PROJECTION EFFECTS IN CORONAL DIMMINGS AND ASSOCIATED EUV WAVE EVENT

Coronal mass ejections (CMEs) are large-scale coronal magnetic field structures expelled into the heliosphere. They are the main causes of space weather disturbances since they may drive interplanetary shocks and produce geomagnetic storms. Accompanied with CMEs, we observe activities low in the solar atmosphere including filament eruptions, flares, large-scale coronal EUV waves, and coronal dimmings.

EUV waves are large-scale disturbances propagating through the solar atmosphere, observed as moving fronts of increased coronal EUV emission. They propagate at typical speeds of 200–400 km s⁻¹ (Klassen et al. 2000; Thompson & Myers 2009; Muhr et al. 2014), but on rare occasions, EUV wave events with speeds in excess of 1000 km s⁻¹ have been reported (Nitta et al. 2013). Several models have been proposed to explain the nature of EUV waves, falling into three categories: wave, non-wave, and hybrid models. Most wave models interpret the propagating signatures as fast-mode magnetosonic waves (Thompson et al. 1999; Wang 2000; Warmuth et al. 2001; Wu et al. 2001; Ofman & Thompson 2002), while in the non-wave models they are explained as disturbances due to sucessive restructuring of magnetic field lines during the erupting CME (Delannée & Aulanier 1999; Chen et al. 2002; Attrill et al. 2007). Hybrid models include both interpretations resulting in a bimodal picture, where a fast-mode wave travels ahead of a slower inner component due to magnetic field reconfiguration caused by the erupting CME (Zhukov & Auchère 2004). A detailed discussion on the various theories and the evidence for and against them may be found in recent reviews (e.g., Warmuth 2010, 2015; Gallagher & Long 2011; Zhukov 2011; Patsourakos & Vourlidas 2012; Liu & Ofman 2014).

The relationship between CMEs and EUV waves was investigated by several statistical studies (Biesecker et al. 2002; Cliver et al. 2005; Nitta et al. 2013, 2014; Muhr et al. 2014). It was found that fast and wide CMEs are, in general, accompanied by well-observed EUV waves that are associated with shocks, i.e., related to type II radio bursts (Cliver et al. 2005). The unprecedented multi-vantage-point observations from the Solar TErrestrial RElations Observatory (STEREO, Kaiser et al. 2008) and the high-cadence imagery by the Atmospheric Imaging Assembly (AIA, Lemen et al. 2012) on board the Solar Dynamics Observatory (SDO, Pesnell et al. 2012) improved our understanding of how the EUV wave formation and kinematics are related to the CME dynamics. Initially, the EUV wave front is found to be closely attached to the laterally expanding CME flanks. After an impulsive driving phase, the CME flanks decelerate, and the EUV wave subsequently becomes freely propagating with a velocity close to the fast magnetosonic speed in the quiet corona (Warmuth et al. 2004; Long et al. 2008; Veronig et al. 2008, 2010; Kienreich et al. 2009; Patsourakos & Vourlidas 2009; Patsourakos et al. 2009; Cheng et al. 2012; Olmedo et al. 2012). Quadrature observations of the two STEREO spacecraft unambiguously revealed that the EUV wave is propagating ahead of and is driven by the laterally expanding CME flanks (Kienreich et al. 2009; Patsourakos & Vourlidas 2009; Ma et al. 2011). These findings were also confirmed by MHD simulations of CME eruptions and associated EUV waves (Pomoell et al. 2008; Cohen et al. 2009; Downs et al. 2011, 2012).

Coronal dimmings are regions of reduced emission in the low corona observed in EUV (Thompson et al. 1998, 2000) and soft X-rays (SXR; Hudson et al. 1996; Sterling & Hudson 1997). The appearance of these dimming regions is, in general, interpreted as density depletion caused by the evacuation of plasma during the CME expansion (Hudson et al. 1996; Thompson et al. 1998; Harrison & Lyons 2000). This interpretation is supported by the simulatenous and co-spatial observations of coronal dimmings in different wavelengths (e.g., Zarro et al. 1999), spectroscopic observations showing plasma outflows in dimming regions (Harra & Sterling 2001; Attrill et al. 2010; Tian et al. 2012), and studies on dimming/CME mass relations (Sterling & Hudson 1997; Wang et al. 2002; Harrison et al. 2003; Zhukov & Auchère 2004; Mandrini et al. 2007; Aschwanden 2009; Miklenic et al. 2011). In the literature, two different types of dimmings are defined, core (or twin) dimmings and secondary (or remote) dimmings, respectively. Core dimmings are stationary, localized regions that are often present on both sides of an erupting configuration, in opposite magnetic polarity regions. In relatively simple cases, they are interpreted to mark the footpoints of the ejected flux rope (Sterling & Hudson 1997; Thompson et al. 2000; Webb et al. 2000; Mandrini et al. 2005, 2007; Temmer et al. 2011). The more shallow secondary dimmings are diffuse and can extend to significant distances from the source region. They are often observed to follow behind a propagating EUV wave front (Delannée & Aulanier 1999; Wills-Davey & Thompson 1999; Thompson et al. 2000; Attrill et al. 2007; Mandrini et al. 2007; Muhr et al. 2011). Therefore, they could be rarefaction regions that develop behind a compressive wave (e.g., Wu et al. 2001; Muhr et al. 2011; Downs et al. 2012) or formed due to the plasma evacuation behind the flux rope and overlying fields that are erupting.

In this paper, we analyze the flare-CME event and associated large-scale EUV wave that occurred on 2011 September 6. Several aspects of this event have been studied before. For instance, Nitta et al. (2013) investigated the EUV wave kinematics using high-cadence observations from SDO/AIA, and determined this event to be a high-speed EUV wave with a velocity of v ≈ 1250 km s⁻¹. Only 6 out of 171 events reported by Nitta et al. (2013) were identified in this high-speed regime (v > 1200 km s⁻¹). The globally propagating EUV wave caused oscillations of filaments and the launch of a jet (Shen et al. 2014). These results indicate that EUV waves may serve as agents for linking successive solar activities. Jiang et al. (2013) focused on the triggering mechanism of the erupting event using MHD simulations initialized by a nonlinear force-free field (NLFFF) extrapolation. Janvier et al. (2016) studied the morphology and time evolution of photosperic traces of the current density and flare ribbons and compared it with the topological features found by NLFFF modeling. Both, Jiang et al. (2013) and Janvier et al. (2016) identified a spine-fan configuration of the overlying field lines, due to the presence of a parasitic positive polarity, embedding an elongated flux rope. The energy accumulation of the flare and the accompanied CME was studied by Feng et al. (2013). They found that the calculated free magnetic energy is able to power the flare and the CME, and that both phenomena may consume a similar amount of free energy. Romano et al. (2015) studied the evolution of the source active region (NOAA 11283) in relation to its recurrent flaring and CME activity. They found that before the occurence of the X2.1 flare-CME event under study here, the shearing motions seem to inject a larger fraction of energy into the corona than the emergence of the magnetic field.

In this paper, we investigate this extraordinary EUV wave and associated flare-CME event in detail and fully exploit the quadrature view from SDO and STEREO.

2.1. Event Overview

2.2. Data

2.3. Perturbation Profiles

We study the EUV wave pulse quantitatively using the perturbation profile method (e.g., Warmuth et al. 2004; Podladchikova & Berghmans 2005; Veronig et al. 2010; Long et al. 2011; Muhr et al. 2011). The method allows a detailed study of the temporal evolution of the propagating wave front and its changes in amplitude by generating perturbation profiles along a specific sector. We derive the perturbation profiles from base-ratio images by calculating the median of the relative intensity changes of all pixels that are within the selected sector of propagation and successive annuli of 1° width around the origin of the wave. The median absolute deviations are used as 1σ measure errors. To identify the characteristics of the EUV wave, a Gaussian curve is fitted to the positive section of each profile. MPFIT (Markwardt 2009) is used to calculate the parameters of the Gaussian function with error bars. The Gaussian function that is fitted to the data is of the form

$$\begin{eqnarray}&&I(x)={p}_{0}+{p}_{3}\,\exp \left(-\displaystyle \frac{{(x-{p}_{1})}^{2}}{2{p}_{2}^{2}}\right),\end{eqnarray} \tag{ 1 }$$

where, p₀ ... p₃ are the best-fit values returned. We use the perturbation profiles to study the kinematics of the EUV wave pulse as well as the time evolution of its amplitude.

2.4. Projection and LOS Integration

The animated Figure 1 shows the evolution of the EUV wave in SDO/AIA 211 Å and 335 Å base-difference images. The EUV wave occurs as a sharp front, with a first appearance at 22:19 UT and can be followed up to about 22:27 UT. As can be seen in the 211 Å images, the segment of the wave front propagating toward the north "disappears" at 22:20:48 UT and then reappears at 22:22:48 UT. The same behavior is observed in the 193 and 171 Å channels (not shown). In contrast, the entire wave front remains clearly visible in the 335 Å channel. Coronal dimming regions are observed behind the EUV wave front. The shape and extension of the dimming regions appears differently in various SDO/AIA channels (e.g., 211 Å versus 335 Å; Figure 1).

3.1. EUV Wave

To derive the kinematical evolution of the EUV wave, we calculate perturbation profiles for the direction in which we observe the disappearance of the wave front over a sector of 15° angular width (indicated with red lines in Figure 1). The animated Figure 2 shows the time evolution of these profiles for the 211 and 335 A channels. To determine the velocity of the EUV wave, we define the leading front of the wave x_lead to be a function of the parameters of the Gaussian fit, namely

$$\begin{eqnarray}&&{x}_{\mathrm{lead}}={p}_{1}+2{p}_{2}.\end{eqnarray} \tag{ 2 }$$

Accordingly, the $1\sigma$ errors in x_lead are

$$\begin{eqnarray}&&{\rm{\Delta }}{x}_{\mathrm{lead}}=\pm ({\rm{\Delta }}{p}_{1}+2{\rm{\Delta }}{p}_{2}).\end{eqnarray} \tag{ 3 }$$

The resulting wave kinematics for SDO/AIA 211 and 335 Å are plotted in the middle and bottom panels of Figure 3, respectively. The position of the leading edge of the wave front determined by this procedure (indicated by the vertical lines in Figure 2) for each perturbation profile is plotted together with the 1σ error against time. The velocities obtained from the linear fit to the time–distance data are v₂₁₁ = 1070 ± 70 km s⁻¹ for 211 Å and v₃₃₅ = 1075 ± 80 km s⁻¹ for 335 Å.

Download figure:

Video Standard image High-resolution image

Zoom In Zoom Out Reset image size

The observations from quadrature view enable us to study the wave signature simultaneously off-limb and on-disk. Figure 4 shows a sequence of running-difference images illustrating the evolution of the eruption in ST-A 195 Å. The EUV wave front was tracked manually along the solar limb, indicated by the red crosses. Since the typical heights of EUV waves have been established to be between 80 and 100 Mm (Kienreich et al. 2009; Patsourakos et al. 2009), we also include measurements at a height of 90 Mm (∼130'' above the solar limb, i.e., circle with radius 1100'', marked as blue crosses). For SDO, we obtain the time evolution of the peak and the leading front of the EUV wave pulse from perturbation profiles for the 193 Å channel. Figure 4 also shows that initially the CME front cannot be distinguished from the EUV wave. Around 22:30 UT, the lateral expansion of the CME stops and we identify the EUV wave propagating ahead of the CME.

Zoom In Zoom Out Reset image size

Figure 5 shows the distance–time curves obtained from ST-A 195 Å (off-limb) and SDO 193 Å (on-disk). Different speeds are derived using mesurements obtained from different vantage points. From SDO, we derive a speed of v_lead = 1080 ± 80 km s⁻¹ for the leading front and v_peak = 900 ± 30 km s⁻¹ for the peak of the wave pulse. From ST-A view, the resulting speeds are significantly lower, amounting to v = 520 ± 40 km s⁻¹ for the leading front of the wave propagating along the solar limb, and v = 550 ± 40 km s⁻¹ at a height of 90 Mm. These discrepancies are investigated in more detail. In Figure 4, we show the position of the EUV wave's peak and leading front as observed from SDO transformed to ST-A view (marked as green and orange dots, respectively). We clearly see that the on-disk features neither correspond to a wave signature on the solar limb, nor to a wave signature at a height of ∼90 Mm in ST-A. To estimate the actual heights, where the emission for those signatures come from, we also plot the projected LOS of the transformed SDO/AIA wave positions in ST-A (indicated by the green and orange line) in order to identify where the LOS matches with the outermost visible front in ST-A (indicated by the green and the orange circles). We obtain, for the leading edge of the front, a height of ∼260 Mm and, for the peak of the wave pulse, a height of ∼130 Mm above the solar limb.

Zoom In Zoom Out Reset image size

Another interesting aspect of the EUV wave is the intermittent disappearance of parts of the front. Figure 6 shows the amplitude of the Gaussian fit to each perturbation profile (see orange triangles in Figure 2) plotted against time. The amplitude's time evolution as well as the perturbation profiles themselves (see Figure 2 and movie no. 2) clearly reveal that the EUV wave pulse "disappears," i.e., the relative intensity decreases to the background level of 1.0, in the 211 Å filter at 22:20:48 UT, and reappears again at 22:22:48 UT. The same behavior is observed for the 171 Å and the 193 Å channels. In contrast, the 335 Å profiles show a continuously propagating wave pulse, but the pulse amplitude reveals a decrease during the time of disappearance in the 211 Å filter (marked as black dashed vertical lines in Figure 6). After its "reappearance," the amplitude of the wave pulse increases in all four filters.

Zoom In Zoom Out Reset image size

3.2. CME and Type-II Radio Burst

Figure 7 shows the CME kinematics obtained from ST-A/EUVI, COR1, and COR2 data by tracking the CME frontal part manually along its main propagation direction. Since from the ST-A view, the event occurred on the solar limb, these measurements of the CME propagation are basically uneffected by projection effects. We derive an average speed of v = 990 ± 50 km s⁻¹, which is much higher than the value provided in the CDAW catalog derived from SOHO/LASCO observations (v = 580 km s⁻¹), where the CME is observed as a halo, i.e., propagating toward the observer.

Zoom In Zoom Out Reset image size

The event was also associated with a type II radio burst, observed by different ground-based instruments (Hiraiso Radiospectrograph, Culgoora Spectrograph, Green Bank Solar Radio Burst Spectrometer, and Sagamore-Hill). The patchy and intermittent continuation of the coronal type II emission to the interplanetary space was observed by the Waves instrument on board the WIND spacecraft (WIND/Waves; Bougeret et al. 1995) and the two Waves instruments on board STEREO (STEREO/WAVES; Kaiser 2005; Bougeret et al. 2008; Kaiser et al. 2008). The type II radio burst, a signature of a coronal shock wave, started concurrently with the EUV wave, at about 22:19 UT. The high starting frequency of the complex and structured type II burst (340 MHz for the fundamental emission band) indicates that the shock was formed low in the solar corona.

From the frequency drift rate and the generally used five-fold Saito coronal electron density model (Saito et al. 1970), we obtain a shock wave speed of v = 1020 ± 30 km s⁻¹ and a starting height of h = 1.08 ± 0.02 R_⊙, i.e., ∼60 Mm above the solar surface (marked as black dot in Figure 7). The study of Cho et al. (2013) showed comparable results for the starting height of the high frequency type II radio burst. We note that the speed of the coronal shock is comparable to the speed of the EUV wave signatures as measured from SDO and the CME speed obtained from ST-A observations (1070 and 990 km s⁻¹, respectively).

3.3. Coronal Dimmings

To identify the coronal dimming regions, we apply a thresholding technique that uses, for each pixel, an upper limit on the absolute and relative decrease of the intensity with respect to the pre-event image. A pixel is flagged as a dimming pixel if its intensity decreased by at least 20% compared to its pre-event value and decreased by a certain absolute value A (in DN) from its pre-event value. Within this set of identified dimming pixels, we determine the 5% "darkest" pixels, i.e., the 5% pixels that reveal the largest absolute change of their intensity. Figure 8 illustrates the application of this method and shows a direct image in SDO/AIA 211 Å, the corresponding base-difference image and the SDO/HMI LOS magnetogram at a time where the dimming region is still evolving. The red contours indicate the overall identified dimming regions. The black (left panel), blue, and yellow contours (right panel) mark the 5% darkest pixels. We interpret these regions as "potential" core dimming regions, since it is assumed that dense plasma, previously confined by the flux rope, is evacuated there and the selected pixels show the maximum decrease in intensity of all identified dimming regions.

Zoom In Zoom Out Reset image size

As a next step, we exploit the quadrature configuration of SDO and ST-A to study the three-dimensional extension and position of the identified coronal dimming regions. To this aim, we select regions-of-interest (ROIs) in SDO/AIA observations and perform a coordinate transformation of these ROIs to ST-A/EUVI.

Figure 9 shows SDO/AIA 193 Å base-difference images at two different time steps. We study areas within a potential core dimming region (green circle), within the overall dimming region (red circle), and an ROI that includes both enhanced and reduced intensity (cyan oval). To check for the correctness of the coordinate transformation, we mark an active region as a reference point (blue circle). To investigate the origin of the contributions for the different dimming regions, we overplot the projected LOS for selected points inside each ROI in SDO/AIA (marked as crosses in Figure 9) onto the corresponding image of ST-A/EUVI. This is indicated by lines in the corresponding colors. The sum of all pixel intensities along each LOS is expected to correspond to the observed intensity in the marked locations in SDO/AIA. Indeed, this is qualitatively obtained from the right panels of Figure 9, where we plot the cumulative sum of all pixel intensities along the LOS for each selected location. For instance, the position marked with the purple cross corresponds to a region of enhanced intensity in SDO/AIA; the ST-A view shows that the LOS crosses the CME flanks and EUV wave front. The corresponding curve for the cumulative sum over the LOS pixels shows qualitatively the same result, i.e., an enhanced intensity compared to the background. The location marked with the green cross corresponds to a core dimming region and therefore to a location of much reduced intensity. As can be seen from its cumulative curve, only reduced intensity pixels (dimming pixels) lie along its LOS. The reconstruction of the LOS for the red cross reveals that the dimming contributions result from regions higher up in the corona (see the upper panels in Figure 9). The resulting emission is much higher than the reconstructed intensity of the core dimming region (green cross), indicating that in this dimming region lesser dense plasma is evacuated. The cyan cross marks an even more shallow dimming region, which shows a smaller number of dimming pixels lying along the LOS as well as a contribution of pixels with enhanced intensities with respect to the pre-event state. The orange cross is used as a reference and, as expected, the reconstructed intensity equals zero. The right panels of Figure 9 clearly show that the highest decrease in the emission reconstructed from ST-A comes from the core dimming region, while the other ROIs (see the red and cyan locations and the corresponding curves for the cumulative sum) show regions of secondary dimmings. We note that this procedure is only approximately correct, as we can only make LOS cuts along the LOS-projected image of the approximate quadrature configuration, which by itself is also a result of LOS integration of the emission observed.

Zoom In Zoom Out Reset image size

The EUV wave/CME/coronal dimming event on 2011 September 6 shows several peculiar features. Only by complementing the results of SDO observations with ST-A, we were able to disentangle real signatures from signatures arising from LOS integration and projection effects. Furthermore, we were able to find new aspects for the formation of associated coronal dimming regions.

• 1.  
  For the first time, we identified the "disappearance" of a segment of the EUV wave front followed by its reappearance and further propagation in the absence of any obvious magnetic structures. In the following, we discuss three possible scenarios for this phenomenon.

  • i.  
    Heating at the wave front: Vanninathan et al. (2015) studied the strong EUV wave associated with the X2.2 flare of 2011 February 15 and showed that the passing EUV front adiabatically compresses the ambient coronal plasma, resulting in an increase in density of about 6%–9% and an associated temperature increase of 5%–6%. Using the method described in Downs et al. (2011), we estimated for the event under study the intensity ratio for each AIA channel as a function of temperature, by assuming that the passing EUV wave causes an ∼8% increase in density compared to the pre-event corona (consistent with the values derived by Schrijver et al. 2011; Veronig et al. 2011; and Vanninathan et al. 2015). This leads to an ∼5% increase in temperature. From the estimated intensity ratios, we obtain that for an ∼5% increase in temperature, the intensity in the 171 and 193 Å channels is reducing while that of 211 and 335 Å channels continue to increase. We repeated this test for different densities ranging from 6% to 12% increase and obtained the same qualitative results in all cases. This is contradictory to the observed evolution of the amplitude of the wave pulse revealing a decrease in intensity for all four wavelength channels during its "disappearance" phase (see Figure 6). This result is inconsistent with the disappearance being caused by heating at the wave front.
  • ii.  
    Wave propagation through an inhomogenous medium: EUV waves are known to be affected by inhomogeneities in the corona (in terms of Alfvén velocity), such as active regions, coronal holes, or filaments. Encountering a region of high Alfvén speed would alter the amplitude of the wave pulse. To test this hypothesis, we checked for magnetic obstacles in SDO/HMI LOS magnetograms along the propagation path, which would correspond to regions of enhanced Alfvén speed. Figure 10 shows LOS magnetic field and Hα data together with the distance range, where the wave front "disappears" (marked by the red and green fronts). No magnetic flux enhancements or filaments could be identified along the sector of interest. However, a filament is present along the propagation direction of the EUV wave but not at the locations where we observe the wave front disappearance. We note that the wave front passage causes the filament to oscillate, but it does not erupt (Shen et al. 2014). These observations suggest that the propagation through an inhomogenous medium is not responsible for the disappearance of the wave front.
  • iii.  
    LOS effects: the simultaneous decrease in amplitude for all channels indicates that the decrease in emission results primarily from changes in the density and not changes in temperature. The 171, 193, and 211 Å filters reveal the highest decrease. These channels are most sensitive to temperatures around 1–2 MK, which correspond to quiet-Sun coronal temperatures. During the CME lift-off, such plasma is evacuated, resulting in regions of reduced intensity (coronal dimmings). The observed intensity in SDO/AIA results from the sum of emission along the LOS. Assuming that along a specific LOS, contributions from coronal dimming regions (negative contributions) as well as the EUV wave (positive contributions) are present, the total intensity can be reduced to the background intensity, when summed up. The 335 Å channel measures plasma at higher temperatures around 2.5 × 10⁶ K and is thus less sensitive to the expansion and evacuation of quiet coronal plasma by the erupting CME (see Figure 1).
• 2.  
  Observations from SDO/AIA revealed signatures of different types of coronal dimmings. Using a thresholding technique, we were able to distinguish potential core dimming regions from secondary/remote dimming regions. The identified core dimming regions show evacuation of dense plasma, observed as a maximum decrease in intensity in the overall identified dimming region from SDO/AIA and as the largest decrease in emission from the ST-A LOS reconstructions. They are located in regions of opposite magnetic polarities and lie close to the eruption site (see Figure 8). Therefore, we conclude that these regions point to real footpoints of the erupting flux rope. For the secondary/remote dimmings (red ROI and lower part of the cyan ellipse in Figure 9), the reconstruction of the emission from ST-A revealed that whether the observed intensity in SDO/AIA results in a dimming or not, depends strongly on the structures lying along the corresponding LOS. We conclude that projection effects play an important role in the appearance and correct interpretation of coronal dimming regions.
• 3.  
  The kinematical evolution of the wave is derived to be different when measured from different vantage points. This can be explained by projection effects as shown in Figure 4. Obviously, the on-disk wave signatures cover contributions from the outermost bright front of the CME eruption, when the EUV wave is not yet detached from the CME, which is driving the wave. Ma et al. (2009) and Hoilijoki et al. (2013) found that there exist viewing angles, where the orientation of the LOS is tangential to the erupting dome, producing visible features that can be interpreted as parts of an EUV wave. In a recent study by Delannée et al. (2014), using different techniques for deriving the altitude of an EUV wave, the resulting heights span a distance range of 34–154 Mm above the solar limb, which is consistent with our result of ∼130 Mm for the peak of the wave pulse identified from SDO/AIA.

Zoom In Zoom Out Reset image size

Another important aspect is that the on-disk observations of EUV waves may contain significant intensity contributions of the emission from the expanding CME body integrated along the LOS over several scale heights. The speeds of the EUV wave (as derived from SDO/AIA) and the CME (derived from ST-A with low projection effects) are similar, implying that in both instances we measure the same propagating structure. In addition, the observed "bending" of the wave toward the solar east (see movie no 1. time range ∼22:24 UT until 22:30 UT), is another feature suggestive of projection effects, as no magnetic obstacles are present at that location, that would cause a reflection or change of propagation direction. Assuming that the "bending" results from LOS integration along the outermost front of the CME, we derive that these contributions might come from large heights in the corona. For the time when the wave "bending" is clearly observed, the apex of the CME, as derived from ST-A, is at a distance of about 2–2.5 R_⊙.

The event studied in this paper is a good example of how the use of single-view image data may limit our ability to correctly interpret coronal features, such as EUV waves and coronal dimmings.

We thank the referee for careful consideration of this manuscript and useful comments. This study was funded by the Austrian Space Applications Programme of the Austrian Research Promotion Agency FFG (ASAP-11 4900217) and the Austrian Science Fund FWF (P24092-N16). K.D. is thankful to N. Nitta and K. Kozarev for helpful discussions. A.M.V. and K.V. acknowledge the fruitful discussions during the ISSI Team Meeting "The Nature of Coronal Bright Fronts" led by David Long and Shaun Bloomfield. SDO data are courtesy of the NASA/SDO AIA and HMI science teams, STEREO data are courtesy of the NASA/STEREO SECCHI teams. The CDAW CME catalog is generated and maintained at the CDAW Data Center by NASA and The Catholic University of America in cooperation with the Naval Research Laboratory. SOHO is a project of international cooperation between ESA and NASA. This work utilizes data obtained by the GONG Program, managed by the National Solar Observatory, which is operated by AURA, Inc. under a cooperative agreement with the National Science Foundation. We are also grateful to the staff of the Hiraiso Radiospectrograph, Culgoora Spectrograph, Seagamore-Hill and Green Bank Solar Radio Burst Spectrometer for their open data policy.