Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://star.arm.ac.uk/preprints/2007/503.pdf Äàòà èçìåíåíèÿ: Thu Aug 16 17:57:54 2007 Äàòà èíäåêñèðîâàíèÿ: Tue Oct 2 02:53:57 2012 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: solar eclipse

Solar Physics DOI: 10.1007/·····-···-···-····-·

Present and Future Observing Trends in Atmospheric Magneto-seismology
D. Banerjee1 , R. ErdÈlyi2 , R. Oliver3 , E. O'Shea4
c Springer ····

Abstract With modern imaging and spectral instruments observing in the visible, EUV, X-ray and radio wavelengths, the detection of oscillations in the solar outer atmosphere has become a routine event. These oscillations are considered to be the signatures of a wave phenomenon and are generally interpreted in terms of magnetohydrodynamic (MHD) waves. With multi-wavelength observations from ground and space-based instruments, it has been possible to detect waves in a number of different wavelengths simultaneously and, consequently, to study their propagation properties. Observed MHD waves propagating from the lower solar atmosphere into the higher regions of the magnetized corona have the potential to provide an excellent insight into the physical processes at work at the coupling point between these different regions of the Sun. High-resolution wave observations combined with forward MHD modelling can give an unprecedented insight into the connectivity of the magnetized solar atmosphere, which further provides us with a realistic chance to reconstruct the structure of the magnetic field in the solar atmosphere. This type of solar exploration has been termed atmospheric magneto-seismology. In this review we will summarise some new trends in the observational study of waves and oscillations, discussing their origin, and their propagation through the atmosphere. In particular, we will focus on waves and oscillations in open (e.g., solar plumes) and closed (e.g., loops and prominences) magnetic structures, where there have been a number of observational highlights in the last few years. Furthermore, observations of waves in filament fibrils allied with a better characterization of their propagating and damping properties, the detection of prominence oscillations in UV lines, and the renewed interest in largeamplitude, quickly attenuated, prominence oscillations, caused by flare or explosive phenomena, will be addressed. Keywords: Coronal loops, MHD Waves, MHD Oscillations
Indian Institute of Astrophysics, Koramangala, Bangalore 560034 (e-mail: dipu@iiap.res.in) 2 Solar Physics and Space Plasma Research Centre (SP2 RC), Department of Mathematics, University of Sheffield, Hicks Building, Hounsfield Road, Sheffield, S3 7RH, (England) UK (e-mail: robertus@sheffield.ac.uk) 3 Departament de FÌsica, Universitat de les Illes Balears, E-07122, Palma de Mallorca, Spain (e-mail: ramon.oliver@uib.es) 4 Armagh Observatory, College Hill, Armagh BT61 9DG, N. Ireland (e-mail: eos@arm.ac.uk)
1


Banerjee et al.

1. Intro duction From Solar and Heliospheric Observatory (SOHO) and Transition Region And Coronal Explorer (TRACE) data, new results, that shed light onto dynamical events in the outer solar atmosphere, especially short time-scale variability and/or oscillations at EUV wavelengths, have emerged. The detection of waves in the outer solar atmosphere is made possible by the effect these waves have on the plasma. For example, the presence (or signature) of waves may be detected in the form of variations or oscillations in the radiant flux ("intensity"), due to changes in the plasma density, or in the line-of-sight velocities, due to motions in the plasma, both measurable from spectral lines. These periodic motions are generally interpreted in terms of magnetohydrodynamic (MHD) waves. They carry information from the emitting regions allowing a diagnosis of the frozen-in magnetic fields as well as the plasma contained in different magnetic structures, e.g., coronal loops. The wavelengths of these waves are often comparable to the characteristic sizes of coronal structures, their time scales are in the range of seconds to minutes and they are detectable from space and by ground-based instruments, e.g., the detection of EIT (or coronal Moreton) waves (Thompson et al., 1998), compressible waves in polar plumes (Ofman et al., 1997; DeForest and Gurman, 1998) or periodic phenomenon in the corona from ground-based observations (Aschwanden, 1987). Thus, imaging instruments (from space and ground) have uncovered a myriad of wave detections in the corona, which have been reviewed at length in Aschwanden (2003, 2004, 2006), De Moortel (2005, 2006), De Pontieu and ErdÈlyi 2006, ErdÈlyi, (2006a,b), Nakariakov and Roberts, (2003), Nakariakov and Verwichte, (2005), Nakariakov, (2006) and Wang, (2004). In this review we will report on current trends in the observational study of MHD waves. Summaries will be provided for imaging observations together with a slightly more detailed description of spectral methods as these have not been dealt with in previous reviews. It is not the purpose or intention of this review to make an exhaustive list of all observations. Instead, we seek to present a complementary view to those mentioned above by focusing on some recently reported observations of waves, particularly those related to spectroscopic and not imaging methods. We will also briefly address the status of prominence oscillations in a separate section, stressing their importance as a natural example and as a tool for studying wave signatures.

2. MHD Waves in the Lower Solar Atmosphere The solar atmosphere from its visible lower boundary, the photosphere, through a transitional layer with sharp gradients, up to its open-ended magnetically dominated upper region, the corona, is magnetically coupled. This physical coupling is obvious when one overlays concurrently taken snapshots of the various solar atmospheric layers as a function of height and a magnetogram obtained at the same time at photospheric heights. A typical magnetic field concentration, e.g., an active region or an intense flux tube, will show up as a strong brightening at corresponding locations in the UV, EUV, and X-ray images indicating evidence in support of the coupling of the all pervasive magnetic field. Recent high-resolution satellite, and ground-based technology provides us with unprecedented fine-scale spatial and temporal resolution data of different magnetic

Rev_sol_press_final.tex; 9/08/2007; 19:07; p.2


Observing Trends in Atmospheric Magneto-seismology

structures in the solar atmosphere (e.g., plumes, coronal loops, arcades, and even dynamic features such as spicules) that support periodic motions (propagating waves or oscillations) on wide spatial and time scales. The large concentrated magnetic structures at photospheric to low-TR and coronal heights serve as excellent waveguides for the propagation of perturbations excited at footpoint locations. These observed oscillations within the magnetic structures, being intrinsically locked into them (in contrast to the acoustic solar global oscillations that are ubiquitous in the solar interior) provide us with the tools to diagnose the structures themselves. 2.1. Wave leakage from the photosphere As the acoustic wave frequency increases beyond 5.3 mHz, the upper boundary of subsurface cavities becomes increasingly transparent and the acoustic waves are able to propagate into the Sun's chromosphere. The high-frequency waves may therefore convey information about the properties of the chromosphere. Using time-distance analysis of solar acoustic waves with frequencies above the nominal atmospheric acoustic cutoff frequency (5.3 mHz) Jefferies et al., (1997) showed that the waves can be partially reflected at both the Sun's photosphere and a layer located higher in the atmosphere. From one dimensional spectroscopic observations, Baudin, Bocchialini and Koutchmy, (1996) showed for the first time that upward propagating five minute waves emerge from the deep chromospheric network. They suggested that the waves propagating in the open corona are reminiscent of photospheric oscillations transmitted by the magnetic field of the chromospheric network. Using the Magneto-Optical filters at Two Heights (MOTH) instrument, Finsterle et al., (2004) have recorded simultaneous dopplergrams at a high cadence (ten second sampling intervals) in two Fraunhofer lines formed at different heights in the solar atmosphere. They found evanescent-like waves at frequencies substantially above the acoustic cut-off frequency in regions of intermediate magnetic field. Furthermore, upwardlyand downwardly-propagating waves were detected in areas of strong magnetic field such as sunspots and plage: even at frequencies below the acoustic cut-off frequency. They conjectured that the interaction of the waves with the magnetic field must be a non-linear process depending on the field strength and/or inclination. Very recent observations of the transition region (hereafter, TR), in particular spicules and moss oscillations, detected by TRACE and by SUMER on board SOHO brings us closer to an understanding of the origin of running (propagating) waves in coronal loops. The correlations on arcsecond scales between chromospheric and transition region emission in active regions were studied in De Pontieu, Tarbell, and ErdÈlyi (2003). The discovery of active region moss (Berger et al., 1999), i.e., dynamic and bright upper transition region emission above active region (AR) plage, provides a powerful diagnostic tool to probe the structure, dynamics, energetics and coupling of the magnetized solar chromosphere and transition region. De Pontieu, Tarbell, and ErdÈlyi (2003) studied the possibility of the direct interaction of the chromosphere with the upper TR, by searching for correlations (or lack thereof ) between emission at varying temperatures using concurrently observed EUV lines emitted from the low chromosphere (Ca II K-line), the middle and upper chromosphere (H ), the low TR (C iv 1550 å at 0.1 MK), and from the upper TR (Fe ix/x 171 å at 1 MK and Fe xii 195 å at 1.5 MK). The relatively high cadence (24 to 42 second) data sets obtained with the Swedish Vacuum Solar Telescope (SSVT, La Palma) and TRACE allowed them to find a relationship between upper TR oscillations and low-lying

Rev_sol_press_final.tex; 9/08/2007; 19:07; p.3


Banerjee et al.

Figure 1. Demonstrating the using wavelet power spectra for for TRACE 171 å (full, with tr Units of intensity are arbitrary

correlation between chromospheric and upper-TR oscillations TRACE 171 å, H - 350 må, H + 350 må and light curves iangles), H - 350 må (full blue) and H + 350 må (full red). (From De Pontieu, 2004).

Rev_sol_press_final.tex; 9/08/2007; 19:07; p.4


Observing Trends in Atmospheric Magneto-seismology

photospheric oscillations. Figure 1 (from De Pontieu, 2004) shows a typical example demonstrating the correlation between chromospheric and upper-TR oscillations. The wavelet power spectra for TRACE 171 å (top panel), H ­ 350 må (second from top), H +350 må (third panel from top) and light-curves (bottom panel) for TRACE 171 å (full, with triangles), H ­ 350 må (full blue) and H + 350 må (full red), are quite similar, despite the atmospheric-seeing deformations the ground-based data suffer from. While there is generally a good correlation between the TRACE 171 å signal and the wings of H , there is often a delay between the H ­ 350 må and H + 350 må signals, usually of the order of 60 to 100 seconds. A simple estimate using this phase delay and the physical distance between the line formation of TRACE Fe IX/X 171 å lines has led to the possible conclusion of direct wave leakage. This correlation analysis provides clues to the understanding of the coupling between the different layers of the solar atmosphere. De Pontieu, ErdÈlyi, and de Wn (2003a) analysed intensity oscillations in the upper- TR above AR plage. They suggested the possible role of a direct photospheric driver in TR dynamics, e.g., in the appearance of moss (and spicule) oscillations. Wavelet analysis of the observations (by TRACE) found strong ( 5 ­ 15%) intensity oscillations in the upper TR footpoints of hot coronal loops. A range of periods from 200 to 600 seconds, typically persisting for about four to seven cycles was found. A comparison with photospheric vertical velocities (using the Michelson Doppler Imager onboard SOHO) revealed that some upper TR oscillations showed a significant correlation with solar global acoustic p-modes in the photosphere. In addition, the ma jority of the upper-TR oscillations were directly associated with upper chromospheric oscillations observed in H , i.e., periodic flows in spicular structures. The presence of such strong oscillations at low heights (of the order of 3 000 km) provides an ideal opportunity to study the direct propagation of oscillations from the photosphere and chromosphere into the TR (De Pontieu, ErdÈlyi, and James 2004) and low magnetic corona (see, for example, De Pontieu, ErdÈlyi, and De Moortel 2005). These type of measurements can also help us to (i) understand atmospheric magnetic connectivity, that is so crucial for diagnostic reconstruction in the chromosphere/TR, and shed light on the dynamics of the lower solar atmosphere, e.g., the source of chromospheric mass flows such as spicules (e.g., De Pontieu, ErdÈlyi, and James 2004); (ii) explore the dynamic and magnetised lower solar atmosphere using the method of seismology. This latter aspect is discussed in detail in recent review papers by e.g., De Pontieu and ErdÈlyi (2006) and ErdÈlyi (2006a). On the nature of oscillations in sunspots, Bogdan (2000) has summarized the observational and theoretical components of the sub ject in a coherent, common, and conceptual manner. We will not carry out a detailed review of this sub ject here but we would like to mention some recent developments. O'Shea, Muglach and Fleck (2002) reported oscillations within the umbra at different temperatures, from the temperature minimum as measured by TRACE 1700 å up to the upper corona as measured by CDS Fe xvi 335 å (log T = 6.4 K). Using cross-spectral analysis, time delays were found between low- and high-temperature emission suggesting the possibility of both upward and downward wave propagation. Earlier observations indicated that the waves responsible for these oscillations may not be reaching the corona. Based on a similar observing campaign as O'Shea, Muglach and Fleck (2002), and using TRACE and SOHO, Brynildsen et al., (2002) found that the oscillation amplitude above the umbra increases with increasing temperature, reaching a maximum for emission lines

Rev_sol_press_final.tex; 9/08/2007; 19:07; p.5


Banerjee et al.
a b

Figure 2. Leakage of evanescent photospheric p-mode power into the chromosphere. Distribution of wavelet power (for cases a and b, respectively = 0 and 50 ), in arbitrary units, independent for each height as a function of wave period for different heights above the photosphere. Vertical flux tubes (a) allow minimal leakage of p modes with periods of 300 seconds (> Pc 220 seconds), so that only oscillations with lower periods (< 250 seconds) can propagate and grow with height to dominate chromospheric dynamics. Inclined flux tubes (b) show an increased acoustic cut-off period Pc , allowing enhanced leakage and propagation of normally evanescent p modes. Adapted from De Pontieu, Erd´lyi, and James (2004). e

formed close to 1 ­ 2â105 K, and decreasing for higher temperatures. Furthermore, they report that three minute oscillations fill the sunspot umbra in the transition region, while in the corona the oscillations are concentrated in smaller regions that appear to coincide with the endpoints of sunspot coronal loops. This suggests that wave propagation along the magnetic field makes it possible for oscillations to reach the corona. However, it must be pointed out that Doyle, DzifÀkovÀ, and Madjarska (2003) discussed the possibility that the observed oscillations seen in TRACE 171 å by Brynildsen et al., (2002) and Mg ix 368 å (and other coronal lines) by O'Shea, Muglach and Fleck (2002) may not actually be coronal in origin due to the effect of non-Maxwellian contributions. 2.2. The Source of Propagating Waves In order to answer the question of where propagating coronal waves originate from, and, inspired by the observational findings of similarities between photospheric and TR oscillations, De Pontieu, ErdÈlyi, and James (2004) developed the general framework of how photospheric oscillations can leak into the atmosphere along inclined magnetic-flux tubes. In a non-magnetic atmosphere p modes are evanescent and cannot propagate upwards through the temperature minimum barrier since their period P ( 200 ­ 450 seconds) is above the local acoustic cut-off period Pc 200 seconds. However, in a magnetically structured atmosphere, where the field lines have some natural inclination (), where is measured between the magnetic guide channelling he oscillations and the vertical, the acoustic cut-off period takes the t form Pc T / cos with the temperature T . This inclination will allow some nonpropagating evanescent wave energy to tunnel through the temperature minimum into the hot chromosphere of the waveguide, where propagation is once again possible because of higher temperatures (Pc > 300 seconds). The authors have shown that the inclination of magnetic flux tubes (applicable, e.g., to plage regions) can

Rev_sol_press_final.tex; 9/08/2007; 19:07; p.6


Observing Trends in Atmospheric Magneto-seismology

Figure 3. Wavelet power of loop intensity oscillations as a function of time and wave period, as observed with TRACE (top panel, case 16a of De Moortel et al., (2002a)) and for simulations (bottom panel) driven by MDI velocities at the loop footpoint region. Middle panel shows the running difference (I ) of loop intensity at one location (relative to total intensity I ) as a function of time for observations (full line, diamonds) and simulations (dashed). The area contained between the horizontal axis and cone of influence is free of edge effects introduced by the wavelet analysis. Adapted from De Pontieu, Erd´lyi, and De Moortel (2005). e

dramatically increase tunnelling, and may even lead to direct propagation of p-modes along inclined field lines, as plotted in Figure. 2. McIntosh and Jefferies (2006) have demonstrated observationally that the acoustic cutoff frequency in the lower solar chromosphere can be modified by changing the inclination of the magnetic field in the lower solar chromosphere. Though they have demonstrated this effect from a study of a sunspot with TRACE, they expect a similar modification of cutoff frequency to occur whenever plasma conditions permit (low- , high-inclination magnetic fields) elsewhere on the Sun, in particular in magnetically- intense network bright points anchored in super-granular boundaries. A natural generalisation of the above idea was put forward by De Pontieu, ErdÈlyi, and De Moortel (2005) who proposed that a consequence of the leakage of photospheric oscillations is that spicule driven quasiperiodic shocks propagate into the low corona, where they may lead to density, and thus intensity oscillations with properties similar to those observed by TRACE in one MK coronal loops. In other words, the origin of the propagating slow MHD waves detected in coronal loops (see a recent review on their properties by, e.g., De Moortel, 2006) is linked to wave energy leakage of solar global standing oscillations. De Pontieu, ErdÈlyi, and De Moortel (2005) highlighted that oscillations along coronal loops associated with AR plage have many properties that are similar to those of moss oscillations: (i) the range of periods is from 200 to 600 seconds, with an average of 350 ± 60 seconds and 321 ± 74 seconds, for moss and coronal oscillations, respectively; (ii) the spatial extent for coherent moss oscillations is about 1 ­ 2 ,

Rev_sol_press_final.tex; 9/08/2007; 19:07; p.7


Banerjee et al.

whereas for coronal waves, the spatial coherence is limited to 2 in the direction perpendicular to that of wave propagation. They also point out that, although the oscillations in moss and corona have similar origins, they are the result of different physical mechanisms: moss oscillations occur because of periodic obscuration by spicules, and coronal oscillations arise from density changes associated with the propagating magneto-acoustic shocks that drive the periodic spicules. A typical example of a comparison of the observed properties of coronal intensity oscillations with synthesized observations is shown in Figure. 3.

3. Propagating Waves into the Corona In the pre-SOHO/TRACE era, the first observations of MHD waves in the corona were reported by Chapman et al., (1972) with the GSFC extreme-ultraviolet spectroheliograph on OSO-7 (the spatial resolution was a few arcsec, the cadence time was 5.14 seconds). In Mg vii, Mg ix and He ii emission intensity periodicity of about 262 seconds was detected. The importance of this early work is that within the range of low-frequencies, an analogy to photospheric and chromospheric oscillations was found and, it was further speculated that the photospheric and chromospheric evanescent waves become vertically propagating, gravity-modified acoustic waves at a height in the chromosphere where a temperature rise admits propagation again. Antonucci, Gabriel, and Patchett (1984) using the Harvard College Observatory EUV spectroheliometer on Skylab detected oscillations in the C ii, O iv, and Mg x emission intensity with periods of 117 seconds and 141 seconds. They suggested that the intensity fluctuation of the EUV lines was caused by small amplitude waves, propagating in the plasma confined in the magnetic loop, and that the size of the loop might be important in determining its preferential heating in the active region. A final example from that era, though at much shorter wavelengths, is the observation by Harrison (1987), who detected, with the Hard X-ray Imaging Spectrometer onboard SMM, soft X-ray (3.5 ­ 5.5 keV) pulsations of 24 minute period lasting for six hours. The periodicity was thought to be produced by a standing wave or a travelling wave packet which existed within the observed loop. It was concluded that the candidates for the wave were either fast- or AlfvÈn- MHD modes of AlfvÈnic surface waves. Since the launches of SOHO and TRACE, and the abundant evidence that has emerged for MHD phenomena and, in particular, propagating waves, our views have changed considerably. However, the source of propagating waves still remains a puzzle. 3.1. Waves in Open Structures Propagating waves may propagate in open (e.g., plumes) and closed (e.g., loops) coronal magnetic structures. The first undoubted detection of propagating slow MHD waves was made by the Ultraviolet Coronagraph Spectrometer (UVCS/SOHO). Detection of slow waves in an open magnetic structure high above the limb of coronal holes was reported by Ofman et al., (1997, 2000a). DeForest and Gurman (1998), analysing Extreme-ultraviolet Imaging Telescope (EIT/SOHO) data of polar plumes, detected similar compressive disturbances with linear amplitudes of the order of 10 ­ 20% and periods of 10 ­ 15 minutes. Ofman, Nakariakov, and DeForest (1999)

Rev_sol_press_final.tex; 9/08/2007; 19:07; p.8


Observing Trends in Atmospheric Magneto-seismology Table 1. Overview of the periodicities and propagation speeds of propagating slow MHD waves detected in coronal structures. Authors Berghmans and Clette (1999) Nightingale, Aschwanden, and Hurlburt (1999) Schrver et al. (1999) Banerjee, O'shea, and Doyle (2000) Banerjee et al., (2001a) Banerjee et al., (2001b) De Moortel, Ireland, and Walsh (2000) Robbrecht et al., (2001) Berghmans, McKenzie, and Clette (2001) De Moortel et al., (2002a) De Moortel et al., (2002b) Sakurai et al., (2002) King et al., (2003) Popescu et al., (2005) O'Shea, Banerjee, and Doyle (2006) O'Shea, Banerjee, and Doyle (2007) Perio d (s) 600 ­ 300 600 ­ 1200 (plume) 1200 ­ 1800 (inter-plume) 600 ­ 1200 (coronal hole) 180 ­ 420 (282 ± 93) 172 ± 32 (sunspot) 321 ± 74 (plage) 180 ­ 600 120 ­ 180 & 300 ­ 480 600 ­ 5400 & 10200 (off-limb) 300 ­ 1000(off-limb) 300 ­ 1000(coronal hole) Speed (km/s) 75 ­ 200 130 ­ 190 70 ­ 100 70 ­ 165 65 ­ 150 300 122 ± 43 100 ­ 200 25 ­ 40 ­ 150 ­ 170 50 ­ 70 Wavelength 195 171 & 195 195 629 629 629 171 171 & 195 SXT 171 171 171 5303 171 & 195 SUMER CD S CD S

and Ofman, Nakariakov, and Seghal (2000b) identified the observed compressive longitudinal disturbances as propagating slow MHD waves. We have summarized the main features of the observed oscillations following De Moortel (2006) in Table 1. A number of studies using the CDS and SUMER spectrographs on SOHO have reported oscillations in plumes, interplumes and coronal holes in the polar regions of the Sun (e.g., Banerjee, O'shea, and Doyle 2000; 2001a,b, O'Shea, Banerjee, and Doyle , 2006; 2007). All of these studies point to the presence of compressional waves, thought to be slow magnetoacoustic waves as found by DeForest and Gurman (1998). The detected damping of slow propagating waves was attributed to compressive viscosity. Up to now evidence for the fast magnetoacoustic wave modes in these same regions has been absent, even though recent results by Verwichte, Nakariakov, and Cooper (2005) have shown that propagating fast magnetoacoustic waves can be present in open magnetic field structures, albeit in this instance, in a post-flare supra-arcade. For the fast mode the wavelengths of the propagating wave should be much shorter than the size of the structure guiding the wave. Shorter wavelength implies shorter period, thus it demands high cadence observations. TRACE can work on 20 ­ 30 second cadence, allowing us to detect a wave with a 40 ­ 60 second periodicity at

Rev_sol_press_final.tex; 9/08/2007; 19:07; p.9


Banerjee et al.

best. Thus it is difficult to detect smaller periodicity with the present space-based instruments, whereas ground-based coronagraphs and radioheliographs have much better time resolution and have been used for the detection of short-period waves. 3.2. Waves in Closed Structures Koutchmy, Zhugzhda, and Locans (1983) devoted an experiment to the search for short period coronal waves using the green coronal line 5303 å of Fe xiv. Their power spectra showed evidence of Doppler-velocity oscillations with periods near 300 seconds, 80 seconds, and, especially, 43 seconds. However no prominent intensity fluctuations were reported. Though Koutchmy considered that their oscillations were due to resonant AlfvÈn oscillations viewed at a low level through several legs of coronal arches, later these data were re-interpreted as standing kink waves by Roberts, Edwin, and Benz (1984). The first detection of microwave quasi periodic pulsations, with a periodicity of 6.6 seconds, which could be associated with the fast kink mode was performed by Asai et al., (2001) with the Nobeyama radioheliograph. Four bursts were observed with the hard X-ray telescope onboard YOHKOH and the Nobeyama Radioheliograph during the impulsive phase of the flare. Williams et al., (2001, 2002) and Katsiyannis et al., (2003) reported on the presence of high-frequency MHD waves in coronal loops observed during a total solar eclipse with the SECIS instrument. The detections lie in the frequency range 0.15 ­ 0.25 Hz (4 ­ 7 seconds), last for at least 3 periods at a confidence level of more than 99%, and arise just outside known coronal loops. This led them to suggest that they occur in low emission-measure or different temperature loops associated with active regions. Madjarska et al., (2003), using a number of different transition region and coronal lines from SUMER on SOHO, were the first to report oscillations in coronal bright points, finding a periodicity of six minutes Ugarte-Urra et al., (2004), using data from CDS on SOHO, found evidence of oscillations occurring with periods between 420 ­ 650 seconds in a number of TR lines (O v and O iii) but none in the coronal line of Mg ix. They also report on a separate measurement of an oscillation with a period of 491 second period observed with the transition region line of S iv of SUMER in a bright point. Using EIT/SOHO, Berghmans and Clette (1999) were the first to report on slow modes in closed loop structures. Following the success of SOHO, observers using TRACE also searched successfully for quasi-periodic disturbances in coronal loops (e.g., Schrver et al., 1999; Nightingale, Aschwanden, and Hurlburt 1999; De Moortel, Ireland, and Walsh 2000). A detailed overview of the observed properties of these propagating intensity perturbations is given by De Moortel et al., (2002a, b). From a ground-based coronagraphic observation at the Norikura Solar Observatory, Sakurai et al., (2002) have reported on the detection of coronal waves from Doppler velocity data. The propagation speed of the waves was estimated by correlation analysis. The line intensity and line width did not show clear oscillations, but their phase relationship with the Doppler velocity indicated propagating waves rather than standing waves. In all of the reported cases, the phase speed is of the order of the coronal sound speed. In TRACE observations the propagating waves are observed as intensity oscillations, hence they are likely to be candidates for compressive disturbances. No significant acceleration or deceleration was observed. The combination of all of

Rev_sol_press_final.tex; 9/08/2007; 19:07; p.10


Observing Trends in Atmospheric Magneto-seismology

these facts leads to the most plausible conclusion that the observed propagating waves are indeed slow MHD waves. 3.3. Detection of Waves Through Statistical Methods Most of the aforementioned detection was restricted to a few specific case studies. A new approach has been taken up by O'Shea et al., (2001), where wavelets were used to measure oscillations in a statistical manner. A novel randomisation method was used to test their significance. This form of statistical testing is useful as it provides a more accurate picture of the processes at work in the atmosphere than a smaller number of discrete observations can. Recently McEwan and DeMoortel (2006a) have studied a number of examples of observed longitudinal oscillations in coronal loops to find evidence of the small temporal and spatial scales of these loop oscillations. Increasing the number of observed longitudinal oscillations allowed an improvement in the statistics of the measured parameters, providing more accurate values for numerical and theoretical models. O'Shea et al., (2001) studied several active regions using data from the the Coronal Diagnostic Spectrometer (CDS) (Harrison et al., 1995). Three different lines were used, a TR line of O v and two coronal lines of Mg ix and Fe xvi. For this work three different active regions were studied in a statistical way, using 17 individual datasets in total to build up histograms of the typical oscillation frequencies present in all of the active regions. In Figure. 4, the combined histogram (of primary and secondary) frequencies measured in the intensity (flux) (top row) and the combined histogram of the frequencies measured in the velocities (bottom row) is shown. Comparing these plots, it is clear that the coronal lines of Mg ix and Fe xvi contain more significant oscillations in velocity than in intensity, which suggests that in the velocity additional non-compressive modes are being measured. This suggests that these non-compressive modes are perhaps being produced in, and confined to, the corona. This effect is not seen in the transition region line of O v suggesting a change between the different temperature regimes of the TR and corona. Recently, O'Shea, Banerjee, and Doyle (2006, 2007), have used measurements of spectral lines obtained from CDS to perform a statistical search for the presence of oscillations in off-limb polar regions and in coronal holes. Phase delays were measured using the technique of Athay and White (1979), in which phase delays are plotted over the full ­180 to +180 range and as a function of frequency. An example of the result of this is shown in Figure. 5 from O'Shea, Banerjee, and Doyle (2006). In this figure the combined phase delays measured between different line pairs, e.g.,, between O v and Mg x, are shown. The results shown here are from a number of observations in the northern off-limb polar region, combined to obtain a more statistically significant overview of the processes at work in the Sun's atmosphere. From Figure. 5, it can be seen that the measured oscillations are present over the frequency range of 0 ­ 8 mHz and that the phases line up along roughly straight lines (there is a large scatter in the points around these `straight' lines). This distribution of phases along straight lines indicates the presence of outwardly propagating waves. Measuring the slope of these lines allows one to obtain the time delays between the different lines, based on the phase equation; = 2 f T (1) where f is the frequency and T is the time delay in seconds. From this equation it can be seen that the phase difference will vary linearly with f , and will change by

Rev_sol_press_final.tex; 9/08/2007; 19:07; p.11


Banerjee et al.

Figure 4. Histograms of the combined oscillation frequencies, from the primary and secondary oscillations, obtained from the intensity (top row) and velocity (bottom row) time series of Fe xvi 333å (left panel), Mg ix 368å (middle panel), and O v 629å (right panel) (From O'Shea et al. 2001).

Rev_sol_press_final.tex; 9/08/2007; 19:07; p.12


Observing Trends in Atmospheric Magneto-seismology

Figure 5. Phase delays measured between the oscillations in the different line pairs, as labeled, e.g.,, between O v and Mg x 624 (left panel). Phase delays from radiant flux oscillations are shown as the black circle symbols, while phase delays from L.O.S. velocity oscillations are shown as the grey circle symbols. Phase delays were measured at the 95% and 99% significance levels. Phase delays at the 99% significance level are indicated by the slightly larger symbols. Average uncertainties in the 95% and 99% phase delay estimates are shown by the representative error bars in each plot. Over-plotted on this plot are lines corresponding to fixed time delays (From O'Shea, Banerjee, and Doyle 2006).

360 over frequency intervals of f =1/T . In the case of Figure. 5, the time delay measured between the O v 629 line and Mg x 624 line (the first plot on the left) was found to be 58±7 seconds (17 mHz). Using the measured time delays, in conjunction with height differences measured between the different lines using limb brightening measurements, O'Shea, Banerjee, and Doyle (2006) calculated propagation speeds of 154±18 km s-1 between the O v 629 and Mg x 624 lines, 218±28 km s-1 between the O v 629 and Si xii 520 lines, and 236±19 and 201±17 km s-1 between the Mg x 609 and Si xii 520 and Mg x 624 and Si xii 520 line pairs, respectively. These speeds suggest the presence of slow magnetoacoustic waves in these off-limb locations as being the waves responsible for the observed oscillations. From a study of flux-velocity (I-V) phase plots, O'Shea, Banerjee, and Doyle (2006) found evidence for more transverse-like waves to be present at coronal temperatures while at TR temperatures more longitudinal-like waves were present. They attributed the presence of these more transverse-like waves to be due to fast magnetoacoustic waves, while the more transverse-like were due to slow magnetoacoustic waves. It is not clear if fast magnetoacoustic waves are present. In this context, we would also like to point out the possibility of spicules, in the form of obscuration, having an effect on the measurement of intensity-velocity phase measurements. The concern is that this obscuration could be causing a false

Rev_sol_press_final.tex; 9/08/2007; 19:07; p.13


Banerjee et al.

periodicity and obscuring the actual periodicity. But one should note that spicules do not pro ject more than 10 above the limb on average (see Xia et al., 2005) essentially ruling them out as affecting, substantially, the off-limb results of O'Shea, Banerjee, and Doyle (2006). This is due to the fact that the O v line used (the line that could be directly affected by spicules) is measured out to a height of 50 above the limb where spicules cannot affect its periodicity, while the coronal lines are being measured out to something like 200 above the limb. Even if we assume that spicules are affecting the results at lower altitudes, the fact that the results as presented in O'Shea, Banerjee, and Doyle (2006) are a combination of seven datasets and contain results from all heights, up to 50 above the limb (for O v) and up to 200 for the coronal lines, will tend to reduce the possible effects of these obscuration on the overall results. Any effects from obscuration will essentially be `drowned out' in the large amount of `real' data. In O'Shea, Banerjee, and Doyle (2006), I-V phase measurements found a 180 degree phase difference between I and V for the transition region O V line, but typically a 0 degree phase difference for the coronal lines. From Xia et al., (2005), there is no mention of the velocities measured from the spicules being in any way correlated with the radiance measurements. The fact that O'Shea, Banerjee, and Doyle (2006) see strong correlations between I-V in their statistical results would suggest that the essential nature of what they have reported is not due to spicules but to propagating waves. In a similar work, O'Shea, Banerjee, and Doyle (2007), using the same techniques as O'Shea, Banerjee, and Doyle (2006), found evidence for slow magnetoacoustic waves in equatorial and polar coronal holes. In that work, however, the propagation speeds found were substantially lower than those found off-limb, perhaps related to the presence of a more complicated magnetic geometry in the coronal holes. Again, by examining the I-V phase delays, they found that there was a difference in the distribution of the I-V phases between transition and coronal lines; the transition region line of O v showing phases at ­180 and +180 not present in the coronal lines. This again suggests a change in the ma jority of the waves between the transition region and the corona. They also see an indication for the presence of standing waves at the coronal temperatures of Mg x and Si xii, due to the presence of significant peaks at ­90 and +90 in their phase histograms. The presence of standing waves fits nicely with their discovery that the measured phase delays between line pairs occur at fixed phase intervals of 90 and 135 which, as in O'Shea, Banerjee, and Doyle (2006), were linked with some form of resonant cavity effect on the waves. In these types of off-limb studies a significant concern is the effect pro jection effects may have on any comparison between those propagation speeds measured off-limb and those measured in coronal holes? The effect of the pro jection angle is essentially unquantifiable as the angle of the magnetic fields in which one measures the propagating waves is unknown in both regions. However, one can note that the waves one measures at the poles are essentially propagating at 90 to our line-ofsight (LOS), but being compressional longitudinal (slow) waves are still completely measurable in intensity at this angle. From these measured intensities in lines at different temperatures one can obtain the propagation speeds (as in O'Shea, Banerjee, and Doyle, 2006). One can therefore assume that these speeds measured off-limb are essentially true speeds unaffected by pro jection effects, propagating as they are at almost 90 to the LOS. Those waves measured on-disk in coronal holes, however, are propagating at angles between 0 and 90 , and therefore, will have a propagation speed reduced by the effect of this pro jection angle relative to the LOS. For example, an

Rev_sol_press_final.tex; 9/08/2007; 19:07; p.14


Observing Trends in Atmospheric Magneto-seismology

angle of 60 relative to the LOS would result in the propagation speed being reduced by a factor of two relative to its true speed. So one should keep these facts in mind when interpreting the quoted wave speeds. 3.4. EIT waves One of the earliest observations of global waves known, is that of the chromospheric Moreton waves (Moreton and Ramsey 1960). It was seen in the wings of H , propagating in the hot chromosphere, with speeds of 400 ­ 2000 km s-1 . Based on their propagation characteristics, Moreton waves are interpreted as fast shock waves. Further unambiguous evidence for large-scale coronal propagating disturbances initiated during the early stages of a flare and/or CME, has been provided by recent EUV Imaging Telescope (EIT) observations onboard SOHO in the 195 å bandpass. Thompson et al., (1999) reported first on these phenomena based on their SOHO0EIT observations. Although this instrument has a relatively poor temporal and spatial resolution, there are already more than 200 wavelike events found (Klassen et al. 2000; Biesecker et al. 2002). Since these global waves were first seen by the SOHO-EIT instrument, they were labeled as "EIT waves". EIT waves have circular or arc-shaped fronts of enhanced emission and are generated in or near an active region. An interesting event was observed on 4 November, 1997 (Eto et al. 2002), at the time of an intense flare (X2.1 in the NOAA/GOES classification). A Moreton wave was observed in H - + 0.8 å and H - ­ 0.8 å with the Flare-Monitoring Telescope (FMT) at the Hida Observatory. At the same time, an EIT wave was observed in EUV with EIT. There is an ongoing debate about whether the EIT waves are a coronal counterpart of Moreton waves or not. EIT and Moreton waves are sufficiently different and some have theorized that they are two entirely different populations, which originate from different sources (Eto et al. 2002). Moreton waves are strongly-defined, narrow, semi-circular fronts, while EIT waves are broad ( 100 Mm), extremely diffuse, and usually produce circular wave fronts. Moreton waves have relatively short lifetimes (usually < ten minutes), and have shown cospatial observational signatures between the chromosphere and the soft x-ray corona (Khan and Aurass 2002). EIT waves are primarily visible in the lower corona (at one -two MK), but typically have lifetimes of over an hour and can travel the entire diameter of the Sun while remaining coherent. It appears that there should be two types of wave phenomena in the corona during an eruption, a fast-moving wave which is the coronal counterpart of the H - Moreton wave (or the coronal Moreton wave), and a slower moving one which is the EIT wave, with diffuse fronts. SOHO-EIT may catch several wave fronts and at most one front of the coronal Moreton wave in one event if the coronal Moreton wave is moving very fast. We should also point out that although Moreton waves are always viewed in conjunction with EIT waves, the converse is not true, even in high-cadence data. Opinions are divided on the nature of these global waves between a number of different interpretations (e.g.,, fast magnetohydrodynamic waves, shock waves, nonwave feature, etc.). These global waves originate from impulsive and/or eruptive sources such as flares or coronal mass ejections (CMEs) and are able to travel over very long distances. Sometimes these distances are comparable to the solar radius. It has been proposed, for example, that (i) EIT waves are different entities from Moreton waves, and that (ii) X-ray waves as detected by Yohkoh-SXT are, instead,

Rev_sol_press_final.tex; 9/08/2007; 19:07; p.15


Banerjee et al.

the coronal counterpart of Moreton waves, therefore signifying fast mode MHD waves as predicted by Uchida, Altschuler, and Newkirk (1973). There are also many events in which a sharp EUV wave front is seen to be co-spatial with a soft X-ray (SXR) wave front, the latter exhibiting the characteristics of coronal fast-mode waves (Khan and Aurass 2002). These results tend to favor the coronal fast-mode wave model for EIT waves. Observations show that an EIT wave has two stages: first, there is an early (driven) stage where the wave is correlated with a radio II type burst. This correlation can be attributed to the fact that in the initial stage the propagating wave can excite plasma radiation, accelerate electrons, and create an energized population which serves as the source of the radio emission. The second stage consists of a freely-propagating wavefront. Harra and Sterling (2003) investigated an EIT wave jointly seen by TRACE and CDS/SOHO (JOP70). They concluded that EIT waves consist of a faster propagating, piston-driven portion and a more slowly-propagating portion due to the opening of the field lines associated with an erupting filament. They found that these slowly-propagating waves later interact with coronal loops forcing them to oscillate. Wills-Davey and Thompson (1999) examined observations that show the first evidence of a coronal wave event seen by TRACE. They concluded that the observed disturbance behaves more like a fast-mode magnetoacoustic wave. Their observations support Uchida's (1968) model of the propagation of an AlfvÈnic wave in a medium of non-uniform magnetic field. Wills-Davey (2006) has recently developed mapping algorithms that allow automated tracking of a propagating coronal wave, enabling the finding of reproducible fronts and propagation tra jectories. The debate on the nature of EIT waves seems to have widened now. While studying the same event simultaneously with different EUV instruments, Wills-Davey, DeForest, and Stenflo (2007) have concluded that fast MHD compressional waves do not properly describe the dynamics of many EIT wave events. The physical properties of EIT waves, their single-pulse, stable morphology, the non-linearity of their density perturbations and the lack of a single representative velocity instead suggest that they may be best explained as a soliton-like phenomena. Given their propagation characteristics and ability to convey information about the medium in which they propagate, global EIT waves, if their mode physics is finally identified properly, could be used as an excellent tool for global coronal seismology. Ballai, ErdÈlyi and PintÈr (2005) studied TRACE-EUV data to show that these global coronal disturbances are indeed waves with a well-defined period. They showed that EIT waves interact with the coronal loops, and as a result coronal loops begin to oscillate. These induced oscillations are considered to be fast standing kink modes, in good agreement with the theory developed by Roberts, Edwin, and Benz (1984). Ballai, ErdÈlyi and PintÈr (2005) further conjectured that one possible explanation for the different behavior of the same event seen in two wavelengths is that the waves seen in 195 å (EIT) are just some residuals of a rapid wave propagating in a much denser plasma (e.g., at the chromospheric level in the form of shock waves): these residuals being very similar to a bow wave. The more energetic the wave propagating in the chromosphere is, the larger the amplitude that the EIT waves generate. It is possible that small events do not produce large enough waves in the chromosphere to be detected in the low corona. This would explain the relatively small number of EIT waves seen compared to the flaring frequency.

Rev_sol_press_final.tex; 9/08/2007; 19:07; p.16


Observing Trends in Atmospheric Magneto-seismology

Figure 6. The non-thermal velocity as derived from Si viii SUMER observations, using Ti = 1 106 K. The dashed line is a second order polynomial fits, while the (+) symbols correspond to theoretical values (From Banerjee et al., 1998).

3.5. Detection of Waves from Variation of Line Width Study So far, it has been mentioned that waves may be detected using the oscillatory signatures they impose on the plasma (density changes, plasma motions). Another method of detecting waves is to examine the variation they produce in line widths measured from spectral lines. The measured broadening of the optically thin spectral lines of ions is due to two effects, thermal broadening and non-thermal broadening associated with Doppler shifts due to unresolved line-of-sight motions Tef f = Ti + mi 2 < vL 2k
OS

>

where Ti is the ion temperature, k is the Boltzmann constant, vLOS is the line-ofsight component of the velocity, and, 2/3 1. Propagating waves may even be detected through spectral-line broadenings, if concurrently more than one spectral slit is pointing at the same magnetic waveguide, e.g., a coronal loop, and sampling distinct regions of the waveguide. Let us suppose, that there is a coronal loop at the center of the solar disk and one spectrograph samples the footpoint, as a function of time, while the other samples the apex of the same loop, again as a function of time. Let us suppose there is, e.g., a longitudinal wave excited at the footpoint of the waveguide that will propagate along the magnetic structure. Since the motion is longitudinal, and the first spectrograph points exactly in the direction of propagation, it will detect an increased line broadening with time as long as the wave passes

Rev_sol_press_final.tex; 9/08/2007; 19:07; p.17


Banerjee et al. Table 2. (a) Skylab; (b) SOHO. Source Mariska, Feldman, and Doschek (1979) Hassler et al. (1990) Ofman and Davila (1997) Erd´lyi et al. (1998) e Doyle et al. (1997) Doyle, Banerjee, and Perez (1998) Banerjee et al., (1998) Chae, SchÝhle, and Lemaire (1998) Esser et al. (1999) Moran (2003) Instrument SO82-Ba Sounding Rocket UVCSb SUMERb HRTS SUMERb SUMERb SUMERb UVCSb SUMERb LOS (km/s) 33 20 ­ 25 300 1 ­ 100 19 ­ 27 24 ­ 28 27 ­ 46 20 ­ 30 20 ­ 23 40 ­ 60 Location white limb+12" 1 ­ 1.2 R 1.7R limb quiet sun limb to 25Mm limb+20 ­ 180 Mm disc 1.35 ­ 2.1 R 1.02 ­ 1.3 R

through the slit, in spite of it not being able to actually resolve the wave. The second spectrograph sampling the part of the loop below the apex, as well as the apex itself, will also observe a line broadening with time in the part of the loop below the apex that will, however, decrease at the apex, since the wave perturbation there will be perpendicular to the LOS. Measuring the time difference between the first increase in the line width in the footpoint of the loop and the first decrease at the apex could give information about the average longitudinal wave speed. Unfortunately we are not aware of any experiment that has explored the above described opportunity offered by line-broadening, perhaps due to the practical difficulty involved in arranging for two independent and complementary (spectrally) spectrographs to point at different parts of the same solar structure at the same time. Instead, a popular observational sequence is to point the slit, e.g., at the apex of the loop, and let the Sun rotate the loop so that the slit scans from apex to footpoint. If the loop supports the presence of e.g., longitudinal waves, one would find a systematic line broadening from apex to limb. On the other hand, if the loop supported the presence of transversal (e.g., kink) motion, one would find line narrowing. Although this technique, often referred to as centre-to-limb variation in the literature, does not allow one to deduce the propagation velocity of the observed wave, it may give information about the polarisation of the wave, and, of course, about the rms velocity amplitude. Two studies of this type were carried out by ErdÈlyi et al., (1998) and Doyle, Teriaca, and Banerjee (2000). We should also point out here that at this moment it is still very difficult, if not impossible, to resolve individual loops spectroscopically, but perhaps, using the high-resolution EIS on HINODE together with CDS or instruments on the upcoming Solar Orbiter, individual loops will in the future be able to be resolved and these ideas tested. Table 2. summarizes some results and indicates that either slow MHD waves (i.e. mainly longitudinal wave propagation) or AlfvÈn waves (waves that travel along the field lines but are perpendicularly polarised to them) are detected. Harrison, Hood, and Pike (2002) examined the Mg x 625 å line ( 1 â 106 K) in equatorial regions using the SOHO-CDS instrument on. Their most significant result was the discovery of emission line narrowing as a function of altitude and intensity above 50 000 km.

Rev_sol_press_final.tex; 9/08/2007; 19:07; p.18


Observing Trends in Atmospheric Magneto-seismology

All earlier observations of emission-line broadening with increasing altitude are consistent with the propagation of linear-undamped AlfvÈn waves in open field regions with decreasing density. Harrison, Hood, and Pike (2002) attributed the narrowing as being due to the dissipation of AlfvÈn waves in the corona. One should remember that there is a fundamental difference in the properties of wave propagation in the equatorial corona (closed field regions) when compared to coronal holes (open field regions). Thus it is important to see if one can also observe this narrowing of coronal lines in the coronal hole regions. Both Banerjee et al., (1998) and Doyle, Teriaca, and Banerjee (1999) studied Si viii line profiles with SUMER in the off-limb northern polar hole regions. They recorded line broadening up to 110 000 km (150 off-limb) and then a levelling off in the line widths up to 220 000 km (see Figure. 6), after which there was a faint hint of a fall off in the widths, although this last observation was inconclusive due to uncertainties in the data. O'Shea et al. (2003) measured the variation of Mg x 624 line widths (from CDS) above the north polar limb and found that there was an initial linear increase with altitude, supporting the interpretation of linear undamped AlfvÈn waves propagating outward in open-field regions. Also noted in these results was a turnover point, at a particular altitude, where the line widths suddenly decreased or levelled off. This decrease in the line widths at a particular height is consistent with a dissipation of the AlfvÈn wave energy. In a follow-up paper, O'Shea, Banerjee and Doyle (2005), measuring the line widths of the Mg x 609 and 624 lines from CDS, again found evidence for a decrease in the line widths above a certain height off-limb (cf. Fig. 7). This was again attributed to damping of upwardly propagating MHD waves. In addition, O'Shea, Banerjee and Doyle (2005) measured the ratio of the two Mg x lines as a function of radial distance above the limb. They found that this ratio changed from values expected for a collisionally dominated plasma to one expected from a radiatively- dominant plasma as the same approximate height that the decrease in line width occurred. This suggest that the decrease in the line widths, (the damping of the waves,) may be linked to this change in the dominant excitation, perhaps due to decreases in the electron density. We note in passing that Doyle et al., (2005) have found evidence for some broadening above the limb to be due to spicules. Areas where spicules were absent were found to have lower line widths suggesting that spicules have some part to play in line broadening at least close to the limb. It is very important to understand the mechanism of line- width decrease as it may trigger the acceleration of plasma particles in these regions. In polar coronal holes, where the magnetic field is open and predominantly vertical, AlfvÈn waves mainly contribute to the off-limb line broadening due to their transverse velocity polarisation. Acoustic waves propagating along the magnetic field are unlikely to contribute to the line broadening because their velocity polarisation is predominantly perpendicular to the line of sight. The decrease of the line width in polar coronal holes can then be explained either by the AlfvÈn wave damping or due to the conversion into acoustic waves. Recently Zaqarashvili1, Oliver, and Ballester (2006) have shown that the resonant energy conversion from AlfvÈn to sound waves near the region where the plasma approaches unity (or more precisely, where the ratio of sound to AlfvÈn speeds approaches unity) may explain the observed sudden decrease of the spectral line width in the solar corona.

Rev_sol_press_final.tex; 9/08/2007; 19:07; p.19


Banerjee et al.

Figure 7. Variation of the Doppler width (uncorrected for instrumental width contributions) versus radial distance for the 26478/26479 and 26542/26543 datasets, as indicated by the numbers shown in each plot. The thick black lines show the result of a box-car averaging. Radial distance locations where the radiance fell below a critical S/N value do not show the results of the line width measurements (From O'Shea, Banerjee and Doyle 2005).

4. Observations of Waves and Oscillations in Prominences The solar corona is populated by dense clouds of cold plasma inexplicably floating tens of thousands of kilometres above the photosphere. Such features are routinely seen during solar eclipses, when they can be easily distinguished by their red glow, but they can also be unveiled with the help of filters, such as H , that have been devised to observe the chromosphere. In contrast with the MK temperature of the surrounding corona, prominences remain at a comparatively cool 10 000 K, which prompts one to ask what prevents the mechanisms that heat the corona from also raising the temperature of prominences. The mechanism of their formation and disappearance is not very clearly understood. Firstly, one might wonder why they form in such an environment. Secondly, despite their internal dynamics, prominences that have been stable for weeks suddenly disappear in a spectacular eruption. The processes shaping the lifetime of prominences are largely unknown. Nevertheless, the intervention of one decisive element is quite clear: the magnetic field, which plays a central role in all the prominence processes mentioned here. The reason for our limited insight into the nature of prominences probably has three causes (Vial, 1998): there is no such thing as a canonical prominence, but a wide range of parameters is observed in different ob jects; no prominence has a uniform structure, but they are made of thin threads (or fibrils) and, in addition, different parameter values can be detected in different parts of a prominence; and,

Rev_sol_press_final.tex; 9/08/2007; 19:07; p.20


Observing Trends in Atmospheric Magneto-seismology

no structure is really isolated, so it is necessary to understand the physics of the prominence-corona interface, the effect of the coronal radiation field (e.g., Anzer and Heinzel, 2005) and to trace the magnetic fields permeating the prominence to their origin at the Sun's surface (e.g., Lin et al., 2005b). Our knowledge about prominences has been well reviewed by Tandberg-Hanssen (1995), Martin (1998) and Patsourakos and Vial (2002), where more information on the topic can be found. Where does the study of waves and oscillations in prominences stand in the middle of this panorama? It constitutes a discipline that may complement the direct determination of prominence parameters by providing independent values based on the comparison between observations of oscillations and theory. However, this is more a promise than a reality because of the large gap between observation and theory. Such a gap arises because of the few restrictions imposed by observations (which are sometimes reduced to reporting the period of the detected oscillations) and the simplicity of theoretical studies (which neglect the intricate nature of prominences and represent it with a very simplified physical model). Nevertheless, these two sides are coming together as the complexity of different works increases. Previous advances, both observational and theoretical, have been examined by Oliver and Ballester (2002), Engvold (2004), Wiehr (2004), and Ballester (2006), so it is our purpose here only to review the observational facts of prominence oscillations with a special emphasis on the last few years. ErdÈlyi, Ballester, and Ruderman (2007) should also be considered for a review of the theory. 4.1. Instrumental Setup and Data Analysis Most observational work on prominence oscillations are based on Doppler velocity data acquired with a spectrograph slit. This, in principle, allows one to determine wave properties along the slit (as in Molowny-Horas et al., 1997), but nothing can be said about the propagation properties perpendicular to the slit. As we describe in Sections. 4.5 and 4.6, only on a few occasions has this simple setup been replaced by a two-dimensional one, which obviously results in a much deeper insight into the nature of waves and oscillations. On the other hand, the data analysis has usually been restricted to the computation of power spectra, while other techniques have been rarely used. The advantages of wavelet analysis have been exploited by Molowny-Horas et al., (1997) Bocchialini et al., (2001) and Foullon, Verwichte, and Nakariakov (2004). However, more complex tools have never been used in this sub ject. 4.2. Spectral Indicators Apart from the Doppler velocity, some other spectral indicators have also been used in the search for periodic variations in prominences, and sometimes a periodic signal has been recognised in more than one of these indicators. Landman, Edberg, and Laney (1977) observed periodic fluctuations in the line intensity and width with a period of around 22 minutes, but not in the Doppler shift. In addition, Yi, Engvold, and Keil (1991) detected periods of 5 and 12 minutes in the power spectra of the line-of-sight velocity and the line intensity. Also, Suematsu et al., (1990) found signs of a 60 minute periodic variation in the Doppler velocity, line intensity and line width. Nevertheless, the Doppler signal also displayed shorter period variations (with periods around 4 and 14 minutes) which were not present in the other two data

Rev_sol_press_final.tex; 9/08/2007; 19:07; p.21


Banerjee et al.

sets. We encounter here a perplexing feature of prominence investigations, namely that the temporal behaviour of various indicators corresponding to the same time series of spectra do not agree, either because they show different periods in their power spectra (as in Tsubaki, Ohnishi, and Suematsu 1987) or because one indicator presents a clear periodicity while the others do not (Wiehr et al., 1984; Tsubaki and Takeuchi, 1986; Balthasar et al., 1986; Tsubaki et al., 1988; SÝtterlin et al., 1997). Special mention must be made of the study performed by Balthasar and Wiehr (1994), who simultaneously observed the spectral lines He 3888 å, H8 3889 å , and the Ca+ IR3 8498 å. From this information they analysed the temporal variations of the thermal and non-thermal line broadenings, the total H8 line intensity, the He 3888 å to H8 emission ratio, and the Doppler shift of the three spectral lines, which correlated well and thus reduced to a single data set. The power spectra of all these parameters yielded a large number of power maxima, but only two of them (with periods of 29 and 78 minutes) are present in more than one indicator. The interpretation of the results just summarised appears difficult. First, the theoretical models predict the temporal behaviour of the plasma velocity, and sometimes the density and other physical parameters, in a prominence. The observations, however, yield information on quantities such as the line intensity or the line width. Hence, a clear understanding of the relationship between the spectral parameters and a number of different physical variables (density, pressure, temperature, etc.) is required before any progress can be achieved. Then, perhaps, the presence of a certain period in one or a few signals could be used to infer the properties of the MHD mode involved. 4.3. Periods Periodic variations have been detected in a variety of configurations (in prominences, in filaments, in threads) spanning a range from less than a minute to about 90 minutes, and there is even a reported value around 12 hours. Some of these results are summarised in Table 3, which is by no means exhaustive (see also Oliver and Ballester, 2002). Unfortunately, on its own, a period reveals very little about the conditions in the prominence since it may correspond to almost infinite combinations of density, temperature, magnetic field strength, etc. The detection of oscillations with a period near 12 hours in the intensity of the 195 å line by Foullon, Verwichte, and Nakariakov (2004) is remarkable. These authors exploited the long-term stability provided by a space telescope (in this case SOHO-EIT) to obtain an almost uninterrupted data series lasting 260 hours. Although this is not the first time that a telescope onboard SOHO has been used in connection with prominence oscillations (see Bocchialini et al.,, 2001; RÈgnier, Solomon, and Vial, 2001), it is the only study to uncover such very long period perturbations. Foullon, Verwichte, and Nakariakov (2004), however, discarded an important part of the information in their data by spatially averaging the line intensity and so notably reducing the possibilities of their analysis, as have many other authors. 4.4. Wave Damping While it has been clear for a long time that prominence disturbances last for only a few periods (Oliver and Ballester, 2002), this phenomenon has seldom been quantified. We must emphasise the importance of observing and identifying the process of

Rev_sol_press_final.tex; 9/08/2007; 19:07; p.22


Observing Trends in Atmospheric Magneto-seismology Table 3. Summary of observations of small amplitude prominence oscillations: reported periods and structures in which observations were carried out. Reference Harvey (1969) Bashkirtsev and Mashnich (1984) Tsubaki and Takeuchi (1986) Yi, Engvold, and Keil (1991) Balthasar et al., (1993) Bashkirtsev and Mashnich (1993) SÝtterlin et al. (1997) Terradas et al. (2002) Foullon, Verwichte, and Nakariakov (2004) Lin (2004) Lin et al., (2007) Period (min) 1 ­ 17 42 ­ 82 2.7, 3.5 5, 9, 16 0.5, 12, 20 5 ­ 90 3 ­ 10, 20, 60 30 ­4 0, 75 720 4 ­ 20, 26, 42, 78 3­9 Structure Prominence Prominence Prominence Thread Prominence Prominence/Filament Prominence Prominence Filament Thread Thread

wave damping because this is a process where theory and observations can meet (see ErdÈlyi, Ballester, and Ruderman, 2007 for a discussion of the theory of this topic). Terradas et al. (2002) studied wave motions in a two-dimensional field of view and detected a propagating perturbation which is damped both in space and time. The temporal damping is such that it can be well fitted by an exponential function and, depending on the position in the prominence, varies between two and three times the period. Even though these results are not too restrictive from the theoretical point of view, they are unique in their kind and so similar efforts should be undertaken in the future. 4.5. Wave Propagation Terradas et al. (2002) also provide us with a solid investigation of wave propagation in a prominence. The damped oscillations just described originate in a narrow strip of 3000 km â 10 000 km and then spread out away from this region, which is near the prominence edge and parallel to it. Waves propagate over an area of some 54 000 â 40 000 km. These authors found that wave propagation is quite anisotropic and mostly in the directions parallel and perpendicular to the prominence edge. Terradas et al. (2002) determined the two-dimensional distribution of the phase velocity and obtained values between 10 and 20 km s-1 , where the highest and lowest values take place in the parallel and perpendicular directions, respectively. Once more this is the only work in which the two-dimensional distribution of oscillations is studied (except for the papers described in the next section) so these results must also be confirmed in the future. 4.6. Fibril Structure Solar prominences are formed by many thin, parallel magnetic threads filled with cold plasma, and as a consequence the dynamics of these components can be easier to understand than that of the whole ob ject. Early works (Yi, Engvold, and Keil, 1991; Yi and Engvold, 1991) already noted the possible link between prominence

Rev_sol_press_final.tex; 9/08/2007; 19:07; p.23


Banerjee et al.

oscillations and the fibril structure. Unfortunately, the spatial resolution of the data analysed by Terradas et al. (2002) was not good enough to resolve the prominence threads. It was necessary, hence, to wait until the advent of telescopes with much better spatial resolution to have observations in which the prominence fine structure is well resolved (Lin, Engvold, and Wiik, 2003; Lin, 2004). In the analysis of the Doppler velocity in two threads belonging to the same filament, Lin (2004) found a clear sign of propagating waves and determines their period, wavelength and phase speed. This study was followed by a much more profound one in which the two-dimensional motions and Doppler shifts of 328 features of different threads were examined. These features are observed to flow along the filament axis while oscillating at the same time. To simplify the evaluation of the oscillations, Lin (2004) computed average Doppler signals for each fibril and found that groups of adjacent threads oscillate in phase. This has two consequences: first, since the periodicity is clear in the averaged signal for each thread, the wavelength of oscillations is larger than the length of the thread. Second, fibrils have a tendency to vibrate bodily, in groups, rather than independently, an issue that has been investigated theoretically (DÌaz, Oliver, and Ballester, 2005). H observations conducted with the Swedish 1-m Solar Telescope by Lin et al., (2007) lead to similar results concerning the collective dynamics of fibrils, although propagating Doppler velocity signals with various periods and wavelengths in other threads of the same filament are also detected. All these observations seem to indicate that prominence fibrils sometimes support collective oscillations and sometimes oscillate on their own. This topic deserves a more detailed observational study and, given the simplicity of fibrils compared to the full filament structure, a theoretical investigation could give rise to a fruitful comparison with observations. 4.7. Large Amplitude Prominence Oscillations All the results described above correspond to waves and oscillations with comparatively small amplitudes, i.e. with Doppler velocity peaks typically below one ­ two km s-1 . Nevertheless, prominence oscillations of much larger amplitudes (with oscillatory velocities up to 90 km s-1 ) have also been observed. These large-amplitude oscillations normally occur after an energetic, explosive event disturbs the whole prominence (see the movie in Jing et al.,, 2003 for an illustrative example). The topic of large- amplitude prominence oscillations has remained practically dormant for more than thirty years until its recent revival (see Oliver and Ballester, 2002 for a review of older results). Because of the large velocities involved, large-amplitude oscillations sometimes substantially modify the absorption/emission wavelength of the prominence material. For example, Eto et al., (2002) and Okamoto et al., (2004) observed two filaments as they underwent one of these episodes and could detect their absorption in H ± 0.8 å filters. This indicates that the velocity of the plasma is in excess of 20 km s-1 . During the event studied by Eto et al., (2002) the filament disappeared from the H -line- centre image when the velocity was at its maximum. Such behaviour can periodically repeat in time if the oscillation lasts for a few periods with sufficiently high amplitude, such as observed by Ramsey and Smith (1966). The periods of these oscillations exceed the most common values of small-amplitude oscillations and range from 30 minutes to almost three hours (Isobe and Tripathi, 2006; Jing et al., 2006). In addition, these oscillations are commonly damped with

Rev_sol_press_final.tex; 9/08/2007; 19:07; p.24


Observing Trends in Atmospheric Magneto-seismology

damping times which (as in the case of small-amplitude oscillations) are two ­ three times the corresponding period (for a few examples see Jing et al., 2006). This prompts us to question whether the mechanisms involved in the attenuation of the two types of prominence oscillations are the same. 4.8. Future Directions Some hints for the future development of prominence seismology have been given here. It is particularly important to carry out observations with high spatial and temporal resolution, such as those in Lin et al., (2005a); Lin et al., (2007) using the Swedish Solar Telescope in La Palma. The purpose of this kind of investigation is to characterise the temporal properties of thread oscillations (with a particular emphasis on their excitation and damping) as well as their spatial properties (with a particular emphasis on their collective or individual behaviour). Prominence threads are promising research ob jects because their dynamics can be treated theoretically or numerically using simple models. A second topic that requires some development is the translation of detected variations of the line intensity and line width into variations of physical variables. This probably requires some multilevel non-LTE transfer modelling (e.g., Heinzel and GunÀr , 2005). Finally, observations of prominences from space have the advantage of very good stability and long duration, but they have been seldom used for the study of prominence oscillations. Space instruments do not enjoy the spatial resolution of the best terrestrial telescopes but they should nevertheless be exploited in the future.

5. Concluding Remarks A systematic detection of waves in a wide range of magnetic structures in the corona is now possible with the range of modern space- and ground-based instruments available with a sufficient spatial and temporal resolution, although there are some limitations as outlined in this review. With future observing facilities the field of solar atmospheric seismology is likely to yield better and more accurate diagnostic results. Comparing data from new multi-wavelength observations to MHD wave theories and numerical simulations means the field of atmospheric seismology is going to be a very exciting and interesting future field of study. In this review we have summarized the current trends in the observational study of MHD waves in open (coronal holes) and closed (loops and prominences) solar atmospheric magnetic structures with a little more emphasis on the types of spectral signatures expected for the detection of different types of waves. With strong evidence of fast and slow magnetoacoustic modes arising in the solar atmosphere there is scope for an improvement in determining coronal parameters through atmospheric magneto-seismology. For example the ratio P1 /2P2 , in an homogeneous medium is unity, where P1 is the fundamental mode and P2 is the first harmonic of the standing transverse kink mode. But in a more complex configuration it can be shifted to lower values. Andries et al. (2005), Goossens et al. (2006) and ErdÈlyi and Verth (2007) have pointed out that the identification of harmonics could provide important diagnostic information for the coronal seismology of a loop. McEwan et al. (2006b) have studied how the ratio P1 /2P2 deviates from unity for fast and slow MHD modes in response to such effects as structuring in the longitudinal

Rev_sol_press_final.tex; 9/08/2007; 19:07; p.25


Banerjee et al.

or transversal directions or gravity. They concluded that longitudinal structuring is the most important effect and this can be used in coronal seismology to estimate properties such as the density stratification scale. The future of atmospheric seismology looks bright. The recent launch of the HINODE satellite, containing the X-Ray Telescope (XRT) and the EUV Imaging Spectrometer (EIS), and the upcoming launch of the Solar Dynamics Observatory (SDO), containing the Atmospheric Imaging Assembly (AIA), means that it will soon be possible to obtain slit and image data at a much increased spectral resolution with excellent time resolution. For example, the AIA, offering a replacement for the TRACE instrument, will allow the Sun to be imaged at ten different wavelengths simultaneously, with a time resolution of ten seconds. The time resolution of EIS, in comparison, can go down to one second (depending on the lines chosen), allowing for the measurement of very high frequency oscillations. The good spectral resolution of EIS will, in addition, allow the accurate measurement of non-thermal velocities and allow other studies that are based on the detailed measurement of line widths, e.g., the variation of line widths off-limb caused by wave dissipation, etc. EIS, moreover, has the advantage over previous instruments and observations, e.g., with TRACE, in that it can obtain time series images of different solar structures (with its wide slits) together with a simultaneous measurement of electron density in these structures. EIS can measure electron densities through the presence of a number of excellent density-sensitive line ratios, at coronal temperatures, within its spectral range. Together then, XRT, EIS, and AIA will allow an unprecedented opportunity to observe MHD waves in many solar structures and together they offer the solar community a great opportunity to significantly progress the still nascent field of atmospheric magento-seismology. Acknowledgements R.E. acknowledges M. KÈray for patient encouragement and is also grateful to NSF, Hungary (OTKA, Ref. No. TO43741). The authors also thank PPARC (UK) for the financial support they received. E.O.S. is supported by PPARC grant PP/D001129/1. R.O. acknowledges the financial support received from the Spanish Ministerio de Ciencia y TecnologÌa under grant AYA2006-07637. D.B. and E.O.S. wishes to thank the Royal Society and the British Council for funding visits between Armagh Observatory and the Indian Institute of Astrophysics.

References
Antonucci, E., Gabriel, A.H., Patchett, B.E.: 1984, Solar Phys. 93, 85. Anzer, U., Heinzel, P.: 2005, Astrophys. J. 622, 714. Andries, J., Goossens, M., Hollweg, J. V., Arregui, I., Van Doorsselaere, T.: 2005, Astron. Astrophys. 430, 1109. Asai, A., Shimo jo, M., Isobe, H., Morimoto, T., Yokoyama, T., Shibasaki, K., Naka jima, H.: 2001, Astrophys. J. 562, L103. Aschwanden, M.J.: 1987, Solar Phys. 111, 113. Aschwanden, M.J.: 2003, In ErdÈlyi, R. et al. (eds.), Turbulence, Waves and Instabilities in the Solar Plasma, NATO Sci Ser. 124, 215. Aschwanden, M.J.: 2004, Physics of the Solar Corona: An Introduction, Springer, Berlin. Aschwanden, M.J.: 2006, Phil. Trans. Roy. Soc. A. 364, 417. Athay, R.G., White, O.R.: 1979, Astrophys. J. 229, 1147. Ballai, I., ErdÈlyi, R., PintÈr, B.: 2005, Astrophys. J. 633, L145. Balthasar, H., Wiehr, E.: 1994, Astron. Astrophys. 286, 639. Balthasar, H., Wiehr, E., Schleicher, H. WÆhl, H.: 1993, Astron. Astrophys., 277, 635.

Rev_sol_press_final.tex; 9/08/2007; 19:07; p.26


Observing Trends in Atmospheric Magneto-seismology B B B B B B B B B B B B B althasar, H., KnÆlker, M., Stellmacher, G., Wiehr, E.: 1986, Astron. Astrophys. 163, 343. allester, J. L.: 2006, Phil. Trans. Royal Soc. A 364, 405. ashkirtsev, V.S., Mashnich, G.P.: 1984, Solar Phys., 91, 93. ashkirtsev, V.S., Mashnich, G.P.: 1993, Astron. Astrophys., 279, 610. anerjee, D., Teriaca, L., Doyle, J. G., Wilhelm, K.: 1998, Astron. Astrophys. 339, 208. anerjee, D., O'Shea, E. Doyle, J.G.: 2000, Solar Phys. 196, 63. anerjee, D., O'Shea, E., Doyle, J.G., Goossens, M.: 2001a, Astron. Astrophys. 377, 691. anerjee, D., O'Shea, E., Doyle, J.G., Goossens, M.: 2001b, Astron. Astrophys. 380, L39. audin, F., Bocchialini, K., Koutchmy, S.: 1996, Astron. Astrophys. 314, L9. erger, T.E., De Pontieu, B., Schrver, C.J., Title, A.M.: 1999, Astrophys J. 519, 97. erghmans, D., Clette, F.: 1999, Solar Phys. 186, 207. erghmans, D., McKenzie, D., Clette, F.: 2001, Astron. Astrophys. 369, 291. iesecker, D.A., Myers, D.C., Thompson, B.J., Hammer, D.M. Vourlidas, A.: 2002, Astrophys J. 569, 1009. Bocchialini, K., Costa, A., Domenech, G., Rovira, M., Vial, J. C., Wingfield, K.: 2001, Solar Phys. 199, 133. Bogdan, T.J.: 2000, Solar Phys. 192, 373. Brynildsen, N., Maltby, P., Fredvik, T., Kjeldseth-Moe, O.: 2002, Solar Phys. 207, 259. Chae, J., SchÝhle, U., Lemaire, P.: 1998, Astrohys. J. 505, 957. Chapman, R.D., Jordan, S.D., Neupert, W.M., Thomas, R.J.: 1972, Astrophys J. 174, 97. DeForest, C.E., Gurman, J.B.: 1998, Astrophys. J. 501, 217. DÌaz, A. J., Oliver, R. Ballester, J. L.: 2005, Astron. Astrophys. 440, 1167. De Moortel, I., Ireland, J., Walsh, R.W.: 2000, Astron. Astrophys. 355, L23. De Moortel, I., Ireland, J., Walsh, R.W., Hood, A.W.: 2002a Solar Phys. 209, 61. De Moortel, I., Ireland, J., Hood, A.W., Walsh, R.W.: 2002b,Solar Phys. 209, 89. De Moortel, I.: 2005, Phil. Trans. Roy. Soc. A. 363, 274. De Moortel, I.: 2006, Phil. Trans. Roy. Soc. A. 364, 461. De Pontieu, B., ErdÈlyi, R., de Wn, A.G.: 2003a, Astrophys. J. 595, 63. De Pontieu, B., Tarbell, T., ErdÈlyi, R.: 2003b, Astrophys. J. 590, 502. De Pontieu, B., ErdÈlyi, R., James, S.P.: 2004, Nature 430, 536. De Pontieu, B.: 2004, in (sci. eds.) R. ErdÈlyi, J.L. Ballester, B. Fleck Waves, oscil lations and smal l scale transient events in the solar atmosphere: A joint view of SOHO and TRACE, ESA-SP 547, 25. De Pontieu, B., ErdÈlyi, R., De Moortel, I.: 2005, Astrophys. J. 624, 61. De Pontieu, B., ErdÈlyi, R.: 2006, Phil. Trans. Roy. Soc. A. 364, 383. Doyle, J. G., O'Shea, E., ErdÈlyi, R., Dere, K. P., Socker, D. G., Keenan, F. P.: 1997, Solar Phys. 173, 243. Doyle, J.G., Banerjee, D., Perez, E.: 1998, Solar Phys. 181, 91. Doyle, J.G., Teriaca, L., Banerjee, D.: 1999, Astron. Astrophys. 349, 956. Doyle, J.G., Teriaca, L., Banerjee, D.: 2000, Astron. Astrophys. 356, 335. Doyle, J.G., DzifÀkovÀ, E., Madjarska, M. S.: 2003, Solar Phys. 218, 79. Doyle, J.G., Giannikakis, J., Xia, L.D., Madjarska, M.S.: 2005, Astron. Astrophys. 431, L17. ErdÈlyi, R., Doyle, J.G., Perez, M.E., Wilhelm, K.: 1998, Astron. Astrophys. 337, 287. ErdÈlyi, R..: 2006a, Phil. Trans. Roy. Soc. A. 364, 351. ErdÈlyi, R.: 2006b, in K. Fletcher (ed) Beyond the Spherical Sun: A New era for Helio- and Asteroseismology, Phil. Trans. Roy. Soc. A. 624, 15.1. ErdÈlyi, R., Ballester, J. L. Ruderman, M. S.: 2007, Solar Phys. submitted. ErdÈlyi, R. & Verth, G.: 2007, Astron. Astrophys. 462, 743. Esser, R., Fineschi, S., Dobrzycka, D., Habbal, S.R., Edgar, R.J., Raymond, J.C., Kohl, J.L., Guhathakurta, M.: 1999, Astrohys. J. 510, 63. Engvold, O.: 2004, in Proc. IAU Col loq. on Multiwavelength investigations of solar activity (eds. A. V. Stepanov, E. E. Benevolenskaya & A. G. Kosovichev), 187. Eto, S. et al.,: 2002, PASJ 54, 481. Finsterle, W., Jefferies, S. M., Cacciani, A., Rapex, P., Giebink, C., Knox, A., Dimartino, V.: 2004, Solar Phys. 220, 317. Foullon, C., Verwichte, E. Nakariakov, V. M.: 2004, Astron. Astrophys. 427, L5. Goossens, M., Andries, J., Arregui, I.: 2006, Phil. Trans. R. Soc. 364, 433. Harra, L.K., Sterling, A.C.: 2003, Astrohys. J. 587, 429. Harrison, R.A.: 1987, Astron. Astrophys. 182, 337. Harrison, R.A. et al.,: 1995, Solar Phys. 162, 233. Harrison, R.A., Hood, A.W., Pike, C.D.: 2002, Astron. Astrophys. 392, 319.

Rev_sol_press_final.tex; 9/08/2007; 19:07; p.27


Banerjee et al. Harvey, J. W.: 1969, Ph. D. Thesis , Univ. of Colorado, 166 ­ 176, 295. Hassler, D. M., Rottman, G. J., Shoub, E. C., Holzer, T. E.: 1990 Astrohys. J. 348, L77. Heinzel, P., Anzer, U., GunÀr, S.: 2005, Astron. Astrophys. 442, 331 Isobe, H., Tripathi, D.: 2006, Astron. Astrophys. 449, L17. Jing, J., Lee, J., Spirock, T.J., Xu, Y., Wang, H., Choe, G.S.: 2003 Astrohys. J. 584, L103. Jing, J., Lee, J., Spirock, T.J., Wang, H.: 2006 Solar Phys. 236, 97. Jefferies, S. M., Osaki, Y., Shibahashi, H., Harvey, J. W., D'Silva, S., Duvall, T. L., Jr.: 1997 Astrohys. J. 485, L49. Katsiyannis, A.C., Williams, D.R., McAteer, R.T.J., Gallagher, P.T., Keenan, F.P., Murtagh, F.: 2003, Astron. Astrophys. 406, 709. Khan, J.I. Aurass, H.: 2002, Astron. Astrophys. 383, 1018. King, D.B., Nakariakov, V.M., Deluca, E.E., Golub, L., McClements, K.G.: 2003, Astron. Astrophys. 404, 1. Klassen, A., Aurass, H., Mann, G. Thompson, B.J.: 2000, Astron. Astrophys. Suppl. series, 141, 357. Koutchmy, S.; Zhugzhda, Ia. D.; Locans, V.: 1983, Astron. Astrophys. 120, 185. Landman, D. A., Edberg, S. J., Laney, C. D. 1977, Astrophys. J., 218, 888 Lin, Y. 2004, Ph. D. Thesis , University of Oslo, Norway Lin, Y., Engvold, O., Wiik, J. E.: 2003, Solar Phys. 216, 109. Lin, Y., Engvold, O., Rouppe van der Voort, L., Wiik, J.E., Berger, T.E.: 2005a, Solar Phys., 226, 239 Lin, Y., Wiik, J. E., Engvold, O., Rouppe van der Voort, L., Frank, Z.A.: 2005b, Solar Phys., 227, 283 Lin, Y., Engvold, O., Rouppe van der Voort, L. H. M., van Noort, M.: 2007, Solar Phys., in press Madjarska, M.S., Doyle, J.G., Teriaca, L., Banerjee, D: 2003, Astron. Astrophys. 398, 775. McEwan, M.P., De Moortel, I.: 2006a, Astron. Astrophys. 448, 763. McEwan, M. P., Donnelly, G. R., Döaz, A. J., Roberts, B.: 2006b, Astron. Astrophys., 460, 893. McIntosh, S. W., Jefferies, S. M.: 2006, ApJ 647, L77. Mariska, J. T., Feldman, U. Doschek, G. A.: 1979, Astron. Astrophys., 73, 361. Martin, S. F.: 1998, Solar Phys. 182, 107. Molowny-Horas, R. et al.,: 1997, Solar Phys. 172, 181. Moran, T.G.: 2003, Astrophys. J. 598, 657. Moreton, G.F. Ramsey, H.E.: 1960, PASP 72, 357. Nakariakov, V.M.: 2006, Phil. Trans. Roy. Soc. A. 364, 473. Nakariakov, V.M., Roberts, B.: 2003, in (eds) R. ErdÈlyi et al., Turbulence, Waves and Instabilities in the Solar Plasma, NATO Sci Ser., 124, 167. Nakariakov, V.M., Verwichte, E.: 2005, Liv. Rev. Sol. Phys., 2, 3. Nightingale, R.W., Aschwanden, M.J., Hurlburt, N.E.: 1999, Solar Phys. 190, 249. Ofman, L., Davila, J.M.: 1997, Astrophys. J. 476, 51. Ofman, L., Romoli, M., Poletto, G., Noci, C., Kohl, J.L.: 1997, Astrophys. J. 491, 111. Ofman, L., Nakariakov, V.M., DeForest, C.E.: 1999, Astrophys. J. 514, 441. Ofman, L., Romoli, M., Poletto, G., Noci, C., Kohl, J.L.: 2000a Astrophys. J. 529, 592. Ofman, L., Nakariakov, V.M., Seghal, N.: 2000b Astrophys. J. 533, 1071. Okamoto et al.,: 2004, Astrophys. J. 608, 1124. Oliver, R., Ballester, J. L.: 2002, Solar Phys. 206, 45. O'Shea, E., Banerjee, D., Doyle, J.G., Fleck, B., Mugtagh, F.: 2001, Astron. Astrophys. 368, 1095. O'Shea, E., Muglach, K. Fleck, B.: 2002, Astron. Astrophys. 387, 642. O'Shea, E., Banerjee, D., Doyle, J.G., Poedts, S.: 2003, Astron. Astrophys. 400, 1065. O'Shea, E., Banerjee, D., Doyle, J.G.: 2005, Astron. Astrophys. 436, L35. O'Shea, E., Banerjee, D., Doyle, J.G.: 2006, Astron. Astrophys., 452, 1059. O'Shea, E., Banerjee, D., Doyle, J.G.: 2007, Astron. Astrophys. 463, 713. Patsourakos, S., Vial, J.-C.: 2002, Solar Phys. 208, 253. Popescu, M.D., Banerjee, D., O'Shea, E., Doyle, J.G. Xia, L.D.: 2005, Astron. Astrophys. 442, 1087 Ramsey, H. E., Smith, S. F.: 1966, Astron. J. 71, 197. RÈgnier, S., Solomon, J., Vial, J.-C., Astron. Astrophys. 376, 292. Robbrecht, E., Verwichte, E., Berghmans, D., Hochedez, J.F., Poedts, S.: 2001, Astron. Astrophys. 370, 591.

Rev_sol_press_final.tex; 9/08/2007; 19:07; p.28


Observing Trends in Atmospheric Magneto-seismology Roberts, B. Edwin, P. M., Benz, A. O.: 1984, Astrophys. J. 279, 957. Sakurai, T., Ichimoto, K., Ra ju, K. P., Singh, J.: 2002, Solar Phys. 209, 265. Schrver, C.J., Title, A.M., Berger, T.E., Fletcher, L., Hulburt, N.E. et al.,: 1999, Solar Phys. 187, 261. Suematsu, Y., Yoshinaga, R., Terao, N. & Tsubaki, T. 1990, Publ. Astron. Soc. Japan, 42, 187. SÝtterlin, P., Wiehr, E., Bianda, M. & KÝveler, G. 1997, Astron. Astrophys., 321, 921. Tandberg-Hanssen, E.: 1995, The nature of solar prominences, Kluwer. Terradas, J. et al.,: 2002, Astron. Astrophys. 393, 637. Thompson, B. J., Plunkett, S. P., Gurman, J. B., Newmark, J. S., St. Cyr, O. C., Michels, D. J.: 1998, Geophys. Res. Lett. 25, 2465. Thompson, B.J., Gurman, J.B., Neupert, W.M. et al.,: 1999, Astron. J. 517, 151. Tsubaki, T., Takeuchi, A. 1986, Sol. Phys., 104, 313 Tsubaki, T., Ohnishi, Y., Suematsu, Y. 1987, Publ. Astron. Soc. Japan, 39, 179 Tsubaki, T., Toyoda, M., Suematsu, Y., Gamboa, G. A. R. 1988, Publ. Astron. Soc. Japan, 40, 121 Uchida, Y.: 1968, Solar Phys. 4, 30. Uchida, Y., Altschuler, M. D., Newkirk, G.: 1973, Solar Phys. 28, 495. Ugarte-Urra, I., Doyle, J.G., Madjarska, M.S., O'Shea, E., D: 2003, Astron. Astrophys. 418, 313. Verwichte, E., Nakariakov, V.M., Cooper, F.C.: 2005, Astron. Astrophys. 430, L65. Vial, J.-C.: 1998, in New perspectives on solar prominences. IAU Col loq. 167 (eds. D. Webb, D. Rust, B. Schmieder), ASP Conf. Series 150, 175. Wang, T.J.: 2004, in (eds.) R. ErdÈlyi et al., `Waves, oscil lations and smal l scale transient events in the solar atmosphere: A joint view of SOHO and TRACE', ESA-SP 547, 417. Wiehr, E., Stellmacher, G., Balthasar, H. 1984, Sol. Phys., 94, 285. Wiehr, E. 2004, in SOHO 13 ­ Waves, oscil lations and smal l-scale transient features in the solar atmosphere: a joint view from SOHO and TRACE (ed. H. Lacoste), ESA SP 547, 185. Williams, D.R., Phillips, K.J.H., Rudawy, P., Mathioudakis, M., Gallagher, P.T., OShea, E., Keenan, F.P., Read, P., Rompolt, B.: 2001, Mon. Not. R. Astron. Soc. 326, 428. Williams, D.R., Mathioudakis, M., Gallagher, P.T., Phillips, K.J.H., McAteer, R.T.J., Keenan, F.P., Rudawy, P., Katsiyannis, A.C.: 2002, Mon. Not. R. Astron. Soc. 336, 747. Wills-Davey, M. J., Thompson, B. J.: 1999, Solar Phys. 190, 467. Wills-Davey, M. J.: 2006, Astrophys. J. 645, 757. Wills-Davey, M. J., DeForest, C. E., Stenflo, J. O.: 2007, Astrophys. J. 664, 556. Xia, L.D, Popescu, M.D., Doyle, J.G., Giannikakis, J.: 2005, Astron. Astrophys. 438, 1115. Yi, Z., Engvold, O.: 1991, Solar Phys. 134, 275. Yi, Z., Engvold, O., Keil, S. L.: 1991, Solar Phys. 132, 63. Zaqarashvili1, V., Oliver, R., Ballester, J. L.: 2006, Astron. Astrophys. 456, L13.

Rev_sol_press_final.tex; 9/08/2007; 19:07; p.29


Rev_sol_press_final.tex; 9/08/2007; 19:07; p.30