angle between STEREO A and B is large, finding two corresponding points in a STEREO image pair becomes very difficult. Excluding the local correlation tracking method, the latitude of the CME's centre of gravity derived from the other methods deviates within one degree and longitude differs within 19 degrees.

TEMPORAL POWER SPECTRA OF THE HORIZONTAL VELOCITY OF THE SOLAR PHOTOSPHERE

We have derived the temporal power spectra of the horizontal velocity of the solar photosphere. The data sets for 14 quiet regions observed with the Gband filter of Hinode/SOT are analyzed to measure the temporal fluctuation of the horizontal velocity by using the local correlation tracking (LCT) method. Among the high resolution (~0.2") and seeing-free data sets of Hinode/SOT, we selected the observations whose duration is longer than 70 minutes and cadence is about 30 s. The so-called k-{\omega} diagrams of the photospheric horizontal velocity are derived for the first time to investigate the temporal evolution of convection. The power spectra derived from k-omega diagrams typically have a double power law shape bent over at a frequency of 4.7 mHz. The power law index in the high frequency range is -2.4 while the power law index in the low frequency range is -0.6. The root mean square of the horizontal speed is about 1.1 km/s when we use a tracer size of 0.4" in LCT method. Autocorrelation functions of intensity fluctuation, horizontal velocity, and its spatial derivatives are also derived in order to measure the correlation time of the stochastic photospheric motion. Since one of possible energy sources of the coronal heating is the photospheric convection, the power spectra derived in the present study will be of high value to quantitatively justify various coronal heating models.

Three-Dimensional Reconstruction of Active Regions

The STEREO mission provides an unprecedented opportunity to reconstruct the 3D configuration of solar features. In this work, we combine SECCHI/EUVI data from both spacecraft by means of a local correlation tracking method. The technique allows an automatic (without user intervention) matching of pixels in both images. This information is then used to triangulate the 3D coordinates of each pixel. We use the method in order to reconstruct and analyze the 3D structure of active regions. In particular, we focus on the extraction of coronal loop heights, observed nearly simultaneously in the 171, 195 and 284 A passbands. We compare the properties of loops in the different wavelengths and extract valuable information regarding their geometry. In particular, we demonstrate that some loops that look co-spatial in the 171 A and 195 A images have in fact different heights and thus occupy different volumes. Our results have important implications for multi-wavelength studies of coronal loops, especially for calculations using filter-ratio techniques.

Comparison of Large-Scale Flows on the Sun Measured by Time-Distance Helioseismology and Local Correlation Tracking

We present a direct comparison between two different techniques: time-distance helioseismology and a local correlation tracking method for measuring mass flows in the solar photosphere and in a near-surface layer. We applied both methods to the same dataset (MDI high-cadence Dopplergrams covering almost the entire Carrington rotation 1974) and compared the results. We found that, after necessary corrections, the vector flow fields obtained by these techniques are very similar. The median difference between directions of corresponding vectors is 24°, and the correlation coefficients of the results for mean zonal and meridional flows are 0.98 and 0.88, respectively. The largest discrepancies are found in areas of small velocities where the inaccuracies of the computed vectors play a significant role. The good agreement of these two methods increases confidence in the reliability of large-scale synoptic maps obtained by them.

A Check for Consistency between Different Magnetic Helicity Measurements Based on the Helicity Conservation Principle

Magnetic helicity is a useful quantity in characterizing the magnetic systems of solar active regions. The purpose of the present work is to check for consistency between the local correlation tracking (LCT) method used to measure helicity injection through the photosphere, and the linear force-free field (LFFF) method used to determine helicity in the corona, based on the principle of helicity conservation in the solar corona. We have calculated the amount of magnetic helicity injected through the photosphere during the first disk passage of AR 10696 using the LCT method initially described by Chae. We have also measured the coronal magnetic helicity as a function of time using the LFFF method. With a value for the force-free α, the coronal field is constructed from the extrapolation of the Solar and Heliospheric Observatory (SOHO) MDI magnetograms, then compared with the coronal loops in the EUV images taken by the SOHO EIT. The force-free α that best fits the loops is used to calculate the coronal helicity. From a careful comparison of different helicity measurements during each time interval, we have reached the core conclusion that our measurements follow the helicity conservation principle with an uncertainty of ~15% and hence support the consistency between the two different methods with the same amount of uncertainty.

Measurements of magnetic helicity injected through the solar photosphere

The amount of magnetic helicity injected into the corona through the photosphere can be measured by inferring the apparent motion of photospheric footpoints of fieldlines from a time series of line-of-sight magnetograms, as originally proposed by [Chae, J., 2001. Observational determination of the rate of magnetic helicity transport through the solar surface via the horizontal motion of field line footpoints. Astrophysical Journal 560, L95–L98]. In this paper, we present a revised version of Chae’s original method by adopting a better proxy for the flux density of magnetic helicity that was proposed by [Pariat, E., Démoulin, P., Berger, M.A., 2005. Photospheric flux density of magnetic helicity. Astronomy and Astrophysics 439, 1191–1203]. Moreover it is allowed to choose either the old technique of local correlation tracking (LCT) or the new method called the different affine velocity estimator (DAVE) method that was developed by [Schuck, P. W., 2006. Tracking magnetic footpoints with the magnetic induction equation. Astrophysical Journal 646, 1358–1391.] for inferring transverse velocity. We have applied the revised method using the DAVE method to a set of 96-min cadence, full-disk MDI magnetograms for AR 10696 and compared the results with those obtained using the LCT method. As a result, we have found that the helicity measurements using the DAVE method yield systematically higher values of helicity injection than those using the LCT method, but the discrepancy is moderately small, being less than 10%.

Horizontal and Vertical Flow Structure in Emerging Flux Regions

Abstract In order to obtain an overall view of the flow structure of convective gas in emerging flux regions (EFRs), we studied three EFRs in two solar active regions, NOAA 8218 and NOAA 10774. Using the Local Correlation Tracking method, we found several horizontally divergent flow structures, which were stable over a period of 1 hour, in 2 EFRs in NOAA 8218. The horizontal flow velocities and the sizes of the structures were around $500 \,\mathrm{m} \,\mathrm{s}^{-1}$ and about 4 mm in radius, respectively. We analyzed another dataset of NOAA 10774 using spectroscopic methods and found temporarily stable up-ward gas flows in the central part of the EFR. The line-of-sight velocities were around $150 \,\mathrm{m} \,\mathrm{s}^{-1}$ and the size of the flow patch was 2 to 5 mm in radius. These results support our previous findings that convective-cell-like flow appears in the central part of an EFR. We estimated from these results that the depth of the flow cell in EFRs is about 600 Km, and the turn-over time of the cell is about 2 hours.

Development of Real-Time Frame Selector 2 and the Characteristic Convective Structure in the Emerging Flux Region

Abstract A new image-grabbing system, Real-Time Frame Selector 2 (RTFS2), was developed and installed in the Domeless Solar Telescope at Hida Observatory. The purpose of this system is to acquire high-quality images in ground-based observations. We observed the emerging flux region NOAA 8582 in the G-band wavelengths from 20:59 UT to 23:32 UT on 1999 June 11 with RTFS2. Simultaneous $\mathrm{H}\alpha$ line center and wing wavelength observations with a Lyot filter system showed emerging flux loops and surge activities in this region. We applied a Local Correlation Tracking Method to the G-band data set in order to derive the horizontal convective structure in the upper convective zone. In addition to a meso-scale convective structure, we found divergent-flow structures under emerging flux loops, which were stable during the whole observation period, and located at the middle of the foot points of each flux loop. We suggest that the divergent-flow structure is a newly found characteristic of emerging flux regions.

Determination of magnetic helicity content of solar active regions from SOHO/MDI magnetograms

Chae (2001) first proposed a method of self-consistently determining the rate of change of magnetic helicity using a time series of longitudinal magnetograms only, such as taken by SOHO/MDI. Assuming that magnetic fields in the photosphere are predominantly vertical, he determined the horizontal component of velocity by tracking the displacements of magnetic flux fragments using the technique of local correlation tracking (LCT). In the present paper, after briefly reviewing the recent advance in helicity rate measurement, we argue that the LCT method can be more generally applied even to regions of inclined magnetic fields. We also present some results obtained by applying the LCT method to the active region NOAA 10365 under emergence during the observable period, which are summarized as follows. (1) Strong shearing flows were found near the polarity inversion line that were very effective in helicity injection. (2) Both the magnetic flux and helicity of the active region steadily increased during the observing period, and reached 1.2 × 1022 Mx and 8 ×1042 Mx2, respectively, 4.5 days after the birth of the active region. (3) The corresponding ratio of the helicity to the square of the magnetic flux, 0.05, is roughly compatible with the values determined by other studies using linear-force-free modeling. (4) A series of flares took place while the rate of helicity injection was high. (5) The choice of a smaller window size or a shorter time interval in the LCT method resulted in a bigger value of the LCT velocity and a bigger value of the temporal fluctuation of the helicity rate. (6) Nevertheless when averaged over a time period of about one hour or longer, the average rate of helicity became about the same within about 10%, almost irrespective of the chosen window size and time interval, indicating that short-lived, fluctuating flows may be insignificant in transferring magnetic helicity. Our results suggest that the LCT method may be applied to 96-minute cadence full-disk MDI magnetograms or other data of similar kind, to provide a practically useful, if not perfect, way of monitoring the magnetic helicity content of active regions as a function of time.