Table of contents

An international meeting with the title "Colloquium on Atomic Spectra and Oscillator Strengths for Astrophysics and Fusion Research" was held at the Physics Department, University of Lund during August 17–19, 1983. The purpose of the meeting was to bring together a limited number of "producers and consumers" of spectroscopic data to discuss the needs for data and the possibilities to meet the demands. The scientific committee for the meeting consisted of T Andersen, Aarhus, O Engvold, Oslo, H W Drawin, Fontenay-aux-Roses, N J Peacock, Culham, and J O Ekberg, S Johansson, U Litzén and I Martinson from the Lund Spectroscopy Group.

The meeting was attended by 60 participants from nine countries. General reviews of the meeting topics were presented in twelve invited lectures, while recent results were reported as short talks or posters. All the invited and the majority of the contributed papers are presented in this Topical Issue of Physica Scripta.

The meeting was sponsored by the University of Lund and the Swedish Energy Research Commission.

A survey is given of the numerous forbidden lines that have recently been discovered in the spectra of tokamak plasmas. They represent magnetic-dipole transitions in three different groups of configurations: 2s2p and 2s²2p^k (k = 1–5), 3s3p and 3s²3p^k (k = 1–5), and 3d^k (k = 1, 2, 8, 9). In the first group (n = 2) a total of 62 lines have been reported for Z = 21–32, in the second group 65 lines (some of which are given as tentative) for Z = 29–42, and in the 3d^k group 8 lines for Z = 40, 42. The identifications in the n = 2 configurations have all been confirmed in elaborate isoelectronic studies of the transition energies. A corresponding investigation of the intervals in the second and third group of configurations is hampered by lack of suitable theoretical data, but provisional checks on the isoelectronic consistency of the identifications are presented.

Topical problems involving the interpretation of line emission from impurities or diagnostic test ions in Tokamak fusion devices are discussed briefly. The particular use of Tokamaks for precision (∼ 10 ppm) wavelength spectroscopy of Δn = 0 transitions in highly ionised atoms with between 2 and 12 electrons is described in more detail. Experimental wavelengths are compared with ab initio theory and with semiempirical isoelectronic extrapolations.

The Tokamak data for ions with Z ⩾ 3-electrons lead to improved isoelectronic extrapolations which reveal errors in the previous data and provide transfer standards of wavelength in the XUV and VUV regions.

Wavelengths of the 2-electron systems are of sufficient accuracy to be sensitive to the 2-photon exchange term in the Breit interaction as well as the 2-electron contribution to the Lamb shift. Experimental wavelengths are in agreement with recent theoretical calculations taking these contributions into account.

In order to elucidate the needs for atomic data, in particular for oscillator strengths, an overview is given over the role of line absorption in the quantitative analysis of high-resolution stellar spectra.

The ultraviolet spectra (λ < 3000 Å) of sharp-lined normal and peculiar stars are rich in information about the abundances of the chemical elements, patterns of nucleosynthesis in our galaxy and the mechanisms responsible for the production of abundance anomalies. The quality of data produced by the International Ultraviolet Explorer satellite is marginally good enough to allow profitable research in this area. A quantum jump in observational capability will occur in 1986 with the launch of the High Resolution Spectrograph on the Space Telescope. In our view, progress in the generation of complete and accurate lists of atomic lines and in the computation or measurement of accurate transition probabilities, for the extraordinarily rich ultraviolet spectra of the lower ionization states of numerous elements is lagging badly behind current and near-term observational capabilities. We illustrate our efforts to theoretically synthesize observed line profiles in the spectra of three sharp-lined, normal, moderately hot stars. These computations are seriously complicated and the results made ambiguous by the low accuracy of the oscillator strengths used (primarily those of Kurucz and Peytremann) and by the incompleteness of finding lists. They nevertheless suggest some interesting conclusions about possible star-to-star variations in elemental abundances.

This paper deals with stellar abundance patterns of the lanthanides. It is divided into two parts. In Part A we investigate the accuracy with which partition functions may be calculated from known energy levels. Corrections for "missing levels" are made, and current "best estimates" are tabulated. The current values are compared with partition functions built into the ATLAS code. The latter are generally rather good, except for Ce III. An increase of about a factor of two in published cerium abundances in many CP stars is indicated. In Part B, the systematics of the behavior of La II through Tm II are reviewed one-by-one. The emphasis is on CP stars, but late-type giants and dwarfs are also discussed. It is concluded that the overall "typical" pattern is tolerably well described by varying degrees of enhancement of the solar pattern, including the odd-even effect. Definite non-solar patterns do occur, however, such as the neodymium-samarium holes and the positive europium anomaly. From these abundance patterns we infer processes, nuclear or non-nuclear, that are distinct from those responsible for the lanthanide composition of the majority of stellar atmospheres.

We have obtained a reliable estimate of the platinum abundance in HgMn stars for the first time. It is as high as log A(Pt) = 6, on the scale log A(H) = 12; the Pt abundance often matches the remarkably high abundances of Hg already known in these stars. Our results are based on calculations of theoretical oscillator strengths which were applied to high resolution spectra taken with the International Ultraviolet Explorer satellite. We have used the method of spectrum synthesis to obtain the abundances. Astrophysical absolute oscillator strengths for the optical Pt II lines are derived based on the spectrum of the star χ Lupi.

Observations made with the International Ultraviolet Explorer (IUE) satellite between 1200 Å and 3000 Å have shown that the spectra of giant and supergiant stars later than about K O are dominated by emission from neutral atoms and singly charged ions. This is in contrast to the spectra of G-type and main sequence stars which show species typical of the solar transition region. The cool star spectra show many examples of photo-excitation, line-leaking and fluorescent processes, some of which are discussed here. There is evidence that the H Ly α and H Ly β lines play an important role in the production of emission from O I, S I and Fe II. Some lines which remain without satisfactory explanations in both the stellar and solar spectra are discussed.

In this review we present further requests to atomic spectroscopists for very accurate atomic data, which should "provide a new tool for the investigation of the solar atmosphere". We insist upon the necessity of using very accurate wavelengths, ionization potentials, partition functions, damping constants and transition probabilities in solar photospheric spectroscopy studies in order to refine our knowledge of the physical conditions, physical processes and abundances of the elements in the solar photosphere.

In Table I, we summarize the present solar abundances of all the elements and we compare these data with the meteoritic values.

We will give a brief review of some existing or planned space missions where spectroscopy of celestial objects is involved. The review does not cover all national or international projects but rather focuses on those where European countries participate through the European Space Agency (ESA).

This paper gives a survey of the spectral distribution of emission lines of Fe II, predicted from a single atomic model. The observed differences between the recorded and the predicted spectrum are discussed in terms of deficiencies of the model and interactions within the emitting plasma. We give a number of illustrative examples of unexpected features with applications to astrophysics. Selective population, due to charge transfer and resonant photo excitation, is elucidated. We also discuss the future need of more laboratory data for Fe II as regards energy levels and line classification.

After a short description of the standard methods to determine f-values and atomic lifetimes, the most important development of the last years - the application of pulsed dye lasers - is discussed and some results are compared. The further development in the field of f-value measurement is described by some typical activities: emission and absorption measurements with thermal light sources and the application of emission and absorption branching ratio measurements. The general accuracy which can be obtained with the different measuring methods is discussed.

The article does not comprehend the whole work which has been done in this field since 1977. There appeared a large number of articles which are not mentioned because of the necessarily limited volume of such a survey. Especially, the treatment of the beam foil technique which yields a lot of lifetimes for singly ionized atoms is omitted completely.

Recent developments in the measurement of lifetimes of excited states in highly ionised atoms using fast ion beam excitation methods are reviewed. The applicability and precision of these methods are discussed in the context of specific measurements and compared with existing theoretical capabilities.

By a combination of hook and emission measurements oscillator strengths of 122 Fe I lines have been determined without any assumption concerning the plasma state. For the hook and emission measurements, a graphite furnace and a hollow cathode have been used as light sources. The relative f-values have been converted to an absolute scale by means of lifetime data found in the literature. Comparisons with other experimental data are made, indicating good to excellent agreement.

The intensities of ultraviolet, spin-changing, "intersystem" lines of low-Z atomic ions are frequently used in determinations of electron densities and temperatures in astrophysical plasmas as well as in measurements of element abundances in the interstellar gas. The transition probabilities (A-values) of these lines, which are about five orders of magnitude weaker than allowed lines, have not been measured heretofore and various calculations produce A-values for these lines that differ by as much as 50%. A radio-frequency ion trap has been used for the first measurements of transition probabilities for intersystem lines seen in astronomical spectra. The measurement procedure is discussed and results for Si III, O III, N II, and C III are reviewed and compared to calculated values. Discrepancies exist; these indicate that some of the calculated A-values may be less reliable than has been believed and that revisions to the electron densities determined for some astrophysical plasmas may be required.

The branching ratios of the 2s²3p ²P⁰_1/2,3/2 levels of C⁺ have been measured by use of a Fourier-transform spectrometer and a vacuum ultraviolet spectrometer. Experimental determination of these branching ratios is of interest, because two-electron transitions are involved, and calculations therefore may have larger than usual uncertainties. The strong multiplets occurring at 658, 284, and 176 nm can be used for determining the interstellar extinction between the red and vacuum ultraviolet spectral regions. The reported branching ratios may also be of use for calibrating the spectral radiometric efficiency of laboratory spectrometers with beam-foil or plasma sources.

Absolute transition probabilities for five lines in the 4s-4p transition array of neutral Argon were determined. The plasma source was an Argon cascade arc in local thermodynamic equilibrium (LTE). The diagnostic of the plasma was based on interferometrically measured electron densities. The measured line emission coefficients were corrected for line-wing losses and for optical thin emission. Furthermore the inhomogeneities of the arc plasma were taken into account. Moreover branching ratios for six of the ten 4p levels were determined with a pen-ray.

Using the beam-foil method we have determined the lifetimes of the nd ²D (n = 4–6) and 4s5s ⁴D levels in Cu I and of the 4p ³D, ³F, 4f ³H and 5s ³D levels in Cu II. Special emphasis was placed on the problems of measuring comparatively long lifetimes in heavy atoms with the beam-foil excitation method and corrections of systematic errors were investigated. The final data are in agreement with theoretical values, based on the Coulomb approximation.

Extraction of accurate values of autoionization rates from broadened emission lines in optical spectra is discussed. Results for autoionizing states in He I and Li I are given and compared with theory. Spin-induced autoionization is proposed to cause the discrepancy between theoretical and experimental lifetimes for some Li I quartet levels.

Previously a relative efficiency calibration curve for a 2.2 m McPherson grazing incidence monochromator has been established using cascade relations in Be-like heavy ions. The curve has now been confirmed and extended to cover the wavelength range 10 nm ⩽ λ ⩽ 80 nm using branching ratios for Δn ≠ 0 transition in Li-like ions. A summary of various measurements applicable to problems of astrophysical and plasma physics interest is given.

An up-date is presented of a previous review by the author of atomic transition probability calculations published in 1969. Attention is again restricted, in the main, to allowed electric dipole transitions. Attention is in particular paid to the construction of accurate atomic wave-functions, to the random phase approximation and to recent progress in relativistic calculations. Comparative results for the resonance transition of beryllium-like ions are discussed by way of illustration.

A review of relativistic atomic structure calculations is given with a emphasis on the Multiconfigurational-Dirac–Fock method. Its problems and deficiencies are discussed together with the contributions which go beyond the Dirac–Fock procedure.

Both, astrophysics and theoretical atomic physics, are pursued by the same group at University College London under the auspices of Professor M J Seaton. Through his lifelong research in both fields Mike Seaton could recognise those areas where progress in astrophysics was hampered by lack of sufficiently reliable atomic data and where further research in atomic physics was necessary to make such data available.

This paper reports on methods for the calculation of atomic data employed at University College London, giving the essence of the approximations involved and some guidance as to the accuracy that can be expected from the data so produced.

General expressions (both relativistic and non-relativistic) for the electric multipole transition operators, having not specified values of the gauge condition of the electromagnetic field potential, are presented. The dependence of oscillator strengths for electric dipole transition on the gauge is studied. In the general case it has the form of a parabola. The coefficients of the equation, describing this curve, as well as the values of the gauge for which the oscillator strength equals zero are expressed in terms of the length and velocity forms of the electric dipole transition operators. Transitions in Be-like ions are considered and the results of the calculations are compared with experimentally determined oscillator strengths.