An Alternative View of the Masuda Flare
From RHESSI Wiki
|1st Author:||Nariaki Nitta|
|Published:||2010 April 12|
|Next Nugget:||History of Solar Oblateness|
|Previous Nugget:||Particle Acceleration due to a Plasmoid-Looptop Collision|
The "Masuda flare"  has defined flare physics for a generation of solar astronomers, but do we understand its message? Regular readers of the RHESSI Science Nuggets know that the bulk of hard X-ray emission in solar flares comes from the foot-points of a coronal magnetic loop structure. These X-rays result from interaction of energetic electrons, presumably accelerated in the corona, with the denser solar atmosphere. We also have learned about the hard X-ray sources actually in the corona , which are usually weaker than the foot-point sources. These take a variety of forms and appear at different times or phases in the flare development. Among these the Masuda flare, shown in Figure 1, remains unique. In this Nugget we question how typical this flare is.
The 2D "standard model" of a solar flare
Already in that original Masuda paper, a cartoon described the coronal hard X-ray source as the result of a high-speed reconnection jet colliding with the flare loop seen in soft X-rays. In order to establish the importance of magnetic reconnection in solar flares, the Yohkoh researchers tried to find additional examples: other limb flares that also contained the "Masuda" above-the-loop hard X-ray source. No other good examples were in fact found, but it was nevertheless argued  that impulsive flares, traditionally thought not to be eruptive, might commonly have "plasmoid" ejections. This inspired a re-invention of the wheel, applying or slightly modifying the 2D theory cartoon for eruptive flares from the 1960s and 1970s directly to explain all flares. The "standard" model to date assumes that magnetic field in the corona above the flare region somehow gets destabilized, causing magnetic reconnection associated with plasmoid ejection, which then intensifies the inflow to the reconnection region and drives faster reconnection. Particle acceleration is not an official part of the model, but it is often assumed that the outflow from the reconnection region can account for this as a secondary effect due to turbulence or shock waves. Because the images of the Masuda flare appeared to contain several of the building blocks of this model, it has been extensively cited as a prototype. But then how does the model qualify as standard if a vast majority of flares look different from this prototype?
Revisiting the data
We have re-analyzed the old Yohkoh data for this and other limb flares. Concerning the 13 January 1992 "Masuda" event, the flare itself has been analyzed by many and there is not much to add. We confirm the location of the above-the-loop top source, its dependence on energy, its brightness relative to the foot-points, etc. The only thing to mention is an indication of loop shrinkage in the very early phase as captured in low-energy hard X-ray images, probably adding constraints on how energy release proceeds. We have no space here to elaborate, but this may be a key feature of flare development, as noted in an earlier RHESSI Nugget.
We have (re-)assembled 20 limb flares observed by Yohkoh to determine how unique the Masuda flare is. Most of these flares come from earlier lists compiled by Masuda himself, and also by others, but we have added four more flares with good data coverage and enough counts in the HXT 33-53 keV energy band. Among all of these we conclude that only the original Masuda flare shows a hard X-ray source more than 5000 km above the soft X-ray loop, the key attribute that promoted this flare to its iconic status. So it is definitely unique within the decade-long Yohkoh archive.
Now let us look at large-scale structures around the flare loop. We play with movies of soft X-ray images that cover a 10 arcmin field of view, such as this one. It is very hard to trace any ejection in this example, even though this was cited in  as good evidence. We do encounter other flares that look like plasmoids or loop ejections. This one appears to contain both a wave and a loop ejection, and we found two other similar related events.
So the Masuda flare may not be regarded as a good example of a plasmoid ejection either. One noticeable thing about its movie is that the corona to the north looks quite dynamic. Fortunately, SXT took full-disk images (which were usually disabled during flares) just before and after the flare, letting us know what it the global corona was doing. Figure 2 shows a time sequence of images of the west limb on which the Masuda flare occurred. The northern part of the movie actually covered part of the trans-equatorial loops that connected the active region (AR 6994) with another region in the northern hemisphere. See the largest box in (b) that corresponds to the field of view of the movie. These trans-equatorial loops existed before the flare, although diffuse, and changed shape and brighteness through the flare, suggesting that they are post-flare loops as is the clear loop in Figure 1. This is not predicted in the standad model. It is possible that 3-d quadrupolar reconnection may be needed to explain the observations, although we do not know much about how this works.
We also checked the SXT full-disk images of other limb flares, looking for changes in large-scale loops around but separate from the main flare loop(s). It appears that the changes in brightness and shape of large-scale loops are most prominently seen in the Masuda flare. So they may be the key to the understanding of the above-the-loop hard X-ray sources. For curious readers, here is a table that summarizes the 20 limb flares that were observed by both SXT and HXT from the beginning and that have signals in the HXT M2 channel enough for image reconstruction. All the plots and movies that are supportive of our view are linked in the table.
The main purpose of this Nugget has been to point out that the 2D standard model may not be the only explanation for the Masuda flare. There are probably other flares better described by the model, even though the key clue of coronal hard X-ray emission may be missing. We advocate expanding the temporal and spatial domains in data analysis when studying solar flares; the Masuda event itself has at least two probably important features (see above) that had not been noticed or much discussed in the rush to match the standard model cartoon. Generally following active-region evolution and changes in surrounding regions in multiple wavelengths may give important contextual information on the environment in which the flare occurred.
Back in 1992 the data coverage was not as good as at present time, and it may not be possible ever to understand the Masuda flare much better than we now do. But we point out that these old Yohkoh data may still be valuable especially in terms of the simultaneous soft and hard X-ray images of other events especially. RHESSI may have observed a number of Masuda type events, but it is not straightforward to compare them with the 13 January 1992 flare, since soft X-ray images are often missing. Because of the unfortunate launch delay of RHESSI, we missed a chance to compare its low-energy (e.g., 6-12 keV) images with SXT images.
All of the Yohkoh data are readily and conveniently available at the Yohkoh Legacy Archive site.