Signatures of running penumbral waves in sunspot photospheres
AAstronomy & Astrophysics manuscript no. AA˙RPW˙final˙arxiv c (cid:13)
ESO 2018August 26, 2018
Signatures of running penumbral waves in sunspot photospheres
J. L¨ohner-B¨ottcher and N. Bello Gonz´alez
Kiepenheuer-Institut f¨ur Sonnenphysik, Sch¨oneckstr. 6, 79104 Freiburge-mail: jlb, nbello @kis.uni-freiburg.de
Received 31 March 2015 / Accepted 19 June 2015
ABSTRACT
Context.
The highly dynamic atmosphere above sunspots exhibits a wealth of magnetohydrodynamic (MHD) waves. Recent studiessuggest a coupled nature of the most prominent phenomena: umbral flashes and running penumbral waves (RPWs).
Aims.
From an observational point of view, we perform a height-dependent study of RPWs, compare their wave characteristics, andaim to track down these so far only chromospherically observed phenomena to photospheric layers to prove the upward propagatingfield-guided nature of RPWs.
Methods.
We analyze a time series (58 min) of multiwavelength observations of an isolated circular sunspot (NOAA11823) taken athigh spatial and temporal resolution in spectroscopic mode with the Interferometric BIdimensional Spectro-polarimeter (IBIS / DST).By means of a multilayer intensity sampling, velocity comparisons, wavelet power analysis, and sectorial studies of time slices, weretrieve the power distribution, characteristic periodicities, and propagation characteristics of sunspot waves at photospheric and chro-mospheric levels.
Results.
Signatures of RPWs are found at photospheric layers. Those continuous oscillations occur preferably at periods between4-6 min starting at the inner penumbral boundary. The photospheric oscillations all have a slightly delayed, more defined chromo-spheric counterpart with larger relative velocities, which are linked to preceding umbral flash events. In all of the layers, the powerof RPWs follows a filamentary fine-structure and shows a typical ring-shaped power distribution increasing in radius for larger waveperiods. The analysis of time slices reveals apparent horizontal velocities for RPWs at photospheric layers of ≈
51 km / s , which de-crease to ≈
37 km / s at chromospheric heights. The photospheric distribution of peak periods at the inner penumbra resembles thechromospheric cases. Conclusions.
The observations strongly support the scenario of RPWs being upward propagating slow-mode waves guided by themagnetic field lines. Clear evidence for RPWs at photospheric layers is given. Assuming an inverse proportionality of the peak pe-riod and cut-o ff period on the cosine of the field inclination, the penumbral magnetic field inclination is increasing toward the outerpenumbra. The more rapid increase and the larger horizontal velocities at photospheric heights hint at the more horizontal penumbralfield inclination at lower heights. Key words.
Sunspots – Sun: photosphere – Sun: chromosphere – Sun: oscillations – Techniques: imaging spectroscopy
1. Introduction
Sunspot waves are one of the most spectacular and dynamic phe-nomena in the solar atmosphere. Since their first detection, thequestion about the nature of umbral flashes (Beckers & Tallant1969) and running penumbral waves (Giovanelli 1972; Zirin &Stein 1972, henceforth RPWs) and their relation has been un-der intense discussion. Whereas umbral flashes have been com-monly interpreted as upward propagating magnetoacoustic slow-mode waves guided by magnetic field lines (e.g., Centeno et al.2006; de la Cruz Rodr´ıguez et al. 2013), the case for RPWs isnot yet clear.A first scenario suggests that RPWs are purely chromo-spheric waves excited by umbral flashes and propagate radiallyoutward across the chromospheric penumbra (Tziotziou et al.2006). Secondly, recent studies support the scenario that run-ning penumbral waves, like umbral flashes, are channeled, up-ward propagating waves excited at lower layers (Rouppe vander Voort et al. 2003; Kobanov et al. 2006; Bloomfield et al.2007; Jess et al. 2013, e.g.,). The increasing magnetic field in-clination toward the outer penumbra would explain this visualpattern of horizontal propagation at a chromospheric layer sincethe increasing path length for a coherent wave front delays its oc-currence. However, a detailed photospheric observation focusing on RPWs to disprove the purely chromospheric nature of RPWswas still lacking until now.Under the assumption of field-guided waves, the peak periodor frequency in a power spectrum of sunspot waves can be usedas an indicator for the magnetic field inclination. The acousticcut-o ff , which reduces waves with periods above a certain cut-o ff value, is highly dependent on the field topology (Bel & Leroy1977). The e ff ective cut-o ff period increases with the inclinationangle of the field lines. Observational studies (e.g., Jess et al.2013; Yuan et al. 2014) show that the peak period in chromo-spheric power spectra increases from 3 min in the umbral area toaround 10 min toward the outer penumbra. For further informa-tion, see Bogdan & Judge (2006).Most studies on RPWs have focused on the chromosphere.Only a few attempts (e.g., Lites et al. 1998; Bloomfield et al.2007) have been made to determine their photospheric behaviorin detail. In this work, we perform a novel high-resolution, mul-tilayer study concentrating on the photospheric characteristicsof RPWs. In Sect. 2 of this article, we give a brief description ofthe observations, the data, and calibration techniques. In Sect. 3,we present the wave analysis and discuss the results. Finally, inSect. 4 we draw our conclusions and describe the importance ofthe new findings. a r X i v : . [ a s t r o - ph . S R ] J un . L¨ohner-B¨ottcher and N. Bello Gonz´alez: Signatures of running penumbral waves in sunspot photospheres -50 0 50 100 150 200Solar-X ["]-150-200-250-300 S o l a r- Y [ " ] -50 0 50 100 150 200-150-200-250-300 a) -50 0 50 100 150 200Solar-X ["]-50 0 50 100 150 200 c)b) d) -1.0-0.5+0.0+0.5+1.0+1.5+2.0 B LOS [kG]
Fig. 1.
NOAA11823 as seen in continuum (left panels) and line-of-sight magnetic field strength (right panels) on August 21st 2013at 15:00:45UTC as observed by HMI (a + c + d) and ROSA (b). The black squares mark the analyzed region. The insets (b + d) showthe sunspot with contours indicating the inner and outer penumbral boundary from the time-averaged HMI continuum intensity. TheHMI magnetogram is scaled from −
2. Observations and data analysis
To study the wave phenomena in a sunspot (see Fig. 1) fromphotospheric to chromospheric layers, observations were carriedout on August 21st 2013 from 14:53UTC to 15:51UTC at theDunn Solar Telescope (DST) at the National Solar Observatoryin New Mexico. The atmospheric conditions, which are crucialfor ground-based observations, were excellent and stable allow-ing an e ff ective spatial resolution of up to 0 . (cid:48)(cid:48) even for the non-reconstructed data. Simultaneous multi-instrument, multiwave-length observations were performed with the etalon-based imag-ing spectropolarimeter IBIS (Cavallini 2006) in spectrometricservice mode together with the broadband imaging instrumentROSA (Jess et al. 2010) to sample the sunspot’s photosphereand chromosphere (see left panels of Fig. 2). A complete and de-tailed overview of the data will be given in a forthcoming article,including a full time lapse movie of the observations (L¨ohner-B¨ottcher & Bello Gonz´alez, in preparation).In this work, we focus on the narrowband spectromet-ric observations done with the Interferometric BIdimensionalSpectro-polarimeter (IBIS) for the spectral lines Fe I 630 .
15 nm,Na I D1 589 . . .
15 nm intensities stem from photospheric height of 0 km(line continuum) to 300 km (line minimum) above τ = . . − . + b) and line-of-sight magnetic field strength (panels c + d). The sunspot hasa fully-developed circular penumbra, a diameter of 24 Mm andis located close to disk center at a heliocentric angle θ = ◦ at disk coordinates ( X , Y ) = (63 (cid:48)(cid:48) , − (cid:48)(cid:48) ). The HMI magne-tograms revealed the unipolar direction of the magnetic field anda maximum umbral field strength of + (cid:48)(cid:48) fromthe outer penumbral boundary. The reduced field of view, indi-cated by Figs. 1 b + d, has a size of 42 Mm at a pixel scale of0 . (cid:48)(cid:48) px − .The data calibration of the IBIS data involved backgroundintensity subtraction, flat-field calibration, correction for colli-mated wavelength shifts and prefilter transmission, as well asan iterative reduction of temporal image distortions caused byatmospheric turbulences. Doppler velocities for Fe I 630 .
15 nmand Na I D1 589 . ff ects of the magnetic field onthe line profiles. The accuracy of the photospheric velocitiesis confirmed by the Doppler velocities from the Helioseismicand Magnetic Imager (HMI / SDO). These continuous and stablesatellite data using the Fe I line at 617 . . (cid:48)(cid:48) px − serve as context informationfor our observations (see Fig. 4).For the investigation of characteristic frequencies and powerdistribution, we performed a time-position-dependent wavepower analysis based on wavelet techniques (Torrence & Compo1998) as in Bello Gonz´alez et al. (2010). On the basis of a con-tinuous oscillatory stability, we limit this study to the interpre-tation of the global (time-averaged) wavelet power spectra cal-culated at a fine nonlinear scale sampling. One hundred periodsteps sample the range from 26 s to 14 min; the step size in-creases to higher periods. The characteristic interval of sunspotwaves between 2 min and 6 min is scanned by 34 steps. For thepower analysis with respect to the distance to the sunspot center,we included spherical projection e ff ects implied by the sunspotlocation and orientation. We therefore used foreshortened cir-cles to compute the azimuthal average as in L¨ohner-B¨ottcher &Schlichenmaier (2013). Close to disk center, these e ff ects aresmall but should not be neglected. For the temporal evolution of
2. L¨ohner-B¨ottcher and N. Bello Gonz´alez: Signatures of running penumbral waves in sunspot photospheres − − − − − S o l a r − Y [ " ] − − − − − a) T PEAK (P Int, Ca , (cid:104)
2, C − ) b) T PEAK (P Int, Ca , (cid:104)
3, C − ) c) T PEAK (P Int, Ca , (cid:104)
5, Core ) d) T
PEAK (P Int, H (cid:96) ) − − − − − S o l a r − Y [ " ] − − − − − e) T PEAK (P Int, Na , (cid:104)
2, C − ) f) T PEAK (P Int, Na , (cid:104)
3, C − ) g) T PEAK (P Int, Na , (cid:104)
4, Core )
40 50 60 70 8040 50 60 70 80 h) T
PEAK (P Int, CaK )
40 50 60 70 80Solar − X ["] − − − − − S o l a r − Y [ " ]
40 50 60 70 80 − − − − −
240 DC i) T
PEAK (P Int, Fe , (cid:104)
3, C − )
40 50 60 70 80Solar − X ["]40 50 60 70 80 k) T
PEAK (P Int, Fe , (cid:104)
4, C − )
40 50 60 70 80Solar − X ["]40 50 60 70 80 l) T
PEAK (P Int, Fe , (cid:104)
5, Core ) T PEAK [min] Fe I : λ C-8pm ! Fe I : λ C-4pm ! Na I : λ C-12pm ! Na I : λ C-6pm ! Na I : λ Core ! Ca II : λ Core ! Ca II : λ C-22pm ! Cont ! C h r o m o s phe r e ! P ho t o s phe r e ! x ! y ! z ! T PEAK [min] ! Fig. 2.
Three-dimensional view of intensities and peak periods of intensity wave power of NOAA11823 at various wavelengthpositions. The intensities (left) show the sunspot at August 21st 2013 at 15:00:06UTC ( ± .
15 nm, Na I 589 . , and Ca II 854 . ≈ T PEAK of the intensity wave power is shown on the right. The periods are scaled from 2.5 min (darkblue) to 8 min (dark red). The black contours indicate the location of the umbra (inner) and penumbra (outer) in continuum intensity(bottom panel). Whereas the length of the axis arrows represent distances around 1.5 Mm, the image positions along the z-axis arenot to scale.RPW velocities in radial direction, we also used the projectionof circular sectors.
3. Results and discussion
The analysis of this high-resolution, multiwavelength observa-tion of a perfectly stable and symmetrical sunspot yields clearevidence for photospheric RPWs and their field-guided propa-gation to higher layers.The wave power analysis ( P ) for Doppler velocities from thephotospheric Fe I 630 .
15 nm and photospheric-chromosphericNa I D1 589 . P X shifts toward larger periods. The shift in period at the inner penumbra stands in contrast to the constant p -mode signal inthe 4-6 min range at the outer penumbra and quiet sun. Thisevidence for RPWs is even more prominent in the low chro-mosphere where the shift to higher periods is still traceable tothe outer penumbra up to periods of around 10 min. To verifythis radial increase of wave periods, which for chromosphericheights is in line with recent studies (Jess et al. 2013; Yuan et al.2014), the peak periods in the power spectra of both velocity(Fig. 4) and intensity (Fig. 2) oscillations were determined forall heights. Whereas at photospheric layers the radial transitionto larger peak periods (from 4 min to 8 min) is centered in theinnermost penumbra, at higher layers the increase in period issmoother. In chromospheric heights, larger periods are thereforelocated further outside in the penumbra and beyond.The height-sampled distribution of wave power and shiftin peak period hint at the opening field topology of sunspots.According to cut-o ff theory and observations (Bel & Leroy 1977;De Pontieu et al. 2004; Tziotziou et al. 2006), the observed peakperiod T Peak of field-guided waves depends on the acoustic cut-o ff period T cut ,φ and increases with temperature ϑ K and incli-nation angle φ (with respect to the vertical) qualitatively like T Peak ∼ T cut ,φ / . ∼ √ ϑ K / cos φ . Larger inclinations wouldespecially allow the propagation of overpowering high-periodwaves into higher atmospheric layers. Assuming the acousticcut-o ff already occurring for photospheric RPWs, the larger pe-riods would indicate a more horizontal orientation of the penum-
3. L¨ohner-B¨ottcher and N. Bello Gonz´alez: Signatures of running penumbral waves in sunspot photospheres -205-210-215-220-225-230-235 S o l a r- Y [ " ] -205-210-215-220-225-230-235 a) P [4-6min] Na I
50 55 60 65 70 75Solar-X ["]-205-210-215-220-225-230-235 S o l a r- Y [ " ]
50 55 60 65 70 75-205-210-215-220-225-230-235 b) P [4-6min] Fe I T [ m i n ] c) P X Na I T [ m i n ] d) P X Fe I U PU QS
Fig. 3.
Spatial distribution of the time-averaged wavelet power P of the Doppler velocities from Na I 589.6 nm (upper panels) andFe I 630.15 nm (lower panels). Panels a and b show the power inthe 4-6 min band. The power (bright colors indicate high power)is plotted in logarithmic (arbitrary) units to assure best contrastin the penumbra. The contours mark the inner and outer penum-bral boundary in continuum intensity. Circular sectors 1-3 (whitesolid) follow the foreshortened projection of a circle (dotted).Panels c and d show the azimuthally averaged global power spec-trum P X for periods T (in min) according to the distance X (inarcsec) from the spot center. The white contours mark the impor-tant elements in the spatial power analysis. For spatial compari-son, an arbitrary radial element in continuum intensity is addedbelow. The dashed lines indicate the inner and outer penumbralboundaries.bra at photospheric than at chromospheric levels. In additionto the spectroscopic study of IBIS data, the analysis of HMIDoppler velocities and their power distribution confirm the pho-tospheric signatures of RPWs, though at a lower spatial and tem-poral resolution (compare Fig. 4 a + b).To prove the field-aligned propagation of RPWs in the lowersunspot atmosphere, we study the temporal evolution and trajec-tories of the waves. Therefore, we perform a sectorial time-sliceanalysis for the photospheric and chromospheric Doppler veloc-ities, as shown in Fig. 5. To investigate the fluctuations, we sub-tract the temporally averaged Doppler velocities and focus on therelative velocities. Since the RPWs seem to follow the filamen-tary fine-structure of the penumbra and under the assumptionof coherent wave trains for nearby fields, we select three circu-lar (foreshortened) sectors with opening angles of 10 ◦ for thepenumbral regions with highest velocity power (see Fig. 3 a,b).The temporal behavior of the azimuthal average at each distancefrom the spot center is shown in Fig. 5 for the velocities derivedfrom Na I 589.6 nm (panels a-c) and Fe I 630.15 nm (panels d-f). A detailed study of the oscillatory pattern demonstrates thephotospheric origin of RPWs as follows: – Continuous and clear signatures of penumbral oscillation atphotospheric height are found. – Photospheric oscillations at the inner penumbra have an am-plitude of up to 0 . / s. -210-220-230-240 S o l a r- Y [ " ] -210-220-230-240 a) Fe I I
50 60 70 80Solar-X ["]-210-220-230-240 S o l a r- Y [ " ]
50 60 70 80-210-220-230-240 c) Na I DC T PEAK [min]
Fig. 4.
Spatial distribution of the peak periods from the globalpower spectra of Doppler velocity oscillations. The major pe-riods T PEAK (in min) are shown for a) Fe I 630.15 nm, b) Fe I617.3 nm (HMI), and c) Na I 589.6 nm. The scale ranges from2.5 min (dark blue) to 8 min (dark red). The black contours markthe umbral and penumbral boundaries in continuum intensity.The white arrow is pointing to the disk center. – All sectors provide observational evidence for outward-directed horizontal propagation in the inner penumbra. – In the inner penumbra close to the umbral boundary, allphotospheric waves have a slightly delayed chromosphericcounterpart with larger oscillatory amplitudes of up to0 . / s. – A simple analysis of the slopes of all wave trains over thefirst 4 (cid:48)(cid:48) of the inner penumbra was performed in the sectorialtime slices shown in Fig. 5. The analysis reveals the apparenthorizontal velocities, v
HOR , combined to Fig. 6. For the pho-tospheric layers (black symbols and histogram), representedby the Fe I 630.15 nm line, the apparent horizontal velocitiesaverage to (cid:104) v HOR , Fe I (cid:105) = ±
13 km / s. – At upper photospheric to chromospheric height, representedby the Na I 589.6 nm line (Fig. 6, red symbols and his-togram), the wave trains indicate smaller apparent horizontalvelocities of (cid:104) v HOR , Na I (cid:105) = ±
10 km / s. The smaller veloc-ity is illustrated exemplarily in Fig. 5 by the steeper slopescompared to the photospheric case.At chromospheric layers, the retrieved apparent horizontalvelocities of around 20 −
50 km / s at the inner penumbra are inline with recent studies (e.g., Tziotziou et al. 2006; Kobanovet al. 2006; Jess et al. 2013). Commonly, this apparent propa-gation speed decreases radially toward the outer penumbra. Asobservations at higher chromospheric to transition region layershave shown (Madsen et al. 2015), the retrieved horizontal veloc-ities decrease to around 10 km / s. Vice versa, the larger appar-ent velocities of around 30 −
70 km / s at photospheric layers fitwell into the trend of decreasing horizontal velocities at the in-ner penumbra toward higher layers. The topological model of thesunspot’s magnetic field (e.g., Westendorp Plaza et al. 1997), the
4. L¨ohner-B¨ottcher and N. Bello Gonz´alez: Signatures of running penumbral waves in sunspot photospheres t i m e t [ m i n ] v Na I (t) a) Sector 1
U PU b) Sector 2
U PU c) Sector 3
U PU − − − (cid:54) V LOS [km/s] t i m e t [ m i n ] v Fe I (t) d)U PU e)U PU f)U PU − − − (cid:54) V LOS [km/s]
Fig. 5.
Temporal evolution (in min) of the relative velocities inthe sunspot atmosphere. The analysis of time slices was per-formed for the Doppler velocities V Na I (panels a-c) of Na I589.6 nm and V Fe I (d-f) of Fe I 630.15 nm in all three sectors,also shown in Fig. 3. For each sector, the azimuthal average at aradial distance X (in arcsec) from the spot center was calculated.The dashed lines mark the umbral boundary. The velocity scale ∆ V LOS ranges between ± . / s (a-c) and ± . / s (d-f). Theblack bars trace the apparent wave trains.field-guided propagation of running penumbral waves, and theirvisual appearance at a certain layer can explain this behavior(Bloomfield et al. 2007; Madsen et al. 2015). In the penumbra,the magnetic field inclination increases radially from the ver-tical umbral field. At photospheric layers, this bending of themagnetic field lines is stronger than at higher atmospheric lay-ers in which the radial increase in field inclination happens moresmoothly. As stated by Bogdan & Judge (2006), the large appar-ent horizontal velocity of RPWs (here up to 80 km / s in Fig. 6)rather reflects relative travel time di ff erences for waves guidedby the individual less and more inclined magnetic field lines ofthe inner penumbra.
4. Conclusions
We provide strong evidence for the presence of RPWs in thesunspot photosphere. Our observations conflict with the scenariothat RPWs are a purely chromospheric wave phenomenon andstrongly support the recent theory (Rouppe van der Voort et al.2003; Bloomfield et al. 2007; Jess et al. 2013, e.g.,) that RPWsare upward propagating slow-mode waves guided by the mag-netic field lines. Under the assumption of an expanding magneticfield topology with more inclined field lines with radial distancefrom the vertically oriented umbra, the performed wave poweranalysis substantiates the dependence of the power spectra onthe atmospheric height and inclination of the magnetic field. v H O R [ k m / s ] Fe I , Sector1Fe I , Sector2Fe I , Sector3 Na I , Sector1Na I , Sector2Na I , Sector3 (cid:157) v HOR, Fe I (cid:156) = ± km/s (cid:157) v HOR, Na I (cid:156) = ± km/s Fig. 6.
Apparent horizontal velocities, v
HOR , of running penum-bral waves at two atmospheric layers. The results for the photo-spheric Fe I line (Fig. 5 d-f) are shown in black, for the chromo-spheric Na I line (Fig. 5 a-c) in red. Left panel: the apparent hor-izontal velocities (in km / s) over the first 4 (cid:48)(cid:48) of the inner penum-bra are plotted according to the observation time, t . The aster-isks, diamonds, and triangles show the apparent speeds of thewave trains in sectors 1, 2, and 3. The dashed lines indicate theaverage apparent velocities, (cid:104) v HOR , Fe I (cid:105) and (cid:104) v HOR , Na I (cid:105) for bothlayers. The averages and their standard deviations are given inthe figure legend. Right panel: histogram showing the number N of velocity values v HOR from the left panel within a 5 km / sbinning interval.The solar atmosphere exhibits a wealth of dynamical phe-nomena at all scales. These processes and their e ff ects arestrongly coupled. Especially in the case of sunspots, oscilla-tions, flows, and other dynamical e ff ects can interact and haveto be taken into account in observational studies (e.g., EstebanPozuelo et al. 2015).A forthcoming, extensive study of the presented data(L¨ohner-B¨ottcher & Bello Gonz´alez, in preparation) will includean analysis of umbral flashes in the chromosphere (a bright um-bral flash event can be seen in Fig. 2), their characteristic simi-larities and di ff erences from RPWs, and the possible reconstruc-tion of the magnetic field topology using the wave characteris-tics in sunspots. To verify the existence of photospheric RPWs,we suggest further observations with high spatial and temporalresolution using photospheric spectral lines that are magnetic in-sensitive (with a Land´e-factor g = Acknowledgements.
The data were acquired in service mode operation withinthe transnational ACCESS program of SOLARNET, an EU-FP7 integrated activ-ity project. The instruments IBIS and ROSA at the Dunn Solar Telescope (DST,NSO) were operated by INAF and QUB personnel, with special thanks to GiannaCauzzi and Peter Keys. HMI data were used by courtesy of NASA / SDO andHMI science teams. This work was prepared at the Centre for Advanced SolarSpectro-polarimetric Data Analysis (CASSDA), funded by the Senatsausschussof the Leibniz Association, Ref.-No. SAW-2012-KIS-5. We thank WolfgangSchmidt for his fruitful comments on the manuscript.
5. L¨ohner-B¨ottcher and N. Bello Gonz´alez: Signatures of running penumbral waves in sunspot photospheres
References