Weekly Notes from the Yohkoh Soft X-Ray Telescope

(Week 24, 2002)

Science Nugget: June 14, 2002

Properties of big flares in this cycle


In a previous nugget (2001 August 17), we reported on the Yohkoh observations of X-class flares. Since then an additional 15 X-class flares have occurred (two in 2002), and in the meantime Yohkoh operations were closed in December 2001. Therefore, it is time to give an update to the previous nugget. Unfortunately, the number of the flares observed by Yohkoh in the early phase has increased by only three (to 26), as compared with the total number of X-class flares being 58 now. But we have at least some information on pre- and post-flare morphology (from SXT full-disk images) for most of the X-class flares.



As already noted in the previous nugget, not many of these X-class flares are of long duration, often referred to as LDEs (long decay events or long duration events). This may be seen in the following plots, which show the distribution of decay times in a 30 min bin and the correlation between the rise and decay times. These times refer to the GOES 1-8 flux, and depend, of course, on how to define the start and end times. Here we take the start time from the NOAA event list, but define the end time as the time when the GOES flux drops to 5% (as opposed to 50% as used at NOAA) of the peak. There are only five LDEs with decay time longer than 200 minutes, but in the following discussion let us include those with rise time longer than 20 minutes in the LDE category.


Overall morphology

As we see below, most of these X-class flares are associated with CMEs. But the two soft X-ray morphologies often connected to CMEs seem to avoid X-class flares. The first is the sigmoid (see 1, 2, 3). The definition of a sigmoid is less straightforward than that of an LDE, but even using a favorable eye only a couple (out of the 35 X-class flares in the longitudes range E60-W60) seem to be sigmoids. For example, the five homologous X-class flares in late November 2000 typically lacked this morphology. Another CME-related morphology is trans-equatorial loops (see 4, 5). But again, we could not find the phenomenon of the disappearance of trans-equatorial loops in the X-class flares, except for one (1998 May 6), which has been taken up in these nuggets several times.


High-energy behavior

Out of the 58 X-class flares since July 1996, 27 were observed by HXT and WBS spectrometers on Yohkoh. Seven of them seem to have a bremsstrahlung spectrum hard enough to extend to the MeV range. Although in the previous nugget we reported the lack of the soft-hard-harder spectral behavior (which was characteristically associated with X-class flares in solar cycle 21), we probably have to correct it, since we seem to have find one example. The following plot shows the light curves from HXT (L: 14-23 keV, M1: 23-33 keV, H: 53-93 keV) and the power-law index from the ratio of the H and M2 (33-53 keV) channels. Interestingly, the 33-53 keV images show a compact source at what seems to be a loop footpoint unlike the coronal emission in the 14-23 keV images, as shown in this figure, where the two contours correspond to the two channels. The 1981 May 13 flare, which may be a typical example of the soft-hard-harder pattern, indicated high-energy emission from 20000 km or above. The soft-hard-harder pattern should be confirmed in HXS/GRS data.


CME association

If we pick up as LDEs those flares with decay time longer than 200 minutes or rise time longer than 20 minutes, the break-down is as follows. Here, the "halo CME" includes partial halos. The CME information is taken from the official LASCO site at CUA. LDEs are all associated with halo CMEs, but note that 85% (33/39) of short-duration X-class flares are associated with at least some kind of CMEs.

Association of X-class flares with CMEs
halo CME1225
CME (non-halo)08
No CME06
No LASCO obs07

An interesting thing is that when an X-class flare is preceded by another X-class or weaker flare, the earlier one is usually not associated with a CME.

Association with type II bursts

Distinction or identification of metric and dekametric/hectometric (DH) type II bursts is important for understanding the origin of shock(s). The following table shows that no type II burst in either frequency range is seen (with one exception) when there is no CME. The association with type II bursts in either metric or DH regime is not 100% for halo CMEs, meaning that not all halo CMEs produce shocks. The correlation between metric and DH type II bursts is not high, suggestive of different origins in a larger num,ber of cases. The information is taken from the NOAA ftp site and the Wind/WAVES Web site.

Association of X-class flares with metric and dekametric/hectometric type II bursts
metric, DHmetric, no DH no metric, DHno metric, no DH
halo CME (LDE)7320
halo CME (non-LDE)20410
CME (non-halo)2402
no CME0105

Association with SEP events

Based on the information available at NOAA, we have learned which of the flares are likely to be associated with interplanetary protons. Correlating this with other observations, we find the following.

Concluding remarks

Study of solar flares has gone into a new dimension in recent years, through the recognition that larger-scale disturbances, i.e., CMEs, could result in heating and therefore electromagnetic radiation. "Flares" resulting from CMEs may last long, hence an excellent correlation between LDEs and CMEs. But we should recall that there is a whole spectrum of other flares that are really confined and non-eruptive. These flares may represent "pure" flare processes as opposed to the aftermath of CME eruptions. What if these processes are intense? Can't they launch a CME? The flare origin of CMEs may be an outdated idea, but we wonder if we can generalize this to all CMEs, especially since CMEs are intimately associated with big flares including short ones. Part of the reason we need to study intense flares in the space weather context is to identify processes that have to do with the initiation of CMEs apart from the assumed (i.e., not yet clearly observed) destabilization of large-scale magnetic field. The numbers we have given above suggest that the flare duration may be not an important factor for understanding the possibly different relations between flares and CMEs. The involvement of filament eruptions (at different temperatures) and the way the filament erupts may give us more useful information.

[Topical index] -o- [Chronological index]

June 14, 2002

N. Nitta (NVN) (nitta@lmsal.com)