RHESSI Science Nuggets

Introduction & Theoretical Expectations

In the classical magnetic reconnection model for solar flares, magnetic field annihilates in a (vertical) current sheet, which generates turbulence and outflows of high-speed plasmas in opposite directions. The turbulence can stochastically accelerate particles and heat the background plasma. Observational signatures, such as hard and soft X-rays produced by these high-energy particles and hot plasmas, are thus expected to appear as two distinct X-ray sources, one above and one below the reconnection site (see Figure 1).

Figure 1. Schematic of the stochastic acceleration model (e.g., [1], [2], [3]) that very likely operates in solar flares. The green curves are magnetic field lines in a possible configuration; the red foam (bubbles) represents turbulence or plasma waves that are generated during magnetic reconnection.

Observations of these two distinct coronal X-ray sources, however, have been rarely reported. This may be due to limited sensitivity, dynamic range, and/or spatial resolution of the instruments, because one source may be much dimmer than the other, the two sources may be too close together to be resolved, and each may be much weaker than the footpoint sources (see "How RHESSI Makes Images"). The original discovery of a double coronal source was made early in RHESSI's history by Sui & Holman, but only a handful of other such events ([4], [5], [6]) have been reported so far, five years after RHESSI's launch. In addition, imaging spectroscopy has not been used to determine the spectra of the individual sources in these RHESSI flares, mainly due to the faintness of the second coronal source that would lie above the reconnection site.

We were fortunate to find another RHESSI flare that provides us an excellent opportunity to study these coronal sources. Its second (upper) coronal source was relatively bright and had a long duration of about 12 minutes. Perhaps more importantly, the footpoints were occulted by the solar limb so that they did not outshine the coronal emission (see an earlier nugget). These advantageous features have allowed us to examine both sources in greater detail in space, time, and energy.

RHESSI X-ray Observations

This event occurred on April 30, 2002 and was classified as a GOES M1.4 flare. Figure 2 shows RHESSI images at different energies during the impulsive peak of the flare. Two distinct sources are shown at each energy, both above the solar limb, with the lower source brighter in each case. The upper coronal source appears at lower altitudes with increasing energy while the lower source behaves oppositely, such that the two sources get closer together at higher energies. This energy-dependence of a double source structure was as originally discovered by Sui & Holman, and similar variations in a single loop-top source have been seen in several other flares. The spectra of the two sources are shown in Figure 2 together with the spatially integrated spectrum.

Figure 2. Left panel: Overlay of three images (in different energy bands) of the 2002 April 30 M-1.4 flare, integrated over 29 seconds during the impulsive peak. For each image, two contours appear in the brighter lower coronal source, while only one (lower-level) contour is present in the dimmer upper source. The two plus signs mark the centroids of the two innermost contours at 16-19 keV. The solid white line represents the solar limb. Right panel: Spectra of the lower and upper coronal sources and the spatially integrated spectrum (labeled as "total"). The spectra are shifted vertically for convenience. The two coronal sources have very similar spectra dominated by the power-law (nonthermal) component.

Interpretation & Discussion

These RHESSI observations of this flare strongly suggest that magnetic reconnection and particle acceleration took place between the two coronal X-ray sources. We elaborate our interpretation as follows:

The fact that the two X-ray sources are closer together at higher energies (Figure 2) suggests that the magnetic reconnection site is located within the small region between the two sources.
In our stochastic acceleration model ([3]), we had predicted that higher levels of turbulence tend to produce harder spectra (more acceleration and less heating). The above observations thus support the scenario (see Figure 3) that higher turbulence levels are located closer to the reconnection site, which is expected intuitively.

The light curves (not shown) and the shapes of the nonthermal spectra (Figure 2) of the two X-ray sources are similar. This suggests that intimately related populations of electrons, presumably heated and accelerated in the same acceleration region between the two sources, are responsible for producing both of them.
The thermal emission indicates that the upper coronal source has a smaller emission measure but slightly higher temperature than the lower source. This can be ascribed to the expected different magnetic connectivities of the two sources with the solar surface: the upper source would be associated with longer field lines, implying slower conductive heat losses and perhaps less material being carried up to the coronal by chromospheric evaporation as the shorter closed loops (that are likely associated with the lower source).

Figure 3. Sketch of the physical scenario superimposed on the the 14-16 keV image (green background), overlaid with the corresponding 9-10 (red) and 16-19 keV (blue) contours. These are the same images shown in Figure 2, but their orientation is rotated for demonstration purposes. The hand-drawn dotted curves represent the magnetic field lines in a possible magnetic configuration.

In summary, the proposed scenario (Figure 3) based on the RHESSI observations of a double X-ray source is consistent with the stochastic acceleration model mentioned earlier (Figure 1). Tests against observations of other particle acceleration models (such as shock and DC electric field acceleration) are beyond the scope of this Nugget.

We note in passing that there seems to be a common belief that the "Masuda" type of "above-the-loop" sources constitutes a special class of hard X-ray emission. We should point out that such sources are most likely the commonly observed loop-top sources (i.e., the lower coronal source here) that exhibit harder spectra higher up in the corona.

A more comprehensive imaging spectroscopic study, a sample of brighter flares, and a rigorous theoretical model are required to shed more light on the details of the physical processes involved.

Biographical note: Wei Liu (formerly a graduate student at Stanford University) is a RHESSI team member at NASA Goddard Space Flight Center (GSFC). Vahé Petrosian is Professor of Physics and Applied Physics at Stanford University. This Nugget is based on a paper submitted to ApJ and available online at ADS or astro-ph. Brian Dennis (at GSFC) is a co-author of the paper and has also contributed significantly to writing this Nugget.