Chromospheric Plasma Ejections in a Light Bridge of a Sunspot 1 1 3 2 2 Donguk Song , Jongchul Chae , Vasyl Yurchyshyn , Eun-Kyung Lim , Kyung-Suk Cho , 1 1 1 Heesu Yang , Kyuhyoun Cho , and Hannah Kwak 7 1 0 2 [email protected] n a J Received ; accepted 4 2 ] R S . h p - o r t s a [ 1 v 8 0 8 6 0 . 1 0 7 1 : v i X r a 1 Astronomy Program, Department of Physics and Astronomy, Seoul National University, 1 Gwanak-ro, Gwanak-gu, Seoul 08826, Korea 2 Korea Astronomy and Space Science Institute 776, Daedeokdae-ro, Yuseong-gu, Daejeon 34055, Korea 3 Big Bear Solar Observatory, New Jersey Institute of Technology, 40386 North Shore Lane, Big Bear City, CA 92314-9672, USA – 2 – ABSTRACT Itiswell-known thatlightbridgesinsideasunspot producesmall-scale plasma ejections and transient brightenings in the chromosphere, but the nature and origin of such phenomena are still unclear. Utilizing the high-spatial and high- temporal resolution spectral data taken with the Fast Imaging Solar Spectro- graph and the TiO 7057 ˚A broadband filter images installed at the 1.6 meter New Solar Telescope of Big Bear Solar Observatory, we report arcsecond-scale ′′ chromospheric plasma ejections (1. 7) inside a light bridge. Interestingly, the ejections are found to be a manifestation of upwardly propagating shock waves as evidenced by the sawtooth patterns seen in the temporal-spectral plots of the Caii 8542 ˚A and Hα intensities. We also found a fine-scale photospheric pattern ′′ −1 (1 ) diverging with a speed of about 2 km s two minutes before the plasma ejections, which seems to be a manifestation of magnetic flux emergence. As a response to the plasma ejections, the corona displayed small-scale transient brightenings. Based on our findings, we suggest that the shock waves can be excited by the local disturbance caused by magnetic reconnection between the emerging flux inside the light bridge andthe adjacent umbral magnetic field. The disturbance generates slow-mode waves, which soon develop into shock waves, and manifest themselves as the arcsecond-scale plasma ejections. It also appears that the dissipation of mechanical energy in the shock waves can heat the local corona. Subject headings: shock waves – sunspots – Sun: corona – Sun: chromosphere – Sun: photosphere – 3 – 1. INTRODUCTION Light bridges (LBs) are filamentary bright structures dividing a sunspot umbra into two umbral regions with the same magnetic polarity (Muller 1979). They comprise small-scale successive convective cells and an elongated dark lane running parallel to the central axis. A LB has a special status in that convective motions are not fully inhibited by magnetic fields (Hirzberger et al. 2002; Rimmele 2008). Plasma below it can penetrate into the vertical magnetic fields of umbra (Schu¨ssler & V¨ogler 2006), and the magnetic fields above it form a magnetic canopy structure (Jurˇc´ak et al. 2006). Such a complex topology of the magnetic fields and magnetoconvection in the LB are believed to be responsible for a variety of dynamic phenomena in the chromosphere, but the nature and origin of the chromospheric activities in the LBs still remain unresolved. The most well-known chromospheric activities in the LB are plasma ejections seen as jet-like features in the Hα (Roy 1973; Asai et al. 2001) and Caii H (Shimizu et al. 2009; Shimizu 2011; Louis et al. 2014) images. The plasma ejections intermittently recur with a short lifetime more than one day (Shimizu et al. 2009), and their length are usually shorter than 1000 km (Louis et al. 2014). The plasma ejections are generally observed at one side of the LB, and the ejected plasma moves along the magnetic field lines above the LB (Asai et al. 2001; Louis et al. 2014). Meanwhile, Asai et al. (2001) reported loop-like bright structures in the TRACE 171 ˚A image that traced the edge of plasma ejections seen in the Hα image. They suggested that the loop-like structures indicate the emergence of the bipolar magnetic flux in the LB. Liu (2012) found a coronal jet ejected from a LB in the EUV 171 ˚A image. They inferred that the coronal jet originated from magnetic reconnection between the LB and the sunspot umbra. Magnetic reconnection in the low chromosphere and the upper photosphere of a LB is thought to be responsible for the chromospheric plasma ejections above the LB. – 4 – Bharti et al. (2007) reported an opposite magnetic polarity inside a LB and suggested that the low-altitude magnetic reconnection above the LB may help dense plasma to be injected into the chromosphere. Shimizu et al. (2009) found a strong electric current sheet formed along a LB. They proposed that highly twisted flux tubes trapped below the magnetic canopy are favorable for the magnetic reconnections inside the LB. Meanwhile, Louis et al. (2015) found a small-scale, flat Ω-loop emerging from a granular LB that was associated with the chromospheric emission detected at one side of the Ω-loop. High-resolution observations of LBs from the Hinode (Tsuneta et al. 2008) and Interface Region Imaging Spectrograph (IRIS; De Pontieu et al. 2014) revealed enhanced brightenings at the edge of jets seen in a LB and coordinated behaviors with the adjacent jets (Bharti 2015). Bharti (2015) speculated that waves leaking from the LB produce these kinds of coordinated behaviors between jets above the LB and adjacent jets. The excitation of waves in the low chromosphere above the LB can be regarded as one of the possible scenarios for the generation of plasma ejections. This has been reported by the previous studies of dynamic jet-like features seen at the region outside sunspots that slow-mode shock waves excited in the low chromosphere can play an important role for the evolution of dynamic jet-like features, such as dynamic fibrils (Hansteen et al. 2006; De Pontieu et al. 2007; Langangen et al. 2008), superpenumbral fibrils (Chae et al. 2014, 2015), and surges/jets (Morton 2012; Yang et al. 2014). In this regard, we expect that shock waves genereated in the low chromosphere of a LB may appear as fine-scale chromospheric plasma ejections in the LB, but there have been no clear observational reports on the relation between shock wave and chromospheric plasma ejections above a LB until now. Here, we report that upwardly propagatingshock waves excited inthelow chromosphere appear as arcsecond-scale chromospheric plasma ejections above a LB. For this study, we combine the high resolution spectral data taken with the Fast Imaging Solar Spectrograph – 5 – (FISS; Chae et al. 2013a) installed at the 1.6 meter New Solar Telescope (NST; Goode et al. 2010) of Big Bear Solar Observatory (BBSO) and the extreme-ultraviolet (EUV) data taken by the Atmospheric Image Assembly (AIA; Lemen et al. 2012) on board the Solar Dynamics Observatory (SDO; Pesnell et al. 2012). We investigate in detail the spectral characteristics of the chromospheric plasma ejections detected above the LB and an unique photospheric flow pattern associated with the chromospheric plasma ejections. Our paper is organized as follows. In Section 2, we describe observations and data analysis of the LB. Our observational findings are presented in Section 3. In Section 3.1, shock wave phenomena of the chromospheric plasma ejections are described. The description of the photospheric flow pattern associated plasma ejections is also presented in Section 3.2. In Section 3.3, the response of the corona to the plasma ejections is described. Finally, we summarize and discuss the physical implications of the chromospheric plasma ejections above the LB in Section 4. 2. OBSERVATIONS AND DATA ANALYSIS On 2014 June 6, we observed a leading sunspot (Figure 1) in NOAA Active Region 12082 (N17, E25) with the FISS and the TiO 7057 ˚A broadband filter (Cao et al. 2010) installed at the 1.6 meter NST of BBSO. The FISS is a high dispersion Ech´elle spectrograph that can simultaneously record the Hα and Caii 8542 ˚A lines based on the fast scanning mode. The spectrograms of the Hα and Caii lines respectively cover the wavelength ranges from -5 ˚A to 5 ˚A and from -8 ˚A to 5 ˚A with the spectral dispersion being 0.019 ˚A and 0.026 ˚A. The slit size is 0.′′16 wide and 40′′ long. One scan of the raster image comprises ′′× ′′ 160 steps and covers a 25 40 field of view (FOV). The exposure time was 30 ms. It took about 23 s for completion of one scan. The basic data processing including flat-fielding, distortion correction, compression, and noise reduction was done in the way described – 6 – by Chae et al. (2013a). The line-of-sight (LOS) Dopplergrams of the photosphere and the chromosphere were computed from the lambdameter method (Deubner et al. 1996) at multi-wavelength levels of the Hα and Sii 8536 ˚A lines. Figure 1 shows the photospheric image of the sunspot that we observed at 20:52:18 UT on June 6, 2014. It was taken by the TiO 7057 ˚A broadband filter with the aid of 308 sub-aperture adaptive optics (Shumko et al. 2014). The speckle reconstruction of the TiO image was done by using the Kiepenheuer-Institut Speckle Interferometry Package code reported by W¨oger et al. (2008). We obtained the photospheric images every 15 s from 20:04:50 UT to 21:04:32 UT. The sunspot was rapidly growing during this observational time. It had the varied shapes of LBs dividing a sunspot umbra into several umbral regions. ′′× ′′ A white rectangle in Figure 1 shows the region of our interest, covering 5 5 . The region includes a strong LB embedded in a dark umbra. The LB has negative-polarity magnetic fields with the field strength below 1200 G. The surrounding umbra has magnetic field of the same polarity stronger than 1400 G. From the TiO image, we find a dark lane running parallel to the central axis of the LB and a number of small-scale successive convective cells. The coronal images were acquired with the AIA on board the SDO. The SDO/AIA provides high resolution full-disk data of the Sun in 7 extreme-ultraviolet (EUV; λ94 ˚A, 131 ˚A, 171 ˚A, 193 ˚A, 211 ˚A, 304 ˚A, 335 ˚A) and 3 ultraviolet (UV; λ1600 ˚A, 1700 ˚A, and 4500 ˚A) channels with the pixel scale of 0.′′6 at a time cadence of 12 s (24 s in UV channels). To investigate temperature variations of EUV-emitting plasma in the corona, we determined the filter ratios of the 171 ˚A, 193 ˚A, and 211 ˚A passbands, which have been often used in a determination factor of temperatures in the corona (Chae et al. 2002). In particular, the variations of the filter ratios (193 ˚A/171 ˚A and 211 ˚A/171 ˚A) are very sensitive to the − temperature variations of 1 2 MK. The FISS data were aligned with the SDO/Helioseismic and Magnetic Imager (HMI; – 7 – Schou et al. 2012) data by using cross-correlation between the SDO/HMI intensity image and the FISS/Caii -4 ˚A image, and then the co-alignment between the SDO/HMI and the SDO/AIA data were carried out by using the IDL (Interactive Data Language) programs of the hmi prep.pro and the aia prep.pro. 3. Results 3.1. Chromospheric Plasma Ejections A chromospheric plasma ejection in an arcsecond-scale is visible above a LB in the Caii -0.5 ˚A and Hα -0.7 ˚A images taken by the FISS (Figure 2). The maximum length ′′ ′′ of the plasma ejection is about 1250 km (1. 7), and the width is about 290 km (0. 4). In the Caii -0.5 ˚A image, the chromospheric plasma ejection divides into two distinct parts; a bright patch and a dark patch. On the other hand, the bright patch is not conspicuous in the Hα -0.7 ˚A image. This may be attributed to the difference in temperature sensitivity between the Caii line and the Hα line (Cauzzi et al. 2009). Caii intensity is more sensitive to the temperature of chromospheric structures than Hα intensity, while Hα intensity is more subject to the light-scattering of the chromospheric features (Chae et al. 2013b). The bright patch and the dark patch of the plasma ejection have different spectral ii characteristics in both the lines (Figure 3). First, the Ca spectral profile of the bright patch shows strong emission in the blue and red wings at the wavelength ranges [-0.4 ˚A and -0.3 ˚A] and [+0.3 ˚A and +0.4 ˚A], respectively. Meanwhile, we find a weak absorption ii ii feature in the Ca line core. This shape of the Ca spectral profile – a broad Gaussian ii ii emission at the Ca line wings and a central absorption at the Ca line core – corresponds ii to the shape of the Ca spectral profile at a transient brightening above a LB reported ii by Louis et al. (2015). In addition, it is similar to the well-known Ca spectral profiles of – 8 – the Ellerman bombs (Ellerman 1917) and the penumbral microjets (Katsukawa et al. 2007; Vissers et al. 2015) located outside sunspot umbrae. ii We also find from the Ca spectral profile that the blue wing has stronger emission ii than the red wing. This asymmetric emission is more pronounced in the Ca contrast profile. The asymmetric wing emission generally reflects mass motion at the event formation height, and so we infer that upward plasma motion was predominant in the low chromosphere. In the Hα spectral profile, however, such a strong emission feature is not ii noticeable. This is different from the case of an Ellerman bomb. Second, the Ca spectral profile of the dark patch in the plasma ejections, presents several absorption cores (cyan arrows in Figure 3). These absorption cores are especially well-identified in the contrast profiles of the Caii and Hα lines. Our data suggest that the dark patch of the plasma ejection is composed of plasmas that have several velocities. Figure 4 shows the temporal variations of the chromospheric plasma ejections seen in the TiO, Caii and Hα images from 20:48:04 UT to 20:58:18 UT. We find from the − figure that the plasma ejections in the LB occurred twice within 10 minutes. Rows 4 6 − in Figure 4 present the first plasma ejection, and rows 7 10 present the second plasma ejection. Each ejection lasts about 3 minutes. Moreover, we find that the dark patch in the plasma ejections are seen successively at the blue wings, the cores, and the red wings (see, red and cyan arrows in Figure 4), which shows that the plasma motion switched from upward to downward directions. Figures 5 and Figure 6 present the spatio-temporal patterns of the plasma ejections in the wavelength-time plots (λ−t plots) and the time-distance plots (t−d plots) of the spectral data. The observed patterns are quite like what are expected in shock waves propagating upwards in several aspects. First, the λ−t plots indicate the temporal pattern of downflow, emission, and upflow. This kind of pattern was repeated at least three times – 9 – during our observationas, which constitute a periodic pattern of about three-minute period. In particular, the two recurrent plasma ejections correspond to the upflow phases in two strongest patterns. Second, each pattern includes a velocity jump from fast downflow (large redshift) to fast upflow (large blueshift). It especially appears as a sawtooth or a N-shape pattern in the λ−t plots. Third, the velocity jump is accompanied by an emission in the line cores. The emission core reflects the increase of source function in the line, and may be considered as an evidence for strong compression and heating. Finally, the position of each −1 emission moves in the plane of sky with a speed of 10.5 km s or higher. In particular, the propagation of each emission is observed at the boundary between the large redshift and the large blueshift, which is regarded as shock frount (see, Figure 6). Figure 5 confirms that each plasma ejection starts with a sudden appearance of upward motion, deceleration, gradual switch to downward motion, and then downward acceleration. The peak speed of upflow is estimated at about 26 km s−1 in the Hα data and 20 km s−1 ii in the Ca data, which turns into the downflow of the same speed in about three minutes. −2 −2 Therefore the resulting deceleration is about 290 m s and 220 m s , respectively. These −2 values are close to solar gravitational acceleration 274 m s , but the proximity may be a coincidence, because the measured deceleration is determined by the peak speed for the given period of three minutes. The positive correlation between peak speed and deceleration was previously reported in dynamic fibrils and was regarded as evidence for shock waves based on the comparison with numerical simulations (Hansteen et al. 2006; De Pontieu et al. 2007). 3.2. Pattern of Photospheric Diverging Motion ′′ We find an unique photospheric pattern of diverging motion (1 ) in the surface of the LB before the occurrence of the chromospheric plasma ejections. This is well identified in – 10 – the time sequence of TiO images from 20:50:18 UT (t=2.2 min) to 20:52:18 UT (t=4.2 min) shown in Figure 7. In the figure, the photoshperic pattern first appeared near the dark lane of the LB, and then expanded across the LB. The expanding direction of the pattern is almost the same as the direction of the ejected plasma in the chromosphere. Two small bright points with the diameter of 90 km are visible at the terminations of the photospheric pattern. They are comparable to quiet Sun bright points in the intergranular lanes in size. The chromospheric plasma ejections of our interest occurred at the one side of the photospheric pattern when the edge of the pattern reached the boundary of the sunspot umbra (see, Figure 2). Our data suggest that the chromospheric plasma ejections are closely related with the diverging motion in the photosphere. The photospheric pattern of diverging motion mentioned above is well identified in the time-distance plot (t−d plot) of the TiO intensity in Figure 8. The tilted bright strands indicate the trajectories of two bright points located at the termination positions of the −1 diverging pattern. Its horizontal speed is about 2 km s , which is faster than the typical speed of horizontal flows in a LB (Louis et al. 2014), but is slower than the speed of the expanding granular cells seen in emerging flux regions outside sunspots (Yang et al. 2013; Kim et al. 2015). In the t−d maps of the Caii ±0.5 ˚A and core intensities of Figure 8, we can see successive enhanced brightenings next to the one side of the photospheric pattern. These intense brightenings are the lower part of plasma ejections. We find from Figure 8 that the brightenings in the Caii -0.5 ˚A image are more intense than that of the Caii +0.5 ˚A image. Since the intense brightenings in the line wing images present the plasma motion and temperature in the low chromosphere, we suggest that the plasma in the low chromosphere mostly have upward motion. We also find that the one side of the flow pattern corresponds to the strong redshift elongated feature in the plot of the photospheric i Doppler velocity determined from the lambdameter method in Si line (Figure 8). The redshift feature indicates downflows in the photosphere. The peak speed of the downflows