3. Structure and time evolution of the PFL derived from SXT and CDS observations
The time evolution of the hot parts of the PFL system ( MK), at approximately the same time when the slit of CDS scanned it, is very well visible in the images obtained by SXT (Fig. 2). A rapid evolution of the hot part of the loop system is seen in the first pictures of the sequence, while very little change is seen in the last images. The length of the loop determined from these pictures is approximately cm. It is also apparent that the examined loop system is surrounded by a hot rare coronal plasma. An influence of this plasma has to be taken into account when the temperature, emission measure and electron density are determined. Apart from the PFL system and the rare coronal plasma around it, it is also possible to recognize some loops above PFL and two bright points; one above the northern footpoint of the loop and the second under it. These two bright points brighten when the PFL system is disappearing. In this analysis we discuss only the behaviour of the conspicuous loop-like structure visible in the first images of the sequence.
The CDS images of the PFL system, in eight selected lines which cover the temperature range from K to 2.2 MK, are shown in Fig. 3. Here, in all lines roughly only one half of the loop system is clearly visible. Why this happens in the hot lines ( MK) is apparent from a comparison of the CDS images with the time sequence of images obtained by SXT (Fig. 2). The hot loops in the raster are visible until approximately 13:22 UT, which corresponds to the time when the hot loop observed by SXT starts to disappear as well. It will be shown in the next section that this is mainly due to plasma depletion from the loop. In the cooler lines this could be due to the combined effect of plasma depletion and cooling.
It follows from the way the CDS rasters are built that the eight images with the loop-like structures in Fig. 3 are exactly cospatial. When the positions of the tops of the semi-loops taken in different lines are compared, it is apparent that those in the images which are taken in lines having higher formation temperature lay above those formed at lower temperatures. On the other hand, the structures visible in different lines are not spatially separated, but they are overlapping, which is demonstrated in Fig. 4. This temperature stratification with height can be either real, as it is supposed in theoretical models of PFL (Kopp & Pneumann 1976, Forbes & Malherbe 1986, etc.), or it could be mimicked by the combined effect of cooling and scanning the loop system with the CDS with a finite speed. The scanning across the loop tops in all available lines took approximately 5 min (see Fig. 3). In contrast, SXT measurements (Sect. 4) show that plasma cooling in the hot part of the system is very slow. Also the theoretical estimate of the PFL plasma cooling rate (Sect. 7) does not show any fast cooling in the temperature region above MK which could explain the thermal stratification of the loop system by the combined effect of cooling and scanning speed. We believe that at least in the lines with formation temperature above 1 MK we observe a real temperature stratification with height, as a 'snapshot' of the PFL system evolution.
When the shapes of the semi-loop structures visible in hot and cool lines are compared (Fig. 3), it is apparent that they are much smoother in the hot lines while in the cool lines there are some irregularities in their shapes (structures inside the dotted boxes). They are clearly not a part of the loop system, so they have to result from a combined effect of the PFL system time evolution and the finite scanning speed of CDS. We interpret these irregularities as a manifestation of a rapid cooling of PFL plasma visible in cool lines. If we admit that there is a real temperature stratification with height in the observed PFL system, then plasma with temperature a little higher than the formation temperature of the particular line is located above the bright loop tops seen in that line, and this hotter plasma is at that moment invisible in this line. But since the CDS slit scans the loop system with a finite speed approximately one step per minute (in Fig. 3 from right to left), the plasma with temperature just greater than the formation temperature of the particular line manages to cool down and becomes visible as the irregularities in the loop-like structures. Because the time necessary for the completion of one N-S stripe exposure is approximately only one minute we can interpret these irregularities as a manifestation of very rapid, probably radiative plasma cooling visible in O V, O III and perhaps also He I lines. Since in the hot lines no similar features are seen, we believe that plasma visible here cools much more slowly compared to the CDS scanning speed.
Another prominent feature is the dependence of the sharpness (or diffusivity) of the PFL images on the formation temperature of the line which was used to take the image. From the CDS data available it follows that the images taken in hotter lines tend to be more diffused than the ones taken in cool lines, which are much sharper. On the other hand, for example, in the image taken in He I line (Fig. 3), the loop top where the CDS slit was parallel to the scanned structure is quite sharp but the part of the loop in the lower left corner, where the CDS slit was perpendicular to the scanned structure, is rather diffused. So we believe that this is probably caused by the combined effect of CDS scanning and different speed of plasma cooling in different lines, rather than by a real structural difference of hot and cool loops, although using this data we can not completely exclude the possibility of real structural difference of hot and cool loops.
© European Southern Observatory (ESO) 2000
Online publication: March 9, 2000