Satellite lines from auto-ionizing states of Fe XVI and the problems with the X-ray Fe XVII lines

We present new calculations of atomic data needed to model auto-ionizing states of Fe XVI. We compare the state energies, radiative and excitation data with a sample of results from previous literature. We find a large scatter of results, the most significant ones in the auto-ionization rates, which are very sensitive to the configuration interaction and state mixing. We find relatively good agreement between the auto-ionization rates and the collisional excitation rates calculated with the R-matrix suite of programs and AUTOSTRUCTURE. The largest model, which includes J-resolved states up to $n=10$, produces ab-initio wavelengths and intensities of the satellite lines which agree well with solar high-resolution spectra of active regions, with few minor wavelength adjustments. We review previous literature, finding many incorrect identifications, most notably those in the NIST data base. We provide several new tentative identifications in the 15-15.7 Å range, and several new ones at shorter wavelengths, where previous lines were unidentified. Compared to the previous CHIANTI model, the present one has an increased flux in the 15-15.7 Å range at 2 MK of a factor of 1.9, resolving the discrepancies found in the analysis of the Marshall Grazing Incidence X-Ray Spectrometer observation. It appears that the satellite lines also resolve the long-standing discrepancy in the intensity of the important Fe XVII 3D line at 15.26 Å.