AXAF grating spectra are attractive for benchmarking the plasma codes because of: high spectral resolution in the gratings; excellent flux calibration; broad band spectrometers; and large effective areas. The three stellar calibration sources cover most of the ionization states from collisionally ionized plasmas.
The question arises as to the usefulness of stellar spectra for atomic physics benchmarks, given their intrinsic complexity. Fortunately, although stellar coronae are astrophysically complex (with loop structure, flares, winds) they appear to obey coronal equilibrium and to be optically thin.2 As a result, the emission from an element in any part of the corona is just determined by the local plasma temperature and emission measure Integral (N_e^2 dV) and the emission from all regions at the same temperature can simply be summed. Abundances then affect the relevant contributions from each element. All other physics drops out, isolating the atomic physics. Astrophysics will still come out of AXAF grating observations of stars -- through density diagnostics, variability, eclipse mapping, and not least, through the fitting of physical models to the newly accurate emission measure distributions made possible by the better atomic physics arising from this project.
Procyon. The dominant temperature is about 2 x 10^6 K. With a continuous emission measure distribution, there are emission lines from Fe IX to XI in the softest LETG band (> 170 Å), for which the atomic physics remains controversial (Brickhouse et al. 1995; Laming et al. 1995). The LETG spectral region above 150 eV is rich in diagnostics for plasmas at this temperature. Imagine the rich Fe L shell complex (currently considered a problem, but eventually will be considered a gold mine of spectral diagnostic information). The LETG will observe L shell emission for O, Ne, Mg, Si, S, Ar and Ca. In addition to benchmarking the relative line intensities from a given ion, these data will for the first time ever give us constraints on the ionization balance calculations. There is large uncertainty due to the rates that go into the ionization balance, especially at higher Z (Arnaud & Raymond 1992). By constructing separate emission measure distributions from each element one can tell by the relative shapes how well the ionization balance calculations agree.
Capella. The source has a dominant temperature at 6 x 10^6 K. Strong emission lines of Fe L shell ions are observed with EUVE. Comparison of EUV to X-ray lines from these ions will provide a stringent test of the codes for lines from different energy levels. While the source is somewhat variable, and thus cross comparison of LETG and HEG lines might be compromised, one interesting possibility would be to coordinate both AXAF grating observations with EUVE.3 Capella will also provide an excellent test of the lower temperature Fe L shell emission lines.
HR 1099. The temperature at the peak of the emission measure distribution is 1.5 x 10^7 K. The HEG spectrum includes lines of the higher ionization states of the Fe L shell ions. The spectrum also contains He-like line emission, for which there are large uncertainties in the calculation of satellite line intensities.