... Chandra/LETGS

Appendices A and B are only available in electronic form at http://www.edpsciences.org

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... as

The exact relation between observed line counts and source emission measure contains three other factors due to instrumental effective area, exposure time and interstellar medium absorption. The first and the second are known, while for the latter we have assumed $N_{H}=1.8\times10^{18}~{\rm cm^{-2}}$ (Linsky et al. 1993).

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... abundances

As a matter of fact we can derive the product between the emission measure distribution and the iron abundance and we evaluate the abundance of the other elements as ratios with respect to the iron abundance.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... value

We use as reference value the Fe solar abundance reported by Grevesse et al. (1992), therefore we have converted metallicity reported by other authors to this one.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... range

It is worth noting that within each line spectral range it is not possible to distinguish the source emission in the spectral line from its continuum emission, and from the instrumental and sky background. This fact is not so important because, for all the lines we have considered, the contamination is a small fraction of the total number photons (<25%). Moreover we have verified that the background taken on both sides of the spectrum was constant.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... transition

The measured fluxes reported by Phillips et al. indicate that the Fe XXIV intensity is 0.5 times the Fe XVII one, therefore our measured fluxes ascribed to the Fe XXIV possibly is slightly overestimated. We note however that our measured line flux results compatible with the other Fe XXIV lines, and so, as reported in Appendix B, we use also this line in the emission measure reconstruction.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...$11.17~{\rm\AA}$

Brown et al. do not measure the intensity of the Fe XXIV line at $10.66~{\rm\AA}$ because it is out of wavelength range they have explored.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... method

Each spectral feature, which has been identified with more than one atomic transition, is referred to the ion responsible for the major contribution.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... included

We note however that the total counts in the lines at 66.25 and $66.36~{\rm\AA}$ may have been slightly overestimated because of the presence of the O VII line at $22.10~{\rm\AA}$ at the third order.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.