Astron. Astrophys. 355, 769-780 (2000)

8. Discussions and conclusions

In this paper we used simultaneously taken CDS and SXT data to examine a decaying PFL system resulting from a small single loop flare GOES class C2.9. From the structural features contained in the CDS images taken in lines with different formation temperature we were able to confirm that plasma with higher temperature tends to lie over plasma with lower temperature, as it is expected by the PFL formation theory and which was earlier observed by Cheng (1980) and vestka et al. (1987). Using the time evolution of temperature of hot plasma in the loop system derived from the SXT data and the theoretical estimate of the cooling time of the hot plasma in the PFL system we ruled out the possibility that the observed thermal distribution of plasma, emitting at different formation temperatures, could be mimicked by a combined effect of PFL plasma cooling and the CDS scanning speed. Another argument which supports the reality of the thermal stratification in the loop system is the existence of the irregularities observed in lines with formation temperatures under 0.5 MK which are protruding from the CDS loop-like structures. Providing the distribution of plasma temperature discussed above is real, these structures can be naturally explained as a result of a very fast plasma cooling, which can be expected at these temperatures. A similar result was obtained using the temperature sensitive line pair Fe XVI at 360.8 Å and Si XII at 520.7 Å. Using this line pair we found that the hottest plasma in the loop system is placed above the maximum intensity of the highest loop visible in Fe XVI line.

From the SXT emission measure and temperature analysis we obtained that the main role during the decay of the hot part of the loop system was played by a plasma outflow from the system. During approximately  s the mean emission measure in the loop decreased to roughly one fourth of its original value. The velocity of the plasma outflow from the loop system at the footpoints was estimated to be approximately 10 km s^-1, for a uniform cross-sectional area along the loop. The time evolution of temperature obtained from SXT data shows some signs of slow cooling at the beginning of the observational sequence. Later on, when the emission measure in the loop system substantively decreased, the temperatures are strongly influenced by the behaviour of the surrounding hot coronal plasma. When we admit that the decrease of temperature observed at the beginning of the observational sequence represents a real plasma cooling and compare its speed to the speed of the emission measure loss, we can conclude that this part of the system decayed mainly due to plasma depletion from the system and the decrease of the temperature played a minor or perhaps a triggering role in the decay of the system. From the CDS data also follows that the plasma outflow from the PFL system, responsible for its decay, started in all lines at approximately the same moment.

To obtain the geometrical filling factor at the top of the PFL system in Fe XIV line ( MK), we first determined the electron density and its uncertainty using the density sensitive line pair of Fe XIV. This method reflects the collisional rate in plasma and is not dependent on geometrical fine structural features in the system. Then we calculated emission measure of plasma in the system from the integrated intensities of an allowed line of Fe XIV at 334.2 Å (at the same locations where the electron densities using the density sensitive pair were measured). The emission measure was used to calculate the and its error. Because during the whole procedure of fitting and calculating of all the quantities necessary to determine the filling factor we took a great care to proper treatment of all errors which can influence the results, we obtained a great, but realistic uncertainty of the resulting filling factor. The results of our measurements are that the upper limit of the geometrical filling factor at the top of the loop system in Fe XIV line is while the lower limit at least in regions of maximum electron density is . Of course, the filling factor can be different for loops with different temperatures. Unfortunately the CDS data available did not allow us to carry out a similar study in lines with different formation temperatures. The results show that if the electron density is estimated from for example a SXT emission measure without knowledge of the filling factor, the results can differ of factor from to 10 from the real electron density in the observed region. This can strongly influence any theoretical interpretation of such observations.

© European Southern Observatory (ESO) 2000

Online publication: March 9, 2000
helpdesk.link@springer.de