Iron-manganese-neon spectrum
Individual researchers
and research pages:

Sven Huldt
Henrik Hartman
Hampus Nilsson

PhD students

Lund Observatory
Local events

Lund University

The FERRUM Project

The FERRUM project is an international collaboration in atomic astrophysics initiated by S. Johansson, Lund. It aims at the production of reliable oscillator strengths (f-values) or transition probabilities (A-values) for spectral lines of singly ionized Iron Group Elements of astrophysical significance.
For allowed transitions the goal of the project is to provide:

  • a set of lines in the satellite ultraviolet (l < 3000 Å) region
  • a set of lines in the optical (l > 3000 Å) region
  • lines with a large spread in excitation energy

    The project also includes atomic data for forbidden transitions based on measurements of radiative lifetimes of metastable states. The project is presented and described in Johansson et al, Physica Scripta T100, 71-80, (2002).

    The three requirements listed for the allowed transitions are motivated as follows.
    Astrophysical spectra may contain absorption and/or emission lines superimposed on a continuous spectrum produced by stellar blackbody radiation (BBR). In general, stellar atmospheres give absorption lines in the continuous spectrum, whereas gas clouds around or between stars give a pure emission line spectrum.
    Stellar absorption lines are used to determine the chemical abundance in the star's atmosphere, and that procedure requires gf-values. Stellar emission lines are used for temperature and density diagnostics of diluted plasmas, and they might also be the source of information about the concentration of various chemical elements. Stellar spectra are recorded either from the ground (l> 3000 Å) or from space (down to about 900 Å), and they often lead to independent analyses. This explains the need for separate data in the optical and satellite UV wavelength regions.

    Most abundance analyses assume local thermal equilibrium (LTE) in homogeneous parallel layers of the stellar atmosphere. Thus, by using lines from energy levels of quite different excitation energy in the abundance analysis one can test the validity of the LTE assumption in the stellar model atmosphere by the Boltzmann distribution law. The abundance derived from different spectral lines of a given ion should give a level population that is proportional to the exponential factor (e-hn/kT) in the Bolzmann formula. As we should see later the utilization of lines from high-excitation levels in the abundance analysis may be of great significance, as intrinsically strong lines (large f-values) may result in reasonably faint features in the stellar spectrum and obey the criteria for linear curve of growth analysis.

    t + BF => A-value => f-value

    t : Radiative lifetime - measured by laser induced fluorescence technique at LLC

    BF: Branching fractions from relative intensities (I), where BFik is proportional to Iik, measured using Fourier Transform Spectroscopy Aik

    BFik= Aik/SAil and ti = 1/SAil => Aik = BFik/ ti

    f: Calculate f-values using different theoretical techniques
    Working scheme
    The various steps in the working scheme used to achieve the goal of the FERRUM Project can be listed as follows:
    • Measure f-values for some lines in a given transition array (TA) with a spread in excitation energy
    • Calculate all lines in the same TA using the orthogonal operator technique
    • Compare experimental and theoretical values and estimate the uncertainty of the calculated values
    • Make the list of lines in the TA array as complete as possible and add error bars to all f-values
    • Insert the f-values for all the lines in a synthetic spectrum and compare to a stellar spectrum
    • Investigate whether possible deviations in the previous comparison depend on the synthetic spectrum f-value, model) or the stellar spectrum (blends)
    • Adjust the line list by adding/removing the complementary information obtained from the stellar spectrum.

    Even if we stress the necessity of measurements it is important to underline the significance of incorporating theoretical data in the line lists. Calculated f-values will always constitute the bulk of the database, and the experimental values will be used to normalize them and to assess their accuracy.


    Visiting address: Sölvegatan 27
    Phone: +46 46 22 27300, Fax: +46 46 22 24614
    Publisher: Leif Lönblad