Before evaluating the different electron density diagnostics, it is important to establish the relative pointings of both instruments. The pointing for the s3477r01 dataset was X=537, Y=-205 while for the 165749 dataset it was X=530, Y=-200 (see Table 1). Thus the solar Y location of the slit for dataset 165749 is 5 arc sec above that for dataset s3477r01. It is necessary however to also take into account the slit sizes of the two instruments. The slit of SUMER is 117 arc sec in length while that of CDS is 119 arc sec. (The spatial pixel size of SUMER is 0.97 arc sec while that of CDS is 1.68 arc sec). Thus, it may be seen that position 8 to 49 arc sec in dataset 165749 corresponds in absolute terms along the Y-axis to the region 14 to 55 arc sec in dataset s3477r01. However, the start times for dataset s3477r01 from the CDS data and dataset 165749 from the SUMER data differ by 68 minutes (see Table 1). Taking solar rotation into account there is a distance in X of 16 arc sec between the starting locations of each slit.
In dataset s3477r01, a summed region from 14 to 55 was used, while in dataset 165749 the summed region from 8 to 49 arc sec was measured, which correspond to similar Y locations. Results from each of these datasets are shown separately below and the results discussed in Sect. 5.
The start times of datasets s3479r00 and 173109 are separated by 37 minutes, but they observe the same spatial location, taking solar rotation into account. Region 10 to 55 arc sec in dataset s3479r00 and 4 to 49 arc sec in dataset 173109 are therefore observing the same spatial region, within 1 arc sec.
In the Fe XIII 359/348 ratio it was found that due to blending of the Fe XIII 359.64Å line with a Ne V line at 359.37Å (Thomas & Neupert 1994, Dere 1978), it was necessary to use two Gaussians in the fit in order to successfully de-blend these lines. The Ne V component in the blend contributed 10-15% of the total flux. The Fe XIII 348.18Å line also suffered blending, in this case with a high temperature line of Fe XVII at 347.406Å. This was again taken into account by fitting two Gaussians to the Fe XIII /Fe XVII line profile.
Because the lines of interest are strong and close together in wavelength we assume that any errors due to line fitting and instrument calibration will be small. In this case errors in the line ratio are estimated to be approximately or 0. 3. This corresponds to errors in the derived density from the Fe XIII ratio of approximately 0.3 dex.
The average measured electron density over the observation time in this region, from the QUB results, is Log Ne= cm-3. The equivalent value from CHIANTI is Log Ne= cm-3.
These compare well with the values of Log Ne=9.7 and 9.5 cm-3 found by Keenan et al. (1995) for two active regions on the Sun using the Fe XIII ratios 318.12Å/320.80Å and 256.42Å/251.95Å. It will be noted from the plots that the results derived from CHIANTI give a systematically higher value than those of QUB. Also note that there are gaps in the densities derived from CHIANTI, because a number of the intensity ratios lie above the high density limit and are therefore not plotted. Only those values plotted have been used to derive an average density.
The average electron density calculated along the whole slit (i.e.0-118 arc sec) for the QUB results is Log Ne= cm-3. The equivalent result from CHIANTI is Log Ne= cm-3. Assuming the ionization equilibrium temperature of Fe XIII to be Log Te=6.2 K (Arnaud & Rothenflug 1985), this implies coronal electron pressures of 15.9 cm-3 K and 16.3 cm-3 K, respectively.
In Fig. 4, results are shown for the region 10 to 55 arc sec. The average electron density for this region over the observation time is Log Ne= cm-3 from the QUB results, and Log Ne= cm-3 from CHIANTI.
The average electron density over the whole slit (0-118 arcsec) has a value of Log Ne= cm-3, using the QUB results. The corresponding average value found from CHIANTI is Log Ne= cm-3. These imply coronal electron pressures of 15.7 cm-3 K and 16.2 cm-3 K respectively. Thus there is no long term variation in the coronal electron density in this region over a period from 16:57 UT to 18:04 UT, i.e. a spatial region covering 12 arc sec in the East-West direction.
For this dataset we choose to observe the region from 8-49 arc sec. This corresponds to the whole of a bright region at the bottom of the slit. The theoretical O IV curves used to derive Log Ne are shown in Fig. 5 while the resulting values are shown in Fig. 6. The difference between the CHIANTI and QUB O IV theoretical line ratios apparent at higher electron densities, is due to the different A-values used by these groups.
Errors in the line ratio are estimated to be approximately or 0.03. This corresponds to errors, in the derived density from the O IV ratio, of approximately 0.25 dex.
The electron density, based on the QUB atomic data begins at a maximum value (in this case Log Ne=11.3 cm-3) and then settles down to be roughly at Log Ne=10.8 cm-3. The average value over the full observation time is Log Ne= cm-3. The equivalent average electron density from CHIANTI is Log Ne= cm-3. Small surges in density are present at 15 and 18 minutes, these will be discussed further in another paper (Doyle et al. 1998). They are more apparent in the results from CHIANTI for which an electron density of Log Ne=11.8 cm-3 may be assigned to the peak at 15 minutes.
Along the whole slit (i.e. summing over the full 117 arc sec) the average electron density was found to be Log Ne= cm-3 using the QUB results. The equivalent value from CHIANTI was Log Ne= cm-3. These imply transition region electron pressures of 15.7 cm-3 K and 15.9 cm-3 K respectively, if we assume that the O IV line has a temperature of maximum abundance at Log Te=5.2 (Arnaud and Rothenflug, 1985).
The region 4 to 49 arc sec comprises a bright region at the bottom of the slit. The derived electron densities for this region are shown in Fig. 7. A clear upsurge in density is visible here in the QUB results towards the end of the observation with Log Ne reaching 11.8 cm-3 . It is clear that there is a considerable increase in electron density in a relatively short time which may indicate that a small scale flaring event is occurring.
The average value over the observation time is Log Ne= cm-3, from the QUB results while the CHIANTI results imply Log Ne= cm-3.
Along the whole slit (i.e. summing over the full 117 arc sec) the average electron density was found to be Log Ne= cm-3 using the QUB results. The equivalent value from CHIANTI is Log Ne= cm-3. These densities imply electron pressures of 16.0 cm-3 K and 16.2 cm-3 K respectively in the transition region.
© European Southern Observatory (ESO) 1998
Online publication: September 17, 1998