Forum Springer Astron. Astrophys.
Forum Whats New Search Orders

Astron. Astrophys. 364, 859-872 (2000)

Previous Section Next Section Title Page Table of Contents

5. Conclusions

Using the spatially resolved images of a two ribbon solar flare, from SXT, HXT and TRACE and the spectral information from BCS we have determined the location of the non-thermal soft X-ray broadenings. Using SXT and BCS temperature and emission measure comparisons and TRACE FeXXIV observations we have shown the location of the dominant CaXIX emission is an extended region within and above the SXR flare loops.

We therefore conclude that the source of the non-thermal line broadenings is not at the flare footpoints but a region within and above the SXR flare loops. This conclusion is in agreement with that of Khan et al. (1995) who also found that the source of the [FORMULA] was not located at the flare footpoints but within the flare loops.

Having excluded the flare footpoints as a possible source of [FORMULA], there remain three alternatives: an above the loop SXR source, the SXR loop top, evaporating plasma or a combination. We do not include the reconnection site for the reasons discussed in the introduction. The evidence pertaining to each region is outlined below.

The spatial coincidence of the hottest SXR region and the HXR loop top source, both of which are above the flare loops, suggests the source of [FORMULA] is also above the SXR loops, a region postulated by Tsuneta (1996, 1997) to be heated by the slow shocks extending from the reconnection region. Tsuneta et al. (1997) calculated a temperature of [FORMULA] for this hot above the loop top region, using SXT, for an M2 class single loop impulsive flare. This temperature is not dissimilar to the temperature derived in this flare. This hot above the loop region could form before the start of the detectable HXR burst and hence be accountable for the increased levels of [FORMULA] before the start of the HXR burst, a feature common to many flares (Alexander et al. 1998). Line of sight effects could also be significant in this flare because of its location (N20E40). If the hot SXT component was located above the SXR loops, as has been observed in other flares (Tsuneta 1996, 1997), then seen in projection it could appear to be partially located within the SXR loops.

In support of evaporating plasma as the source of [FORMULA] is the fact that the CaXIX source location extends over a large area that incorporates the tops of the flare loops. Also the observations of the CaXIX line centroid shifts in the BCS data, suggesting bulk plasma motions, which change complementary to the values of [FORMULA] imply a possible causal relationship. Results of recent numerical simulations of chromospheric evaporation in solar flares (Hori et al. 1997, 1998; Yokoyama & Shibata 1998) suggest that in the early phases the hot plasma is located in the upper most flare loops and is evaporating from the chromosphere. Hence the location of the CaXIX plasma derived in this paper is in agreement with these simulations. The simulations of Hori et al. (1997) also predict temperature gradients in the loops that are in agreement with those seen in this flare. We have shown that the approximate travel time of evaporating plasma from the flare footpoints to the loop top is [FORMULA]. The evaporating plasma is driven by the electrons producing the hard X-ray burst, therefore the elevated levels of [FORMULA] after the first small HXR peak are consistent with association to evaporating plasma. We must also not rule out the presence of HXRs before they are detected by HXT if they are below the instrument detection threshold. In this case evaporation would begin earlier and further strengthen the association of [FORMULA] with evaporating plasma.

The finding of Sakao (1994) that greater electron deposition and brighter HXR emission are associated with weaker magnetic field strengths has already been referred to in Sect. 3. It is thus interesting to note that the TRACE FeXXIV observation in Fig. 10 shows that the hot SXR source is not in fact located exactly at the loop top but is slightly offset, along the loop axis, towards the south footpoint which has weaker associated magnetic field and stronger HXR emission. This image was obtained at 18:21:24 UT or very early in the HXR event when evaporating plasma might still be in the process of propagating up from the footpoints. Therefore this hot source could be evaporating plasma and is hence consistent with evaporating plasma as the source of [FORMULA]. This asymmetry is also apparent in the SXT intensity images in the early stages of the flare (Fig. 13b).

We have presented arguments that the source of the [FORMULA] is associated with either an above the loop top source or evaporating plasma. Both sources are consistent with the observations and we believe that neither can be excluded with the present data set. By examining a limb flare using this method it may be possible to eliminate the line of sight effects and more accurately determine the location to be either above or within the flare loops.

These results highlight the need for high spatial resolution observations of line profiles in solar flares, enabling us to obtain SXR spectra for isolated regions in the flare simultaneously, thus allowing the unambiguous determination of the source location of [FORMULA]. Line profiles would be required over the whole flare area at a time resolution down to a few seconds in order to separate the pre-flare, impulsive and decay phases of the flare. Observations of transition region and chromospheric line profiles will also help in understanding the origins of [FORMULA]. The Solar-B/EUV Imaging Spectrometer (EIS) scheduled for launch in 2004 will possess this capability. In the near future images from the High Energy Solar Spectroscopic Imager (HESSI) may indicate the location of hot thermal plasma in flares. These could also be used in comparison with BCS data to help locate the source of the non-thermal broadenings.

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 2000

Online publication: January 29, 2001