EIT Waves - Cadence issues
From RHESSI Wiki
|1st Author:||David Long|
|2nd Author:||Peter Gallagher|
|Next Nugget:||The McClymont Jerk|
|Previous Nugget:||Faculae and microflares|
Globally propagating coronal disturbances as observed by the SOHO/EIT telescope are known as "EIT waves". They are spectacular events, observed in the EUV, in which structures appear to sweep across the surface of the Sun on a timescale of minutes. Many theories have been suggested to explain them, with one suggestion linking them with the long-known "Moreton Wave," a phenomenon seen in the chromosphere. The standard theory for such a wave is that it is the skirt of a coronal magnetohydrodynamic (MHD) wave. The EIT wave could be that, or it could be a low-coronal manifestation of a CME (Coronal Mass Ejection) as it departs from the corona. A discrepancy has been noted between the high Alfvén speed in the corona and the much lower observed speeds of most EIT waves, calling into question the direct interpretation as an MHD wave.
Figure 1: Three running difference images showing the 2007 May 19 event as observed using the 195 Å passband onboard STEREO-A.
In this Nugget we discuss observations of an event from 2007 May 19 (see Figure 1) made using the STEREO observations. These indicate that the low time cadence of EIT may have produced erroneously low values for the "EIT wave" speed, and thus confused the interpretation of the phenomenon.
"EIT Wave" Theories
There are currently two classes of theories that attempt to provide a physical explanation for "EIT Waves". A theory proposed by Uchida in 1968 argues that fast-mode MHD waves can propagate freely in the solar corona, following a pressure pulse delivered by a flare. Such a process would explain the Moreton wave phenomenon, as well as the metric type II radio burst. Another class of theory proposes that the disturbances are not true waves, but simply evolving disturbances which show the skirt of a CME sweeping the solar atmosphere during its eruption. Both sets of theories have their arguments for and against. The major argument against the wave interpretation is that the measured wave speeds are often below the estimated Alfvén speed in the corona, suggesting that the disturbance cannot be a fast-mode MHD wave. The non-wave interpretation has been shown in simulations to have both Moreton waves (observed in Hα) and EIT Waves observed for the same event. However, this is not always the case.
Figure 2 (taken from a recent paper by Veronig et al.) below gives some input on this question by linking the wave initiation to the impulsive phase, identified via RHESSI hard X-ray emission, of the associated flare.
Figure 2: Time history of global wave development in the event of Figure 1, from a recent paper by Veronig et al. The curves labeled "CME 1," "CME 2," and "Coronal Wave" show image positions in the three available imaging spacecraft: SOHO , STEREO Ahead, and STEREO Behind. The GOES soft X-ray (smooth curve) and RHESSI hard X-ray (noisy curve) time histories show how the thermal and non-thermal aspects of the flare evolved. It is a good example of the Neupert effect.
We used data from the STEREO spacecraft for our analysis of the 2007 May 19 event. STEREO provides observations in the 304 Å, 171 Å, 195 Å and 284 Å passbands at a much higher temporal and spatial cadence than SOHO/EIT. We obtained the distance, velocity and acceleration graphs seen in Figure 3 using the point-and-click method.
Figure 3: Three plots showing the variation in distance (top), velocity (middle) and acceleration (bottom) with time for the 2007 May 19 event as observed by STEREO-A (The "Ahead" spacecraft). Each of the passbands are shown here at normal cadence.
As can be seen, although they record the same event, there appears to be little correlation between the different passbands, especially between the 171 Å passband and the others. To examine this difference, we articficially degraded the cadence of the 171 Å images by taking every fourth image and using our analysis technique as before. We then plotted the variation of the distance, velocity and acceleration with time as before, and obtained the graph seen in Figure 4.
Figure 4: In this case, the 171 Å data have been artificially degraded to the same cadence as the 195 Å and 304 Å data (10 minutes). The 284 Å data has not been plotted in this case as the 20 minute cadence is not comparable to the 10 minute cadence of the 304, 171 and 195Å data.
The graphs now show a remarkable correlation with each other. This implies that what we observe in the 171 Å data is comparable to the 195 Å and 304 Å data. It also implies that the previously observed wave speeds may have been underestimated by the low temporal cadence of SOHO/EIT. The actual wave speeds may in fact be closer to the Alfvén speed than previously thought, opening the door for the fast-mode MHD wave interpretation in this example. A more detailed discussion of these findings may be found in Long et al. (2008).
The initial acceleration and subsequent deceleration shown in Figure 3 has been a topic of much debate in the community. There are a number of likely scenarios which may explain this behaviour:
- One explanation is that proposed by Veronig et al. (2008), where an initial source region expansion drives a freely-propagating decaying shock wave which is observed as a decelerating "EIT wave". It should be noted that the initial data points presented here were not included in the Veronig et al. analysis.
- An alternative is that our initial data points depict the wave accelerating before peaking and beginning to decelerate. It must be noted that the wave feature is followed by a pronounced dimming region in the running difference images for all the 171 Å images used. If the feature was, e.g., a moving loop, the dimming would not grow with the propagation of the feature.
This data shows the importance of finding a way of determining wave motion from plain rather than difference images. It also shows the importance of factoring cadence into calculations of wave kinematics.
Biographical note: David Long is a PhD student in the Astrophysics Research Group in Trinity College Dublin. Peter Gallagher is a Lecturer in Astrophysics at Trinity College. This work was carried out in association with James McAteer and Shaun Bloomfield. James is a Marie Curie Research Fellow and Shaun is a Postdoctoral Research Fellow, both based at Trinity College.