Solar Spectrum Synthesis. I. A Sample Atlas from 224 to 300 nm
Abstract
We have developed sophisticated computer programs for determining solar and stellar atmospheric structure through the analysis of spectra. These programs allow us to treat the spectrum as a whole and to draw much stronger conclusions than would be apparent from individual spectral features. For a given LTE or non-LTE model atmosphere, the programs compute the emergent flux or the specific intensity at up to 20 angles. The spectrum can be broadened by macroturbulence and rotation; it can be modified by transmission through the Earth's atmosphere; it can be convolved with the instrumental profile; and it can finally be plotted together with the observed spectrum with each line labeled. In the opacity calculation, the lines are broadened by radiative, Stark, and van der Waals damping, and they can have isotopic and hyperfine splitting, autoionization, partial redistribution, or be merged into a continuum. The departure coefficients for ions treated in non-LTE in the model atmosphere calculation can be used in the spectrum synthesis programs for all lines of these ions, and highly ionized lines can be treated in the coronal approximation. The model atmosphere can have depth-dependent doppler shifts corresponding to large-scale motions. Using the Vernazza, Avrett, and Loeser models for the average quiet sun, we have computed theoretical solar spectra that include all available atomic and molecular line data. In this atlas we compare with the best available observed spectra in the 224- to 300-nm wavelength range, namely, the Kohl, Parkinson, and Kurucz (Harvard) center and limb rocket spectra in the range 224 to 300 nm; the Allen, McAllister, and Jefferies (Hawaii) disk center rocket spectrum for 268 to 293 nm; and the Brault and Testerman disk center spectrum taken at Kitt Peak for 294 to 300 nm. We also compare the observed spectra with each other. The existing spectra are noisy and do not have adequate resolution, so that it is difficult or impossible to identify weak features, to resolve blends, to study velocity fields, to search for variability; or to do any of the projects that can be routinely done in the visible where high quality spectra are available. Because the ultraviolet spectrum provides diagnostics for the upper photosphere, the temperature minimum, and the chromosphere, our understanding of these regions of the solar atmosphere is seriously impaired. One-half the lines in the observed spectrum are not identified. The overall level of the calculation lies considerably above the observed. The discrepancy is caused mainly by missing atomic and molecular lines and possibly by errors in the measured continuum opacities. Laboratory spectral analyses are seriously incomplete, especially for the iron group atoms and for high J and V molecular levels of CO, SiO, and the hydrides. Any model atmosphere or non-LTE rate calculation that depends directly on available ultraviolet opacities should not be trusted.
- Publication:
-
SAO Special Report
- Pub Date:
- May 1981
- Bibcode:
- 1981SAOSR.391.....K