3. Ion fractions
From Eq. 2 it is possible to see that the ion fraction is a very important parameter for the calculation of the theoretical emitted line intensity. Ion fractions have been calculated in the literature by a number of authors, using the most accurate ionisation and recombination rates available at the time. Most of these authors have assumed that the plasma is in ionization equilibrium, and have also neglected density effects in the calculation, so that this quantity is provided as a function of electron temperature only. Improved rates have been appearing in the literature during the periods between each of the ion fraction calculations, so that the ion fractions resulting from each computation may differ significantly from the previous ones.
The ion fraction datasets which have been considered for our test are:
All these computations assume that in a low-density plasma photoabsorption and three-body recombination are negligible. Charge transfer reactions are also neglected by Shull & Van Steenberg (1992), although they can be important in low temperature plasmas. Thus, differences between RAY and MAZ are due both to improvements in the ionization and recombination data (theoretical and experimental), and in the analytical formulae used to represent the data and interpolate values for the missing ions. Shull & Van Steenberg (1982) adopt a "total" collisional ionization rate to represent both direct ionization and autoionization processes, while all the more recent works adopt two separate cross sections, leading to more accurate and physically meaningful formulae. The more recent works include a larger amount of experimental cross sections to derive their analytical formulae, so that the results should be more accurate.
3.1. Comparison of different ion fraction datasets
A quick comparison between ion fractions for several ions calculated in the SHU, RAY and MAZ dataset shows that in some cases differences are not huge, but for some ions quite dramatic changes are found between the curves provided by each of these computations. A thorough and systematic comparison of these three datasets is beyond the scope of this work; in the following only a few examples of this comparison are reported, that are relevant to the present work. These examples clearly indicate how large the differences might be in some cases, so that large effects on plasma diagnostics are expected.
Fig. 1 and Fig. 2 report the ratio between the ion fractions of Shull & Steenberg (1982) and Mazzotta et al. (1998) relative to the RAY data for a few important ions observed in the CDS spectrum. The scale in Fig. 1 and Fig. 2 is logarithmic. These figures show clearly that differences are usually of the order of 20-40% between these datasets, although for instance iron ions usually agree within 10% between Arnaud & Raymond (1992) and Mazzotta et al. (1998), in some cases differences may rise up to a factor 3-4 (Ne vii) or even to order of magnitudes (Ca x).
Given the size of the changes in the ion fraction datasets, it is to be expected that plasma diagnostics techniques based on the use of EUV emission lines will be significantly affected and will provide different results when each of these ion fraction datasets are adopted.
© European Southern Observatory (ESO) 2000
Online publication: August 23, 2000