Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Statistics of Photospheric Supergranular Cells Observed by SDO/HMI

Statistics of Photospheric Supergranular Cells Observed by SDO/HMI Aims: The statistics of the photospheric granulation pattern are investigated using continuum images observed by Solar Dynamic Observatory (SDO)/Helioseismic and Magnetic Imager (HMI) taken at 6713 Å. Methods: The supergranular boundaries can be extracted by tracking photospheric velocity plasma flows. The local ball-tracking method is employed to apply on the HMI data gathered over the years 2011-2015 to estimate the boundaries of the cells. The edge sharpening techniques are exerted on the output of ball- tracking to precisely identify the cells borders. To study the fractal dimensionality (FD) of supergranulation, the box counting method is used. Results: We found that both the size and eccentricity follow the log-normal distributions with peak values about 330 Mm and 0.85, respectively. The five-year mean value of the cells number appeared in half-hour ′′ ′′ sequences is obtained to be about 60 ± 6 within an area of 350 × 350 . The cells orientation distribution presents the power-law behavior. Conclusions: The orientation of supergranular cells (O) and their size (S ) follows a power-law function 9.5 2 as |O| ∝ S . We found that the non-roundish cells with smaller and larger sizes than 600 Mm are aligned and perpendicular with the solar rotational velocity on the photosphere, respectively. The FD analysis shows that the supergranular cells form the self-similar patterns. Sun: photosphere - Sun: granulation - Methods: data analysis - Techniques: image processing I. Introduction and splitting old ones (Javaherian et al., 2014). The size of supergranules covers the extended range of scales, with a typi- The solar granulation is the upper side of cal diameters of 20-70 Mm (Priest, 2014; convective cells produced based on travel- Ryutova , 2015). The horizontal velocity ing hot plasma currents from the solar in- of the plasma flows from the cell centers to- terior (convective zone) to the photosphere ward the edges are estimated to be 0.2 − 0.5 (Priest, 2014). Hot plasma rises up to the −1 km s (Simon & Weiss, 1968; Ryutova , surface and transfers energy. Then, cold 2015). plasma returns to the interior within the dark boundaries. The granulation is a tur- Granulation process is linked to the mag- bulent process done by merging new grains netic flux distributed ubiquitously through- arXiv:1807.07479v1 [astro-ph.SR] 19 Jul 2018 Running title • May 2016 • Vol. XXI, No. 1 out the solar surface. It has been shown used Michelson Doppler Imager (MDI: that the emergence of magnetic flux is re- Scherrer et al., 1995) onboard on Solar lated to the cells where the plasma flows and Heliospheric Observatory (SOHO), to positively diverge (Stangalini, 2013). The show that the size of supergranules are large-scale type of granulation occurs in smaller at the maximum of solar activ- supergranular cells which is one of the ity. The intensity variation of supergranu- characteristics of the quiet Sun (Priest, lar cells from center to boundary is com- 2014). Studying the statistics of the surface parable with that of obtained for granules magneto-convective features lead to better (Del Moro et al., 2007). understanding of their evolution. Meunier et al. (2007) have presented re- Photospheric convective pattern high- sults about cell-size distribution and found lights the evolution of other phenomena, a correlation between horizontal veloc- such as coronal bright points, magnetic ity of plasma and supergranular radius cancelation, nanoflares and network extracted from MDI/SOHO. In recent flares in different layers (Ryutova et al., decades, by increasing received data taken 2003; Tajfirouze & Safari , 2012; from ground-based and space telescopes, Yousefzadeh et al., 2016). Tian et al. the automatic detection methods are inten- (2010) used data in various passbands sively expanded to analyze data and ex- to find out a correlation between the tract statistics with higher accuracy (e.g., horizontal velocities of plasma in both see Aschwanden, 2010; Alipour & Safari, photospheric and chromospheric super- 2015; Arish et al., 2016; Javaherian et al., granulation. In one of the statistical works, 2017), and also, reducing costs of data clas- the average diameter and lifetime of su- sification. So, the numerous methods are pergranular cells were found to be 25 Mm progressed to extract supergranular cells and 1.5 days, respectively (Roudier et al., from intensity continuum images. One of 2014). The relation between the cell size the important approaches, known as local and the magnetic field is unclear; but, in correlation tracking (LCT), is employed recent works done based on local corre- to recognize the boundaries based on flux lation tracking (LCT), the cell size and current in data (Papadimitriou et al., 2006). velocity are linked to the intensity map of The velocity of pixels of interests are com- both supergranular vertical and horizontal puted by capturing the transform motion of flows (Rincon et al., 2017). This indirect pixels in two consecutive frames. A two- relation between the cell size and magnetic dimensional (2-D) flow field can represent flux inside the cells leads to anti-correlated the boundaries of cells as the output of the dependence, so the large cells can be method. Fisher & Welsch (2008) extended emerged where the local magnetic field is the LCT method with application of the weak. fourier transform named fourier local cor- relation tracking (FLCT). The relation between supergranular at- tributes and the solar cycle, specially, the One of the promising algorithms for de- cell size and intensity variation have been termining the supergranular boundaries is studied by Meunier et al. (2008). They ball-tracking method. This method consid- 2 Running title • May 2016 • Vol. XXI, No. 1 ers imaginary balls on the gridded surface recorded by HMI. The HMI provides dif- independently moving with plasma flows ferent level of full-disk images in the Fe I (Potts et al., 2004). According to the in- absorption line at 6173 Å with a resolution tensities, the balls move in directions to of 0.50 ± 0.01 arcsec and cadence of 45 sec- settle in local minima where the bound- onds. Some corrections, such as exposure aries are elongated. Using balls track- time, dark current, flat field, and cosmic- ing, the locations of some coronal small- ray hits, are done in level-1 data. scale features (bright points, mini-coronal For our purpose, we used 30-minute con- mass ejections, etc.) were carried out secutive continuum HMI data with a time by Innes et al. (2009), Yousefzadeh et al. lag of 45 seconds in every two days from (2016), and Honarbakhsh et al. (2016). the year 2011 to 2015. Since the measure- We investigate the supergranules mor- ment of morphological parameters of the phological parameters and velocities dur- supergranules, such as size and orientation, ing a five-year period of the solar activ- and also, velocity on the surface very sensi- ity. So, the statistical parameters of su- tive to the projection effect, variation of the pergranular cells, such as size−frequency solar radial over time, and the B angle evo- distribution, fractal dimension (FD), orien- lution effect (Roudier et al., 2013), the par- ′′ ′′ tation, and their eccentricities are studied. tial area with a size of 350 × 350 at the Moreover, the correlations between quan- solar disk center (a region centered with tified parameters and the solar activity are longitudes ±11 around the central merid- computed. We used data recorded by Solar ian, and latitudes limited in ±11 around Dynamic Observatory (SDO)/ Helioseis- the equator) (Fig. 1, red box) was selected mic and Magnetic Imager (HMI) taken at to focus on photospheric flows. To coalign 6173 Å. the sequential data to a reference one, all The layout of this article is as follows: images are derotated using drot map.pro the description of data sets is given in Sect. available in the SSW/IDL package. II. The brief review of the methods are pre- pared in Sect. III. The results are presented III. Methods in Sect. IV. Concluding remarks are ex- The ball-tracking method, edge sharpening plained in Sect. V. technique, and box-counting algorithm are explained briefly as follows. II. Description of Datasets Ball-tracking One of the applicable Solar Dynamic Observatory (SDO) uti- methods developed for computing velocity lized one of the three instruments named fields is ball-tracking method (Potts et al., Helioseismic and Magnetic Imager (HMI: 2004). The ball-tracking method is ap- Schou et al., 2012) to investigate the pho- plied on continuum HMI images to track tospheric oscillations and magnetic fields velocity fields. Using fast fourier transform (Wachter et al., 2012). So, to study which is a part of the ball-tracking code, the supergranular cells using photospheric photospheric p-mode oscillations are at- continuum images, we employed the tenuated to remove features moving faster −1 high-spatial and temporal resolution data than 7 km s . In the code, the constructed 3 Running title • May 2016 • Vol. XXI, No. 1 (a) (b) Figure 1: The continuum SDO/HMI full-disk image of the Sun recorded on 30 December 2015 (00:00-00:30 UT) taken ′′ ′′ at 6173 Å (left panel). The red rectangle is selected with area of 350 × 350 (region centered with longitudes ◦ ◦ ±11 around the central meridian, and the latitudes limited in ±11 around the equator). The output of the ball-tracking method is shown with green face (right panel). The blue edges are the cells border obtained by the morphological filters (see text). data cube with two spatial and one tempo- oscillations (υ = ω/k) indicates a cut−off ral dimension is converted to a cube in fea- cone lateral surface in k − ω space, we ture space of wave vector (k) and frequency are able to discard components covering −1 (ω) by Fourier transform. As photosheric motions greater than υ = 7 km s cut−off 4 Running title • May 2016 • Vol. XXI, No. 1 (Tian et al., 1989; Roudier et al., 2013). tion. To speed up performing the binary dilation, erosion procedure is applied on Ball-tracking method delineates velocity output image and the lanes were nearly fields in a fraction of time by moving spher- closed (Boomgaard & Balen, 1992). To ical balls as float tracers on a surface. In find perimeter of structures in 2-D images, this method, the granulation pattern is con- the algorithm uses connectivities to specify sidered as a criterion for 3-D tracers on sur- the edges with more width. The output im- face. Small floating balls move on the so- age including boundaries is presented with lar photosphere based on bouncy laws. In solid blue lines in Fig. 1 (right panel). other words, merging and splitting of gran- Box-counting Fractals have geomet- ules reveal bumps on surface wherever the ric repetitive patterns in different scales time-evolving granulation form ripples on wherein the whole of structure can be gen- the surface. These movements are very erated by small parts in non-integer dimen- similar to a ball moving on fluid surface. sions (Aschwanden, 2011). One of the This ball has a given mass and momen- methods that estimates the fractal dimen- tum. If granular cells push the ball, the sion of 2-D data is box-counting (Molteno, sphere will continue traveling based on in- 1993). This method is applied on image coming force. So, the path and direction to breaks it into smaller parts with differ- of ball are estimated as a function of time ent resolutions step-by-step. In different (Potts & Diver, 2008). Both the diameter resolutions, the code considers the boxes of balls and average surface penetration de- consisting of image components. In each pend on the resolution of the image. The ra- step, the size of box (ǫ) and the number of dius of ball is chosen to match the typical boxes (N(ǫ)) is computed. Thus, the image size of granular cells. The smaller the ra- dimension (D) is defined as dius, the cell borders are found with more sensitivity to short wavelength noise. Typi- cal value for the radius is half of center-to- log N(ǫ) D = lim . (1) center granular distance. For high resolu- −1 ǫ→0 log(ǫ ) tion HMI data, the ball diameter is around two pixels. Time range of 30 minutes is IV. Results selected to prepare images as a datacube. Figure 1 (right panel) exhibits the ball- To determine the boundaries, an auto- tracking output extracted from HMI data matic recognition method is applied to displayed as regions with green face. HMI intensity-continuum data recorded at 6173 Å. We focused on rectangular area Edge Detection To sharpen the edges, ′′ ′′ of 350 × 350 to extract statistical prop- the binary format of images underwent the bridge algorithm that fill the blank erties of supergranular cells (Fig. 1, left panel). For our purpose, 40 images (half- space appeared between unclosed bound- hour data) for every two days were gath- aries (Gonzalez et al., 2008). In the next step, to fill remained gaps in boundaries, a ered with a time lag of 45 seconds during the years 2011 to 2015. decomposition algorithm is used to struc- turalize edges (if it exists) by image dila- The log-normal function is fitted to 5 Running title • May 2016 • Vol. XXI, No. 1 the size-frequency distribution of super- The peak value of the eccentricity distri- granules, (Fig. 2, upper panel). The bution of supergranular cells is 0.8. The log-normally distributed function is given skewness takes a positive value and kur- by (Newman, 2006; Bazargan et al., 2008; tosis is 1.8 for overall five years. The Aschwanden, 2015), skewness and kurtosis of eccentricity distri- butions are approximately constant during five years (Fig. 2, lower panel). 1 (ln(x) − µ ) f (x, µ , σ) = exp − , (2) 2 The cell orientation between the major 2σ σx 2π and horizontal axis (west-east latitude on where µ is the mean value, and σ is the solar disk) and an estimation of the the standard deviation. The fit parame- error value of the orientation angles are ters µ and σ for overall five-year size- specified for each region (see Appendix A). frequency are obtained to be 5.208 ± 0.038 The orientation distribution of cells follows and 0.757 ± 0.031, respectively, with the −α a power-law fit as y ∝ x , wherein α peak value of 330 Mm . The variation of is the power exponent. Using the meth- these values is approximately constant dur- ods introduced in Aschwanden (2015) and ing five years (Fig. 2, upper panel). Farhang et al. (2018), the power-law expo- 2011 with μ ≈ 5.21 and σ ≈ 0.76 nents were obtained. In Fig. 3, the fitted 2012 with μ ≈ 5.19 and σ ≈ 0.76 2013 with μ ≈ 5.30 and σ ≈ 0.73 power-law function is shown with the ex- 2014 with μ ≈ 5.28 and σ ≈ 0.74 2015 with μ ≈ 5.19 and σ ≈ 0.77 ponent α = 0.1613 ± 0.0416 which seems 0 1000 2000 3000 4000 5000 6000 7000 to be constant over the period. The errors Size (Mm ) -16 for the orientations of the cells were in the 2011 with kurtosis ≈ 1.79 and skewness ≈ 5× 10 -16 2 2012 with kurtosis ≈ 1.79 and skewness ≈ 5× 10 ◦ ◦ -16 range of 0.0001 to 1.7321 with mean val- 2013 with kurtosis ≈ 1.79 and skewness ≈ 4× 10 -16 2014 with kurtosis ≈ 1.79 and skewness ≈ 6× 10 -16 2015 with kurtosis ≈ 1.79 and skewness ≈ 3× 10 ues of 0.0584 . In Fig. 4, the log-normal function is fit- ted on the photospheric horizontal veloci- 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 Eccentricity ties of plasma for each year (Fig. 4). The overall five-year parameters µ and σ are Figure 2: The histograms of cells size (top) and cells eccentricity (bottom) for each year. The log- equal to 4.634 ± 0.039 and 0.556 ± 0.031, normal function (Eq. 2) is fitted to the size dis- respectively. The time series of daily (blue) tributions. and monthly (red) number of cells for five years are shown in Fig. 5. The mean value Each segmented region can be sur- is equal to 59.6 ± 5.62 cells ranged from rounded by an ellipse that can be at- 43 to 77 in each frame. tained by morphological moments of a shape (Emre Celebi & Alp Aslandogan , As we see in Fig. 6, the time series 2005). Thus, the major and minor axis of of the plasma velocity and the number of contained area is computed to obtain the ec- sunspots are presented. To test Pearson centricity. All cells eccentricities take val- correlation between time series, we em- ues ranged from 0 and 1. The value 0 rep- ployed the box-Cox transformation to con- resents the shape as a circle, and 1 returns vert the distributions of time series with a segmented shape as line. positive values (number of cells, number Number (× 10 ) Number (× 10 ) Running title • May 2016 • Vol. XXI, No. 1 4 4 10 10 (a) power index, α ≈ - 0.188 ± 0.086 (b) Power index, α ≈ - 0.1091 ± -0.0282 3 3 Data point Data point Power-law fit Power-law fit 10 10 4 4 10 10 (c) (d) power index, α ≈ - 0.2096 ± 0.0541 power index, α ≈ - 0.1355 ± 0.0350 3 3 10 10 Data point Data point Power-law fit Power-law fit 2 2 10 10 (e) power index, α ≈ - 0.1357 ± 0.0350 (f) Data point Power-law fit Data point Power-law fit power index, α ≈ - 0.1613 ± 0.0416 2 2 10 10 0 1 2 0 1 2 10 10 10 10 10 10 Orientation (degree) Orientation (degree) Figure 3: The orientation distribution of supergranules for the years of 2011 − 2015 (a: 2011, b: 2012, c: 2013, d: 2014, e: 2015, f: 2011-2015). The minimum and maximum of orientation errors computed for regions are ◦ ◦ ◦ about 0.0001 and 1.7321 , respectively. The maximum error values belong to the cells orientated in ±45 . × 10 3.5 2011 with μ ≈ 4.66 and σ ≈ 0.55 y − 1 2012 with μ ≈ 4.70 and σ ≈ 0.53 3   , if λ , 0, 2013 with μ ≈ 4.66 and σ ≈ 0.54 Y(λ) = 2014 with μ ≈ 4.68 and σ ≈ 0.53 2.5 2015 with μ ≈ 4.65 and σ ≈ 0.54 log(y), if λ = 0, where λ usually varies from -5 to 5. The op- 1.5 timal value of λ is obtained as the best ap- proximation fitted on the normal distribu- tion curve (Everitt, 2002). To do this, first, 0.5 kurtosis for each transformed time series with value about three are chosen for the 0 200 400 600 800 1000 1200 best λ. Then, the correlation coefficients Velocity (m/s) are computed for time series (Press et al., Figure 4: The histograms of horizontal velocities for 2007). each year during the years 2011 to 2015. The To validate the values attained by Pear- log-normal function (Eq. 2) is fitted to the dis- tributions. son correlation, a hypothetical test called p-value (probability value) is exploited. It specifies that whether there is a meaningful of sunspots, etc.) to normal distributions. relation between time series or computed This transformation considers a range for correlation has been occurred by accident. exponents (λ) defined in following equa- The p-value smaller than 0.05 shows the tion higher validity of the correlation (Everitt, Velocity frequency Distribution of orientations Running title • May 2016 • Vol. XXI, No. 1 Maximum of the solar cycle Towards the solar minimum 2011 2012 2013 2014 2015 Time (daily) Figure 5: Time series of daily (blue) and monthly smoothed (red) number of cells during the years 2011 to 2015. ◦ ◦ 2002). range of 0 − 45 with mostly non-roundish shapes (Fig. 7). In Fig. 8, the direction As seen in Table 1, the correlations of cells (orientations) and surrounded el- between the cells size and orientations, lipses are displayed by black lines and red sunspots number and velocities, and also, ellipses, respectively. sunspots number and cells number are about 0.3, 0.2, and 0.3, respectively. There is an anti-correlated behavior be- V. Conclusions tween eccentricities and cells size, and also, sunspots number and cells size with about We used the ball-tracking method, edge - 0.1 and -0.4, respectively. sharpening technique, and box-counting al- gorithm to study the morphological param- As shown in Fig. 7 (upper panel), for eters of photospheric supergranular cells. eccentricities smaller than 0.2, the sizes The code identified the number of 53651 are more fluctuated. As we see in Fig. 7 individual cells from 900 data cubes includ- (lower panel, blue line), by increasing the ing SDO/HMI continuum images during size, the orientation rises. The small cells the years 2011 to 2015. To avoid the projec- with eccentricities around 0.55 have ori- tion effect (sphere to plan projection), B entations close to 0 , and with increasing 0 angle evolution effect, variation of the so- eccentricity, the orientation rises and ap- lar radius during time, a box with the size proaches to 0.8 (Fig. 7, red line in lower ′′ ′′ of 350 × 350 around the central equa- panel). The size (S ) and the orientation 9.509±0.865 torial region (Fig. 1, red box) were stud- (O) are related by |O| ∝ S . Re- ied. Since the fractal dimensions fluctu- lationships between the size of cells, orien- ation is around 1.8 in each frame, it im- tation, and eccentricity show that the large plies that the supergranulation pattern oc- cells are commonly included orientations ◦ ◦ cur with self-similarities. around 45 − 90 with shapes similar to ellipse, and the smaller ones are in the Size-frequency distribution of super- Number of cells Running title • May 2016 • Vol. XXI, No. 1 2500 200 Velocity Sunspot number 2000 150 1500 100 1000 50 500 0 2011 2012 2013 2014 2015 Time (daily) Figure 6: Time series of both smoothed monthly number of sunspots (red) and surface velocities (blue). Table 1: Pearson correlations between statistical parameters. λ Correlation P-value Orientations and cells size -0.700 0.296 0.001 Eccentricities and cells size 5.000 -0.121 0.001 Sunspots number and velocities 2.400 0.244 0.001 Sunspots number and cells size 0.300 -0.410 0.001 Sunspots number and cells number 0.300 0.320 0.001 granules follows the log-normal function law relation between size (S ) and orienta- similar to that of obtained for gran- tion (O) of the cells can be expressed by 9.5 ules (Berrilli et al., 2002; Javaherian et al., |O| ∝ S . The results show that most 2014). The eccentricity distributions of of the small cells have small values of ori- the cells possibly did not undergo the entations (Fig. 7), and their average ec- changes affected by high-activity or low- centricities are larger than 0.55 supporting activity years. The power exponent of ori- non-roundish shapes. We concluded that entation distributions and the parameters the small cells align with the solar rota- of velocity-distributions don’t vary signif- tional velocity, while the larger ones are icantly during the years 2011 to 2015. The mostly orientated towards to the rotational orientation distribution of cells indicates axis. It seems that the solar rotation has the power-law behavior. Cells with smaller not enough force to rotate the large cells sizes (< 600 Mm ) have small angle of the along its rotational velocity. More quanti- ◦ ◦ orientations from 0 to 5 on the solar sur- tative studies are needed to investigate the face (Fig. 7, lower panel). The cells ec- influence of both differential rotation and centricities fluctuate approximately around photospheric magnetic field on cells orien- 0.66 (Fig. 7, upper panel). The power- tations. Velocity (m/s) Sunspot number Running title • May 2016 • Vol. XXI, No. 1 0 0.2 0.4 0.6 0.8 Eccentricity 0.7 Figure 8: The fitted ellipses to the cells and their major axis. The orientations of the cells − the an- 0.65 gle between the major axis of the ellipse and the Sun-X (west-east direction) are ranged be- ◦ ◦ tween -90 to +90 . Power-law fit power-law index 600 Power-law fit 0.6 α ≈ 9.51 ± 0.86 Size pixels labeled one. So, we can introduce Eccentricity 500 the centroid of object by following equa- -100 -100 -50 -50 0 0 50 50 100 100 tions Orientation (degree) 1 1 x = Σ Σ x f (x, y), y = Σ Σ y f (x, y), Figure 7: Relationship between the size of cells (Mm ) CI x y CI x y A A and their eccentricity (upper panel). The mean (4) ˘ ˘ values for each bin (0 âAS¸ 0.02, 0.02 âAS¸ 0.04, where x and y are the coordinates of CI CI etc.) for cells have been displayed. The rela- the intensity-weighted centroid. For an im- tionship between the size of cells (Mm ) and the orientation (degree) as mean values for age, the central moments µ is expressed pq ˘ ˘ each bin (0 âAS¸ 5, 5 âAS¸ 10, etc.) shows that as (Emre Celebi & Alp Aslandogan , the smaller sizes have no preference direction 2005) on the Sun (lower panel, blue line). The rela- tionship between the eccentricity of cells and p q µ = Σ Σ (x− x ) (y− y ) f (x, y). their orientation (degree) for mean values of pq x y CI CI each bin as mentioned are shown (lower panel, (5) red line). The angle φ between the major axis of an object and horizontal axis (positive x-axis) is given as (e.g., Stojmenovic & Nayak , Appendix A: Angle measurements and 2007) error analysis 1 2µ φ = arctan . For a binary image f (x, y), the area of an 2 µ − µ 20 02 object A = Σ Σ f (x, y) is a summation of (6) x y Size (Mm ) Size (Mm ) Eccentricity Running title • May 2016 • Vol. XXI, No. 1 The error of the angle φ as a function of error of orientation angle (Fig. 9). Us- the moments can be obtained by the error ing above-mentioned equations, the angle propagation method (e.g., Hughes & Hase, φ and the error value are obtained to be 2010; Mumford, 2017) defined as follows 31.6231 ± 0.0032 degrees. ! ! 2 2 ∂µ ∂µ pq pq Δµ (x, y) = ± Δx + Δy , pq ∂x ∂y (7) where Δx and Δy are spatial resolution in x and y direction, respectively. Since in our analysis, Δx = Δy, and are equal to one (pixel), the error of moments of inter- est takes the following form 10 20 30 40 50 Δµ (x, y) = ± 11 Figure 9: An artificial binary image with 54 × 40 pix- els including an object described by a func- 2 2 2 2 (Σ Y f (x, y)) + (Σ X f (x, y)), Y X tion f (x, y). The red point is representative of the intensity-weighted centroid. The angle be- 2 2 tween the major axis of the object and the east- Δµ (x, y) = ± Σ 4X f (x, y), 20 X west direction is 31.6231 ± 0.0032 degrees. 2 2 Δµ (x, y) = ± Σ 4Y f (x, y), 02 Y where X = (x − x ) and Y = (y − y ). CI CI So, the error of angle is obtained as follows Δφ(µ , µ , µ ) = ± 11 20 02 ! ! ! 2 2 2 ∂φ ∂φ ∂φ Δµ + Δµ + Δµ , 11 20 02 ∂µ ∂µ ∂µ 11 20 02 where the partial derivative are expanded as ∂φ µ − µ 20 02 = , ∂µ 11 (µ − µ ) + 4µ 20 02 ∂φ −µ = , ∂µ (µ − µ ) + 4µ 20 02 ∂φ µ = . ∂µ 02 (µ − µ ) + 4µ 20 02 As an example, we created 2-D binary form of artificial image (mimicking data: Javaherian et al., 2014) to test the validity of computing moments and estimate the y Running title • May 2016 • Vol. XXI, No. 1 References Fisher, G. H., & Welsch, B. T. 2008, ASP Conference Series, 383, 373 Alipour, N., & Safari, H. 2015, ApJ, 807, Gonzalez, R. C., Woods, R. E. 2008, Dig- ital Image Processing, Pearson Prentice Arish, S., Javaherian, S., Safari, H. & Hall, Upper Saddle River, New Jersey Amiri, A. 2016, Solar Phys, 291, 1209 07458, 631 - 633, 666 Honarbakhsh, L., Alipour, N., & Safari, H. Aschwanden, M. J. 2010, Sol. Phys., 262, 235 2016, Sol. Phys., 291, 941 Hughes, I. G. & Hase, T. P. A. 2010, Mea- Aschwanden, M. J. 2011, Self-Organized suremants and Their Uncertainities, First in Astrophysics, Springer Heidelberg edition, Oxford University Press, New Dordrecht, London, UK, 257 York, 37-44 Aschwanden, M. J. 2015, ApJ, 814, 19 Javaherian, M., Safari, H., Amiri, A., & Zi- Bazarghan, M., Safari, H., Innes, D. E., aei, S. 2014, Sol. Phys., 289, 3969 Karami, E., & Solanki, S. K. 2008, A&A, Javaherian, M., Safari, H., Dadashi, N., & 492, L13 Aschwanden, M. J. 2017, Sol. Phys., 292, Berrilli, F., Consolini, G., Pietropaolo, E., Caccin, B., Penza, V., & Lepreti, F. 2002, Innes, D. E., Genetelli, A., Attie, R., & A&A, 381, 253 Potts, H. E. 2009, Sol. Phys., 495, 3191 Boomgaard, R. V. D., & Balen, R. V. 1992, Meunier, N., Roudier, T., & Tkaczuk, R. Computer Vision, Graphics, and Image 2007c, A&A, 466, 1123 Processing: Graphical Models and Im- Meunier, M., Roudier, T., & Rieutord, M. age Processing, 54(3), 254 2008, A&A, 488, 1109 Del Moro, D., Giordano, S., & Berrilli, F. Molteno, T. C. A. 1993, Physical Review 2007, A&A, 472, 599 E, 48, R3263 Emre Celebi, M., & Alp Aslandogan, Y. Mumford, J. R. 2017, Error Propaga- 2005, Conference: Coding and Comput- tion Reference, John Hopkins University, ing (ITCC’05) International Conference Physics and Astronomy, Basic Physics on Information Technology. December, Laboratory, 2-4 Las Vegas, NV, USA. Newman, M. E. J. 2006, Contemporary Everitt, B. S. 2002, The Cambridge Dic- Physics E, 46, 323 tionary of Statistics, Second edition, Papadimitriou, S., Sun, J., & Philip, S. Cambridge University Press, New York, 2006, Conference: Proceedings of the USA, 53, 304 6th IEEE International Conference on Farhang, N., Safari, H., & Wheatland, M. Data Mining. December, Hong Kong, 2018, accepted for publication in ApJ China 12 Running title • May 2016 • Vol. XXI, No. 1 Potts, H. E., Barrett, R., & Diver, D. A. Tajfirouze, E., & Safari, H. 2012, ApJ, 744, 2004, A&A, 424, 253 113 Potts, H. E., & Diver, D. A. 2008, Sol. Tian, H., Potts, H. E., Marsch, E., Attie, R., Phys., 248, 263 & He, J. 2010, A&A, 519, 10 Press, W. H., Teukolsky, S. A., Vetterling, Title, A. M., Tarbell, T. D., Topka, K. P., W. T., & Flannery, B. P. 2007, Numerical et al. 1989, ApJ, 336, 475 Recieps: The Art of Scientific Comput- Wachter, R., Schou, J., Rabello-Soares, ing, Third edition, Cambridge university M. C., Miles, J. W., Duvall Jr. T. L., & Press, New York, USA, 724 Bush, R. I. 2012, Sol. Phys., 275, 261 Priest, E., 2014, Magnetohydrodynamics Yousefzadeh, M., Safari, H., Attie, R., & of the Sun, First edition, Cambridge Uni- Alipour, N. 2016, Sol. Phys., 291, 29 versity of the Sun, New York, USA, 21, Rincon, F., Roudier, T., Schekochihin, A. A., & Rieutord, M. 2017, A&A, 599, A69 Roudier, T., Rieutord, M., Prat, V., et al. 2013, A&A, 552, A113 Roudier, T., Svanda, M., Rieutord, M., et al. 2014, A&A, 567, A138 Ryutova, M., Tarbell, T. D., Shine, R. 2003, Sol. Phys., 213, 231 Ryutova, M. 2015, Physics of Magnetic Flux Tubes, Springer-Verlag GmbH Berlin Heidelberg, 12 Scherrer, P. H., Bogart, R. S., Bush, R. I., et al. 1995, Sol. Phys., 162, 129 Schou, J., Scherrer, P. H., Bush, R. T., et al. 2012, Sol. Phys., 275, 229 Simon, G. W., & Weiss, N. O. 1968, Zeitschrift fÃijr Astrophysik, 69, 435 Stangalini, M. 2013, A&A, 561, L6 Stojmenovic, M., Nayak, A., 2007, Direct Ellipse Fitting and Measuring Based on Shape Boundaries, Springer, Berlin, Hei- delberg, 221-223 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Astrophysics arXiv (Cornell University)

Statistics of Photospheric Supergranular Cells Observed by SDO/HMI

Download PDF
13 pages

Loading next page...
 
/lp/arxiv-cornell-university/statistics-of-photospheric-supergranular-cells-observed-by-sdo-hmi-VUQVFB5Huf

References (53)

ISSN
0273-1177
eISSN
ARCH-3330
DOI
10.1016/j.asr.2019.04.027
Publisher site
See Article on Publisher Site

Abstract

Aims: The statistics of the photospheric granulation pattern are investigated using continuum images observed by Solar Dynamic Observatory (SDO)/Helioseismic and Magnetic Imager (HMI) taken at 6713 Å. Methods: The supergranular boundaries can be extracted by tracking photospheric velocity plasma flows. The local ball-tracking method is employed to apply on the HMI data gathered over the years 2011-2015 to estimate the boundaries of the cells. The edge sharpening techniques are exerted on the output of ball- tracking to precisely identify the cells borders. To study the fractal dimensionality (FD) of supergranulation, the box counting method is used. Results: We found that both the size and eccentricity follow the log-normal distributions with peak values about 330 Mm and 0.85, respectively. The five-year mean value of the cells number appeared in half-hour ′′ ′′ sequences is obtained to be about 60 ± 6 within an area of 350 × 350 . The cells orientation distribution presents the power-law behavior. Conclusions: The orientation of supergranular cells (O) and their size (S ) follows a power-law function 9.5 2 as |O| ∝ S . We found that the non-roundish cells with smaller and larger sizes than 600 Mm are aligned and perpendicular with the solar rotational velocity on the photosphere, respectively. The FD analysis shows that the supergranular cells form the self-similar patterns. Sun: photosphere - Sun: granulation - Methods: data analysis - Techniques: image processing I. Introduction and splitting old ones (Javaherian et al., 2014). The size of supergranules covers the extended range of scales, with a typi- The solar granulation is the upper side of cal diameters of 20-70 Mm (Priest, 2014; convective cells produced based on travel- Ryutova , 2015). The horizontal velocity ing hot plasma currents from the solar in- of the plasma flows from the cell centers to- terior (convective zone) to the photosphere ward the edges are estimated to be 0.2 − 0.5 (Priest, 2014). Hot plasma rises up to the −1 km s (Simon & Weiss, 1968; Ryutova , surface and transfers energy. Then, cold 2015). plasma returns to the interior within the dark boundaries. The granulation is a tur- Granulation process is linked to the mag- bulent process done by merging new grains netic flux distributed ubiquitously through- arXiv:1807.07479v1 [astro-ph.SR] 19 Jul 2018 Running title • May 2016 • Vol. XXI, No. 1 out the solar surface. It has been shown used Michelson Doppler Imager (MDI: that the emergence of magnetic flux is re- Scherrer et al., 1995) onboard on Solar lated to the cells where the plasma flows and Heliospheric Observatory (SOHO), to positively diverge (Stangalini, 2013). The show that the size of supergranules are large-scale type of granulation occurs in smaller at the maximum of solar activ- supergranular cells which is one of the ity. The intensity variation of supergranu- characteristics of the quiet Sun (Priest, lar cells from center to boundary is com- 2014). Studying the statistics of the surface parable with that of obtained for granules magneto-convective features lead to better (Del Moro et al., 2007). understanding of their evolution. Meunier et al. (2007) have presented re- Photospheric convective pattern high- sults about cell-size distribution and found lights the evolution of other phenomena, a correlation between horizontal veloc- such as coronal bright points, magnetic ity of plasma and supergranular radius cancelation, nanoflares and network extracted from MDI/SOHO. In recent flares in different layers (Ryutova et al., decades, by increasing received data taken 2003; Tajfirouze & Safari , 2012; from ground-based and space telescopes, Yousefzadeh et al., 2016). Tian et al. the automatic detection methods are inten- (2010) used data in various passbands sively expanded to analyze data and ex- to find out a correlation between the tract statistics with higher accuracy (e.g., horizontal velocities of plasma in both see Aschwanden, 2010; Alipour & Safari, photospheric and chromospheric super- 2015; Arish et al., 2016; Javaherian et al., granulation. In one of the statistical works, 2017), and also, reducing costs of data clas- the average diameter and lifetime of su- sification. So, the numerous methods are pergranular cells were found to be 25 Mm progressed to extract supergranular cells and 1.5 days, respectively (Roudier et al., from intensity continuum images. One of 2014). The relation between the cell size the important approaches, known as local and the magnetic field is unclear; but, in correlation tracking (LCT), is employed recent works done based on local corre- to recognize the boundaries based on flux lation tracking (LCT), the cell size and current in data (Papadimitriou et al., 2006). velocity are linked to the intensity map of The velocity of pixels of interests are com- both supergranular vertical and horizontal puted by capturing the transform motion of flows (Rincon et al., 2017). This indirect pixels in two consecutive frames. A two- relation between the cell size and magnetic dimensional (2-D) flow field can represent flux inside the cells leads to anti-correlated the boundaries of cells as the output of the dependence, so the large cells can be method. Fisher & Welsch (2008) extended emerged where the local magnetic field is the LCT method with application of the weak. fourier transform named fourier local cor- relation tracking (FLCT). The relation between supergranular at- tributes and the solar cycle, specially, the One of the promising algorithms for de- cell size and intensity variation have been termining the supergranular boundaries is studied by Meunier et al. (2008). They ball-tracking method. This method consid- 2 Running title • May 2016 • Vol. XXI, No. 1 ers imaginary balls on the gridded surface recorded by HMI. The HMI provides dif- independently moving with plasma flows ferent level of full-disk images in the Fe I (Potts et al., 2004). According to the in- absorption line at 6173 Å with a resolution tensities, the balls move in directions to of 0.50 ± 0.01 arcsec and cadence of 45 sec- settle in local minima where the bound- onds. Some corrections, such as exposure aries are elongated. Using balls track- time, dark current, flat field, and cosmic- ing, the locations of some coronal small- ray hits, are done in level-1 data. scale features (bright points, mini-coronal For our purpose, we used 30-minute con- mass ejections, etc.) were carried out secutive continuum HMI data with a time by Innes et al. (2009), Yousefzadeh et al. lag of 45 seconds in every two days from (2016), and Honarbakhsh et al. (2016). the year 2011 to 2015. Since the measure- We investigate the supergranules mor- ment of morphological parameters of the phological parameters and velocities dur- supergranules, such as size and orientation, ing a five-year period of the solar activ- and also, velocity on the surface very sensi- ity. So, the statistical parameters of su- tive to the projection effect, variation of the pergranular cells, such as size−frequency solar radial over time, and the B angle evo- distribution, fractal dimension (FD), orien- lution effect (Roudier et al., 2013), the par- ′′ ′′ tation, and their eccentricities are studied. tial area with a size of 350 × 350 at the Moreover, the correlations between quan- solar disk center (a region centered with tified parameters and the solar activity are longitudes ±11 around the central merid- computed. We used data recorded by Solar ian, and latitudes limited in ±11 around Dynamic Observatory (SDO)/ Helioseis- the equator) (Fig. 1, red box) was selected mic and Magnetic Imager (HMI) taken at to focus on photospheric flows. To coalign 6173 Å. the sequential data to a reference one, all The layout of this article is as follows: images are derotated using drot map.pro the description of data sets is given in Sect. available in the SSW/IDL package. II. The brief review of the methods are pre- pared in Sect. III. The results are presented III. Methods in Sect. IV. Concluding remarks are ex- The ball-tracking method, edge sharpening plained in Sect. V. technique, and box-counting algorithm are explained briefly as follows. II. Description of Datasets Ball-tracking One of the applicable Solar Dynamic Observatory (SDO) uti- methods developed for computing velocity lized one of the three instruments named fields is ball-tracking method (Potts et al., Helioseismic and Magnetic Imager (HMI: 2004). The ball-tracking method is ap- Schou et al., 2012) to investigate the pho- plied on continuum HMI images to track tospheric oscillations and magnetic fields velocity fields. Using fast fourier transform (Wachter et al., 2012). So, to study which is a part of the ball-tracking code, the supergranular cells using photospheric photospheric p-mode oscillations are at- continuum images, we employed the tenuated to remove features moving faster −1 high-spatial and temporal resolution data than 7 km s . In the code, the constructed 3 Running title • May 2016 • Vol. XXI, No. 1 (a) (b) Figure 1: The continuum SDO/HMI full-disk image of the Sun recorded on 30 December 2015 (00:00-00:30 UT) taken ′′ ′′ at 6173 Å (left panel). The red rectangle is selected with area of 350 × 350 (region centered with longitudes ◦ ◦ ±11 around the central meridian, and the latitudes limited in ±11 around the equator). The output of the ball-tracking method is shown with green face (right panel). The blue edges are the cells border obtained by the morphological filters (see text). data cube with two spatial and one tempo- oscillations (υ = ω/k) indicates a cut−off ral dimension is converted to a cube in fea- cone lateral surface in k − ω space, we ture space of wave vector (k) and frequency are able to discard components covering −1 (ω) by Fourier transform. As photosheric motions greater than υ = 7 km s cut−off 4 Running title • May 2016 • Vol. XXI, No. 1 (Tian et al., 1989; Roudier et al., 2013). tion. To speed up performing the binary dilation, erosion procedure is applied on Ball-tracking method delineates velocity output image and the lanes were nearly fields in a fraction of time by moving spher- closed (Boomgaard & Balen, 1992). To ical balls as float tracers on a surface. In find perimeter of structures in 2-D images, this method, the granulation pattern is con- the algorithm uses connectivities to specify sidered as a criterion for 3-D tracers on sur- the edges with more width. The output im- face. Small floating balls move on the so- age including boundaries is presented with lar photosphere based on bouncy laws. In solid blue lines in Fig. 1 (right panel). other words, merging and splitting of gran- Box-counting Fractals have geomet- ules reveal bumps on surface wherever the ric repetitive patterns in different scales time-evolving granulation form ripples on wherein the whole of structure can be gen- the surface. These movements are very erated by small parts in non-integer dimen- similar to a ball moving on fluid surface. sions (Aschwanden, 2011). One of the This ball has a given mass and momen- methods that estimates the fractal dimen- tum. If granular cells push the ball, the sion of 2-D data is box-counting (Molteno, sphere will continue traveling based on in- 1993). This method is applied on image coming force. So, the path and direction to breaks it into smaller parts with differ- of ball are estimated as a function of time ent resolutions step-by-step. In different (Potts & Diver, 2008). Both the diameter resolutions, the code considers the boxes of balls and average surface penetration de- consisting of image components. In each pend on the resolution of the image. The ra- step, the size of box (ǫ) and the number of dius of ball is chosen to match the typical boxes (N(ǫ)) is computed. Thus, the image size of granular cells. The smaller the ra- dimension (D) is defined as dius, the cell borders are found with more sensitivity to short wavelength noise. Typi- cal value for the radius is half of center-to- log N(ǫ) D = lim . (1) center granular distance. For high resolu- −1 ǫ→0 log(ǫ ) tion HMI data, the ball diameter is around two pixels. Time range of 30 minutes is IV. Results selected to prepare images as a datacube. Figure 1 (right panel) exhibits the ball- To determine the boundaries, an auto- tracking output extracted from HMI data matic recognition method is applied to displayed as regions with green face. HMI intensity-continuum data recorded at 6173 Å. We focused on rectangular area Edge Detection To sharpen the edges, ′′ ′′ of 350 × 350 to extract statistical prop- the binary format of images underwent the bridge algorithm that fill the blank erties of supergranular cells (Fig. 1, left panel). For our purpose, 40 images (half- space appeared between unclosed bound- hour data) for every two days were gath- aries (Gonzalez et al., 2008). In the next step, to fill remained gaps in boundaries, a ered with a time lag of 45 seconds during the years 2011 to 2015. decomposition algorithm is used to struc- turalize edges (if it exists) by image dila- The log-normal function is fitted to 5 Running title • May 2016 • Vol. XXI, No. 1 the size-frequency distribution of super- The peak value of the eccentricity distri- granules, (Fig. 2, upper panel). The bution of supergranular cells is 0.8. The log-normally distributed function is given skewness takes a positive value and kur- by (Newman, 2006; Bazargan et al., 2008; tosis is 1.8 for overall five years. The Aschwanden, 2015), skewness and kurtosis of eccentricity distri- butions are approximately constant during five years (Fig. 2, lower panel). 1 (ln(x) − µ ) f (x, µ , σ) = exp − , (2) 2 The cell orientation between the major 2σ σx 2π and horizontal axis (west-east latitude on where µ is the mean value, and σ is the solar disk) and an estimation of the the standard deviation. The fit parame- error value of the orientation angles are ters µ and σ for overall five-year size- specified for each region (see Appendix A). frequency are obtained to be 5.208 ± 0.038 The orientation distribution of cells follows and 0.757 ± 0.031, respectively, with the −α a power-law fit as y ∝ x , wherein α peak value of 330 Mm . The variation of is the power exponent. Using the meth- these values is approximately constant dur- ods introduced in Aschwanden (2015) and ing five years (Fig. 2, upper panel). Farhang et al. (2018), the power-law expo- 2011 with μ ≈ 5.21 and σ ≈ 0.76 nents were obtained. In Fig. 3, the fitted 2012 with μ ≈ 5.19 and σ ≈ 0.76 2013 with μ ≈ 5.30 and σ ≈ 0.73 power-law function is shown with the ex- 2014 with μ ≈ 5.28 and σ ≈ 0.74 2015 with μ ≈ 5.19 and σ ≈ 0.77 ponent α = 0.1613 ± 0.0416 which seems 0 1000 2000 3000 4000 5000 6000 7000 to be constant over the period. The errors Size (Mm ) -16 for the orientations of the cells were in the 2011 with kurtosis ≈ 1.79 and skewness ≈ 5× 10 -16 2 2012 with kurtosis ≈ 1.79 and skewness ≈ 5× 10 ◦ ◦ -16 range of 0.0001 to 1.7321 with mean val- 2013 with kurtosis ≈ 1.79 and skewness ≈ 4× 10 -16 2014 with kurtosis ≈ 1.79 and skewness ≈ 6× 10 -16 2015 with kurtosis ≈ 1.79 and skewness ≈ 3× 10 ues of 0.0584 . In Fig. 4, the log-normal function is fit- ted on the photospheric horizontal veloci- 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 Eccentricity ties of plasma for each year (Fig. 4). The overall five-year parameters µ and σ are Figure 2: The histograms of cells size (top) and cells eccentricity (bottom) for each year. The log- equal to 4.634 ± 0.039 and 0.556 ± 0.031, normal function (Eq. 2) is fitted to the size dis- respectively. The time series of daily (blue) tributions. and monthly (red) number of cells for five years are shown in Fig. 5. The mean value Each segmented region can be sur- is equal to 59.6 ± 5.62 cells ranged from rounded by an ellipse that can be at- 43 to 77 in each frame. tained by morphological moments of a shape (Emre Celebi & Alp Aslandogan , As we see in Fig. 6, the time series 2005). Thus, the major and minor axis of of the plasma velocity and the number of contained area is computed to obtain the ec- sunspots are presented. To test Pearson centricity. All cells eccentricities take val- correlation between time series, we em- ues ranged from 0 and 1. The value 0 rep- ployed the box-Cox transformation to con- resents the shape as a circle, and 1 returns vert the distributions of time series with a segmented shape as line. positive values (number of cells, number Number (× 10 ) Number (× 10 ) Running title • May 2016 • Vol. XXI, No. 1 4 4 10 10 (a) power index, α ≈ - 0.188 ± 0.086 (b) Power index, α ≈ - 0.1091 ± -0.0282 3 3 Data point Data point Power-law fit Power-law fit 10 10 4 4 10 10 (c) (d) power index, α ≈ - 0.2096 ± 0.0541 power index, α ≈ - 0.1355 ± 0.0350 3 3 10 10 Data point Data point Power-law fit Power-law fit 2 2 10 10 (e) power index, α ≈ - 0.1357 ± 0.0350 (f) Data point Power-law fit Data point Power-law fit power index, α ≈ - 0.1613 ± 0.0416 2 2 10 10 0 1 2 0 1 2 10 10 10 10 10 10 Orientation (degree) Orientation (degree) Figure 3: The orientation distribution of supergranules for the years of 2011 − 2015 (a: 2011, b: 2012, c: 2013, d: 2014, e: 2015, f: 2011-2015). The minimum and maximum of orientation errors computed for regions are ◦ ◦ ◦ about 0.0001 and 1.7321 , respectively. The maximum error values belong to the cells orientated in ±45 . × 10 3.5 2011 with μ ≈ 4.66 and σ ≈ 0.55 y − 1 2012 with μ ≈ 4.70 and σ ≈ 0.53 3   , if λ , 0, 2013 with μ ≈ 4.66 and σ ≈ 0.54 Y(λ) = 2014 with μ ≈ 4.68 and σ ≈ 0.53 2.5 2015 with μ ≈ 4.65 and σ ≈ 0.54 log(y), if λ = 0, where λ usually varies from -5 to 5. The op- 1.5 timal value of λ is obtained as the best ap- proximation fitted on the normal distribu- tion curve (Everitt, 2002). To do this, first, 0.5 kurtosis for each transformed time series with value about three are chosen for the 0 200 400 600 800 1000 1200 best λ. Then, the correlation coefficients Velocity (m/s) are computed for time series (Press et al., Figure 4: The histograms of horizontal velocities for 2007). each year during the years 2011 to 2015. The To validate the values attained by Pear- log-normal function (Eq. 2) is fitted to the dis- tributions. son correlation, a hypothetical test called p-value (probability value) is exploited. It specifies that whether there is a meaningful of sunspots, etc.) to normal distributions. relation between time series or computed This transformation considers a range for correlation has been occurred by accident. exponents (λ) defined in following equa- The p-value smaller than 0.05 shows the tion higher validity of the correlation (Everitt, Velocity frequency Distribution of orientations Running title • May 2016 • Vol. XXI, No. 1 Maximum of the solar cycle Towards the solar minimum 2011 2012 2013 2014 2015 Time (daily) Figure 5: Time series of daily (blue) and monthly smoothed (red) number of cells during the years 2011 to 2015. ◦ ◦ 2002). range of 0 − 45 with mostly non-roundish shapes (Fig. 7). In Fig. 8, the direction As seen in Table 1, the correlations of cells (orientations) and surrounded el- between the cells size and orientations, lipses are displayed by black lines and red sunspots number and velocities, and also, ellipses, respectively. sunspots number and cells number are about 0.3, 0.2, and 0.3, respectively. There is an anti-correlated behavior be- V. Conclusions tween eccentricities and cells size, and also, sunspots number and cells size with about We used the ball-tracking method, edge - 0.1 and -0.4, respectively. sharpening technique, and box-counting al- gorithm to study the morphological param- As shown in Fig. 7 (upper panel), for eters of photospheric supergranular cells. eccentricities smaller than 0.2, the sizes The code identified the number of 53651 are more fluctuated. As we see in Fig. 7 individual cells from 900 data cubes includ- (lower panel, blue line), by increasing the ing SDO/HMI continuum images during size, the orientation rises. The small cells the years 2011 to 2015. To avoid the projec- with eccentricities around 0.55 have ori- tion effect (sphere to plan projection), B entations close to 0 , and with increasing 0 angle evolution effect, variation of the so- eccentricity, the orientation rises and ap- lar radius during time, a box with the size proaches to 0.8 (Fig. 7, red line in lower ′′ ′′ of 350 × 350 around the central equa- panel). The size (S ) and the orientation 9.509±0.865 torial region (Fig. 1, red box) were stud- (O) are related by |O| ∝ S . Re- ied. Since the fractal dimensions fluctu- lationships between the size of cells, orien- ation is around 1.8 in each frame, it im- tation, and eccentricity show that the large plies that the supergranulation pattern oc- cells are commonly included orientations ◦ ◦ cur with self-similarities. around 45 − 90 with shapes similar to ellipse, and the smaller ones are in the Size-frequency distribution of super- Number of cells Running title • May 2016 • Vol. XXI, No. 1 2500 200 Velocity Sunspot number 2000 150 1500 100 1000 50 500 0 2011 2012 2013 2014 2015 Time (daily) Figure 6: Time series of both smoothed monthly number of sunspots (red) and surface velocities (blue). Table 1: Pearson correlations between statistical parameters. λ Correlation P-value Orientations and cells size -0.700 0.296 0.001 Eccentricities and cells size 5.000 -0.121 0.001 Sunspots number and velocities 2.400 0.244 0.001 Sunspots number and cells size 0.300 -0.410 0.001 Sunspots number and cells number 0.300 0.320 0.001 granules follows the log-normal function law relation between size (S ) and orienta- similar to that of obtained for gran- tion (O) of the cells can be expressed by 9.5 ules (Berrilli et al., 2002; Javaherian et al., |O| ∝ S . The results show that most 2014). The eccentricity distributions of of the small cells have small values of ori- the cells possibly did not undergo the entations (Fig. 7), and their average ec- changes affected by high-activity or low- centricities are larger than 0.55 supporting activity years. The power exponent of ori- non-roundish shapes. We concluded that entation distributions and the parameters the small cells align with the solar rota- of velocity-distributions don’t vary signif- tional velocity, while the larger ones are icantly during the years 2011 to 2015. The mostly orientated towards to the rotational orientation distribution of cells indicates axis. It seems that the solar rotation has the power-law behavior. Cells with smaller not enough force to rotate the large cells sizes (< 600 Mm ) have small angle of the along its rotational velocity. More quanti- ◦ ◦ orientations from 0 to 5 on the solar sur- tative studies are needed to investigate the face (Fig. 7, lower panel). The cells ec- influence of both differential rotation and centricities fluctuate approximately around photospheric magnetic field on cells orien- 0.66 (Fig. 7, upper panel). The power- tations. Velocity (m/s) Sunspot number Running title • May 2016 • Vol. XXI, No. 1 0 0.2 0.4 0.6 0.8 Eccentricity 0.7 Figure 8: The fitted ellipses to the cells and their major axis. The orientations of the cells − the an- 0.65 gle between the major axis of the ellipse and the Sun-X (west-east direction) are ranged be- ◦ ◦ tween -90 to +90 . Power-law fit power-law index 600 Power-law fit 0.6 α ≈ 9.51 ± 0.86 Size pixels labeled one. So, we can introduce Eccentricity 500 the centroid of object by following equa- -100 -100 -50 -50 0 0 50 50 100 100 tions Orientation (degree) 1 1 x = Σ Σ x f (x, y), y = Σ Σ y f (x, y), Figure 7: Relationship between the size of cells (Mm ) CI x y CI x y A A and their eccentricity (upper panel). The mean (4) ˘ ˘ values for each bin (0 âAS¸ 0.02, 0.02 âAS¸ 0.04, where x and y are the coordinates of CI CI etc.) for cells have been displayed. The rela- the intensity-weighted centroid. For an im- tionship between the size of cells (Mm ) and the orientation (degree) as mean values for age, the central moments µ is expressed pq ˘ ˘ each bin (0 âAS¸ 5, 5 âAS¸ 10, etc.) shows that as (Emre Celebi & Alp Aslandogan , the smaller sizes have no preference direction 2005) on the Sun (lower panel, blue line). The rela- tionship between the eccentricity of cells and p q µ = Σ Σ (x− x ) (y− y ) f (x, y). their orientation (degree) for mean values of pq x y CI CI each bin as mentioned are shown (lower panel, (5) red line). The angle φ between the major axis of an object and horizontal axis (positive x-axis) is given as (e.g., Stojmenovic & Nayak , Appendix A: Angle measurements and 2007) error analysis 1 2µ φ = arctan . For a binary image f (x, y), the area of an 2 µ − µ 20 02 object A = Σ Σ f (x, y) is a summation of (6) x y Size (Mm ) Size (Mm ) Eccentricity Running title • May 2016 • Vol. XXI, No. 1 The error of the angle φ as a function of error of orientation angle (Fig. 9). Us- the moments can be obtained by the error ing above-mentioned equations, the angle propagation method (e.g., Hughes & Hase, φ and the error value are obtained to be 2010; Mumford, 2017) defined as follows 31.6231 ± 0.0032 degrees. ! ! 2 2 ∂µ ∂µ pq pq Δµ (x, y) = ± Δx + Δy , pq ∂x ∂y (7) where Δx and Δy are spatial resolution in x and y direction, respectively. Since in our analysis, Δx = Δy, and are equal to one (pixel), the error of moments of inter- est takes the following form 10 20 30 40 50 Δµ (x, y) = ± 11 Figure 9: An artificial binary image with 54 × 40 pix- els including an object described by a func- 2 2 2 2 (Σ Y f (x, y)) + (Σ X f (x, y)), Y X tion f (x, y). The red point is representative of the intensity-weighted centroid. The angle be- 2 2 tween the major axis of the object and the east- Δµ (x, y) = ± Σ 4X f (x, y), 20 X west direction is 31.6231 ± 0.0032 degrees. 2 2 Δµ (x, y) = ± Σ 4Y f (x, y), 02 Y where X = (x − x ) and Y = (y − y ). CI CI So, the error of angle is obtained as follows Δφ(µ , µ , µ ) = ± 11 20 02 ! ! ! 2 2 2 ∂φ ∂φ ∂φ Δµ + Δµ + Δµ , 11 20 02 ∂µ ∂µ ∂µ 11 20 02 where the partial derivative are expanded as ∂φ µ − µ 20 02 = , ∂µ 11 (µ − µ ) + 4µ 20 02 ∂φ −µ = , ∂µ (µ − µ ) + 4µ 20 02 ∂φ µ = . ∂µ 02 (µ − µ ) + 4µ 20 02 As an example, we created 2-D binary form of artificial image (mimicking data: Javaherian et al., 2014) to test the validity of computing moments and estimate the y Running title • May 2016 • Vol. XXI, No. 1 References Fisher, G. H., & Welsch, B. T. 2008, ASP Conference Series, 383, 373 Alipour, N., & Safari, H. 2015, ApJ, 807, Gonzalez, R. C., Woods, R. E. 2008, Dig- ital Image Processing, Pearson Prentice Arish, S., Javaherian, S., Safari, H. & Hall, Upper Saddle River, New Jersey Amiri, A. 2016, Solar Phys, 291, 1209 07458, 631 - 633, 666 Honarbakhsh, L., Alipour, N., & Safari, H. Aschwanden, M. J. 2010, Sol. Phys., 262, 235 2016, Sol. Phys., 291, 941 Hughes, I. G. & Hase, T. P. A. 2010, Mea- Aschwanden, M. J. 2011, Self-Organized suremants and Their Uncertainities, First in Astrophysics, Springer Heidelberg edition, Oxford University Press, New Dordrecht, London, UK, 257 York, 37-44 Aschwanden, M. J. 2015, ApJ, 814, 19 Javaherian, M., Safari, H., Amiri, A., & Zi- Bazarghan, M., Safari, H., Innes, D. E., aei, S. 2014, Sol. Phys., 289, 3969 Karami, E., & Solanki, S. K. 2008, A&A, Javaherian, M., Safari, H., Dadashi, N., & 492, L13 Aschwanden, M. J. 2017, Sol. Phys., 292, Berrilli, F., Consolini, G., Pietropaolo, E., Caccin, B., Penza, V., & Lepreti, F. 2002, Innes, D. E., Genetelli, A., Attie, R., & A&A, 381, 253 Potts, H. E. 2009, Sol. Phys., 495, 3191 Boomgaard, R. V. D., & Balen, R. V. 1992, Meunier, N., Roudier, T., & Tkaczuk, R. Computer Vision, Graphics, and Image 2007c, A&A, 466, 1123 Processing: Graphical Models and Im- Meunier, M., Roudier, T., & Rieutord, M. age Processing, 54(3), 254 2008, A&A, 488, 1109 Del Moro, D., Giordano, S., & Berrilli, F. Molteno, T. C. A. 1993, Physical Review 2007, A&A, 472, 599 E, 48, R3263 Emre Celebi, M., & Alp Aslandogan, Y. Mumford, J. R. 2017, Error Propaga- 2005, Conference: Coding and Comput- tion Reference, John Hopkins University, ing (ITCC’05) International Conference Physics and Astronomy, Basic Physics on Information Technology. December, Laboratory, 2-4 Las Vegas, NV, USA. Newman, M. E. J. 2006, Contemporary Everitt, B. S. 2002, The Cambridge Dic- Physics E, 46, 323 tionary of Statistics, Second edition, Papadimitriou, S., Sun, J., & Philip, S. Cambridge University Press, New York, 2006, Conference: Proceedings of the USA, 53, 304 6th IEEE International Conference on Farhang, N., Safari, H., & Wheatland, M. Data Mining. December, Hong Kong, 2018, accepted for publication in ApJ China 12 Running title • May 2016 • Vol. XXI, No. 1 Potts, H. E., Barrett, R., & Diver, D. A. Tajfirouze, E., & Safari, H. 2012, ApJ, 744, 2004, A&A, 424, 253 113 Potts, H. E., & Diver, D. A. 2008, Sol. Tian, H., Potts, H. E., Marsch, E., Attie, R., Phys., 248, 263 & He, J. 2010, A&A, 519, 10 Press, W. H., Teukolsky, S. A., Vetterling, Title, A. M., Tarbell, T. D., Topka, K. P., W. T., & Flannery, B. P. 2007, Numerical et al. 1989, ApJ, 336, 475 Recieps: The Art of Scientific Comput- Wachter, R., Schou, J., Rabello-Soares, ing, Third edition, Cambridge university M. C., Miles, J. W., Duvall Jr. T. L., & Press, New York, USA, 724 Bush, R. I. 2012, Sol. Phys., 275, 261 Priest, E., 2014, Magnetohydrodynamics Yousefzadeh, M., Safari, H., Attie, R., & of the Sun, First edition, Cambridge Uni- Alipour, N. 2016, Sol. Phys., 291, 29 versity of the Sun, New York, USA, 21, Rincon, F., Roudier, T., Schekochihin, A. A., & Rieutord, M. 2017, A&A, 599, A69 Roudier, T., Rieutord, M., Prat, V., et al. 2013, A&A, 552, A113 Roudier, T., Svanda, M., Rieutord, M., et al. 2014, A&A, 567, A138 Ryutova, M., Tarbell, T. D., Shine, R. 2003, Sol. Phys., 213, 231 Ryutova, M. 2015, Physics of Magnetic Flux Tubes, Springer-Verlag GmbH Berlin Heidelberg, 12 Scherrer, P. H., Bogart, R. S., Bush, R. I., et al. 1995, Sol. Phys., 162, 129 Schou, J., Scherrer, P. H., Bush, R. T., et al. 2012, Sol. Phys., 275, 229 Simon, G. W., & Weiss, N. O. 1968, Zeitschrift fÃijr Astrophysik, 69, 435 Stangalini, M. 2013, A&A, 561, L6 Stojmenovic, M., Nayak, A., 2007, Direct Ellipse Fitting and Measuring Based on Shape Boundaries, Springer, Berlin, Hei- delberg, 221-223

Journal

Astrophysics – arXiv (Cornell University)

Published: Jul 19, 2018

There are no references for this article.