Confined Flares versus Eruptive Flares
From RHESSI Wiki
|Confined Flares versus Eruptive Flares|
|1st Author:||Jie Zhang|
|Published:||31 August 2009|
|Next Nugget:||RHESSI in Cycle 24|
|Previous Nugget:||Solar Cosmic Rays of the GLE on 20 January 2005|
This is a follow-up to the Nugget CMEless Flares, which demonstrated that CMEless flares had the standard soft-hard-soft evolution typical of the hard X-ray spectrum in the impulsive phase of a flare. This differs from the often soft-hard-harder evolution of flares associated with CMEs. In this Nugget, I focus on the possible differences of the source region magnetic field distribution. Most of the discussions are from reference .
Before going further, I would like to clarify the usage of terminology. Flares are historically divided into two types: eruptive flares and confined . The former type is associated with various phenomena involving the disruption and reconstruction of the large-scale coronal magnetic field, including coronal mass ejections, filament eruptions, post-eruption loop arcades and gradually separating ribbons. The latter type, as indicated by its name, is restricted to a relatively small physical volume in the corona. Routine observations of CMEs with high quality provided by SOHO/LASCO now enable us to make definitive classification of flares into these two types. In all future discussions, a confined flare is strictly defined as a flare not associated with a CME, regardless its possible association with filament eruption or loop arcade; similarly, an eruptive flare is strictly associated with a CME of definitive observations. In this sense, a confined flare would be just CMEless flare. But I would like to use the “confined type” or “eruptive type” in order to retain the historic context.
It has been well known that the occurrence rate of eruptive flares increase as the intensity and duration of the flares increase (, ). Based on a large number of events with SOHO observations, the flare eruption rates are 16-25%, 42-55% and 90-92% for C, M and X-class flares, respectively . The interesting question is why certain flares are confined, while others are eruptive. This question shall be addressed in both observational and theoretical aspects. In this Nugget, we focus on observational aspects, in particular, the magnetic environment of the most energetic X-class flares, both eruptive and confined. The most interesting result is that the observations suggest that the overlying magnetic field play an important role in preventing or allowing a flare to be eruptive or confined.
Among the 104 X-class flares from 1996 to 2004, 11 of them were confined events . All of these had short durations in soft X-rays: their rise times did not exceed 13 minutes (except one event of 23 minutes), and the decay times (using the standard GOES definition of return to half of peak flux) did not exceed 10 minutes. The highest peak intensity of these events was X3.6. We selected four out of the 11 events for further in-depth analysis on the following criteria: (1) the source region was relatively close to the central meridian, allowing better characterization of magnetic fields, (2) the flares were isolated in time; there were no other flares from other active regions immediately proceeding or following the events studied, and (3) the flares were isolated in space; ie there was no other coronal disruption event in the vicinity of the flare region within a certain period. To make a comparative study with these confined X-class flares, we chose four events out of the 93 eruptive X-class flares. The four eruptive events were chosen such that they had almost identical GOES X-ray profiles as the four confined events, similar intensities (between X1.0 and X2.0), and rise time and decay times less than 13 minutes (Figure 1). It is interesting to point out that such short-duration or impulsive flares can also be eruptive. The fact that the longer duration flares have a higher eruption rate is useful only in a statistical sense, but carries less significance for theoretical modeling since there is no simple cut-off of the durations for the two types of flares.
We have studied the source region magnetic field properties of the two sets of flares. There is no apparent difference in the total magnetic flux between the two types. However, we found two magnetic parameters which do effectively discriminate between the types (Figure 2). The first parameter is the displacement distance, which is measured as the distance between the location of the flare and the center of mass (COM) of the magnetic flux distribution in its active region. For the four confined events, the displacement parameter is from 6 to 17 Mm, while for those eruptive events, the parameter is apparently larger, from 22 to 37 Mm. This simply implies that a flare is less likely to to be eruptive if it occurs closer to the centroid of the active region.
The second parameter is the ratio of coronal flux between that in the low corona and that in the high corona, as described in reference . The magnetic flux ratios for the confined events vary from 1 to 6, while the ratios for the eruptive events are larger, varying from 7 to 10. The magnetic flux ratio essentially serves as a proxy of the relative strength (or weakness) of the overlying magnetic field, compared with the strength of the inner core magnetic field. The lower the ratio, the weaker the overlying magnetic field, and thus the more difficult the eruption.
We conclude that magnetic field environment of the flare region is important in determining whether a flare would be confined or eruptive. The overlying magnetic field may play a critical role. A conceptual understanding seems straightforward: a relatively stronger overlying field makes the eruption more difficult. The eruptiveness may be less sensitive to the total magnetic flux (in other words, total magnetic energy) of the source region. Instead, it is more sensitive to the distribution in the 3-D corona. This scenario helps explain the observational facts described above. It seems that this scenario is consistent with the theoretical model of torus instability, as described in reference . A flare is likely to be eruptive if the overlying flux is relatively weak and/or the occurring location is farther from the center of the flux concentration. The observational test of this idea can be extended to a much larger sample of events when M-class flares are used (in the order of one thousand versus one thousand events), but it is more time consuming.