What is there before the flare?
From RHESSI Wiki
|1st Author:||Hugh Hudson|
|Published:||29 January 2007|
|Next Nugget:||RHESSI and submillimeter waves|
|Previous Nugget:||The Next Generation - FOXSI|
A solar flare results from the sudden conversion of energy stored in the solar corona. The energy comes from the magnetic field, and goes into forms that we can observe (X-rays, EUV, H-alpha;, radio waves, etc.). How does this happen? The coronal plasma must undergo an instability, subject to its global constraints. We do not know the nature of this instability (but there might be more than one); in any case the restructuring of the coronal magnetic field has to be such that its energy density, in an integral sense, decreases. The best way to identify the nature of the instability might be to compare the pre-flare and post-flare structures. Can we do that well enough with RHESSI and Hinode images?
These are the best data yet for this purpose. The physical problem is to identify the magnetic structure before and after; in principle this can be done by precisely locating the footpoints of a flaring loop. Because of the "line-tying" requirement of ideal MHD, we can think of the magnetic flux passing through the solar photosphere as unchanging, up to a certain time scale. On shorter time scales the coronal field can tilt and sway (but probably not twist). Thus we might be able to trace the field - even if perturbed - into the coronal locations where the interesting physics might happen. Because of the way that solar flares work, hard X-rays may provide the most understandable marker for a flare's footpoint regions.
Practicing with a microflare
We have started with a microflare well-observed by Hinode. Alas, there were no RHESSI observations of this event, but it is for practice; please see future Nuggets in which we plan to use RHESSI to identify the footpoints more precisely. Figure 1 shows the microflare and its pre-event active region, in soft X-rays as detected by the X-ray telescope on Hinode.
Figure 1: The active region (left) and a difference image (right) showing the locations of the flaring loops. The two images are separated by five minutes (5:02, actually), during which time the loop footpoints in the photosphere must remain "frozen" in place.
This figure shows pretty clearly that the clearly the flaring structure is grossly different from the preflare structure. Can one see the flaring loops at all in the in the preflare image, or were those parts of the corona that flared simply invisible? And if the latter, what does that mean? In fact, if one looks closely at the images, more closely than is possible just from these representations, there are some pre-flare features that are similar to and even close to the positions of the flare features. In most cases, though, these new data seem to confirm the pre-RHESSI and pre-Hinode conclusion that the preflare magnetic structure is typically not detectable in soft X-rays. This means that it is tenuous, cool, and has a low gas pressure.
So: we don't see the preflare plasma in soft X-rays. Just how tenuous and cold can it be, i.e. what upper limits can we place upon the gas pressure in these structures? For reference, the coronal gas pressures in visible active-region loops, including bright flaring loops, lie in the range 10-2 to 103 dynes/cm2. Note that these pressures are much smaller than Earth's atmospheric pressure, although that is not particularly relevant here. The lower limit to active-region loop pressures is actually not known, or easily knowable, because they become too faint to see. The brightness of the flaring loops (see Figure 1) results from their higher temperature, which correlates with higher density - the pressure (2nT) is the main thing. An X-ray telescope responds to a plasma at a given temperature and density via I = n2 x f(T), where f(T) describes the detector efficiency for the given range of photon energies. It turns out that, for an equilibrium loop model, the sensitivity f(T) must be multiplied by T4, which is shown in Figure 2.
Figure 2: Detector responses for the Hinode soft X-ray telescope, for three of its filters. Each filter has a different response to the source temperature, such that the intensity varies as n2 x f(T) (with n the density) and for our purposes the response function f(T) must be multiplied by T4 according to theory. Thus a wholesale variation in loop brightness can result from a trivial investment in temperature change!
We can crudely compare the visible flare loop with an invisible preflare loop by assuming that they have the same geometry. In some cases this looks like a good assumption, but it is common belief that magnetic reconnection plays a role in flare physics, and in any case some sort of restructuring is certainly necessary. By making this assumption, we learn that in this case the preflare temperature and density appear to be at or below the minimum ever recorded in the corona, with a correspondingly low pressure. These are only upper limits at present, but they will be greatly improved or even converted to real measurements in the future, by use of other data from Hinode, namely from its EUV imaging spectrograph EIS.
Biographical note: Hugh Hudson is a senior RHESSI team member at UC Berkeley.