Periodic table of elements

Click on a yellow case corresponding to the chosen element and then on an ionization degree. There are no data on the non-coloured cases. Then a new page appears, requiring your detailed choice.

Line widths and shifts table

Column 1

Perturber density N in cm-3

Column 2

Lower level or lower term

Column 3

Upper level or upper term

Comments to columns 2 and 3

When the fine structure splitting is small, namely if the difference between energy levels of the same multiplet is small compared to the distance to the next level linked by an allowed transition, all the fine structure lines of the same multiplet have the same width and shift. In that case the data are given for the multiplet only and for an average wavelength for the whole multiplet. If needed, the width value for a particular line within a multiplet can be obtained from :
Wline = Wmult l2line / λ2mult
Idem for the shift

Column 4

Multiplet when it is available


It is the multiplet number generated on line from the NIST Atomic Spectra Database *. Therefore we have chosen not to select any wavelength range, because the multiplets numbers vary if the the selected wavelength range varies. In addition, the data for multiplets as a whole are only generated in the NIST Atomic Spectra Database if all fine structure components are known and if it is in LS coupling. * NIST Atomic Spectra Database (version 3.1.5), [Online].
Available: Ralchenko, Yu., Kramida, A.E., Reader, J., and NIST ASD Team (2008), National Institute of Standards and Technology, Gaithersburg, MD.

Column 5

Wavelength in Å

Comment to column 5

These wavelengths are calculated wavelengths with the computer code. In particular, they are averaged over the multiplet when multiplet data are given. The data provided in columns 9-13 and following are obtained with these wavelengths. So, for obtaining data for other wavelengths (measured ones for instance), one has to use the formulae : W1 = Wl21 / l2

Column 6

Parameter C for the validity condition of the isolated line approximation

Comment to column 6

The isolated line approximation is valid for a kind of perturbers a (a = electrons, protons, He II, ...) if C/Wa is higher than the corresponding perturber density. For a perturber density N lower than Nl (cm-3)=C/Wa,the line can be treated as isolated even if a weak forbidden component due to the failure of this approximation remains in the wing. Wa is the full width at half-maximum intensity given in the coresponding following columns (9, 11, 13...). See the "Introduction" for definition of C and for the validity condition of the isolated line approximation.

Column 7

Temperature T in Kelvin

Column 8

A (quasistatic parameter for neutral atoms, cf. Introduction for details) if available

Column 9

Full width at half intensity We in Å (electron colliders)

Column 10

Shift de in Å (electron colliders). A positive shift is towards the red, a negative one is towards the blue

  • Empty cells which are not preceded by an asterisk mean that the data are not available
  • Empty cells which are preceded by an asterisk mean that the impact approximation is not valid, because NV > 0.5 (cf. Introduction for details), and thus the coresponding data are not provided
  • Non-empty cells preceded by an asterisk mean that the impact approximation reachs its limit of validity, 0.1 < NV ≤ 0.5 (cf. Introduction for details) (cf. Introduction for details)

When the shift is negative, due to the additional minus sign, only the width value is marked with the asterisk.

Columns 11 and 12

same as columns 9 and 10, but for protons colliders (subscript p)

Columns 13 and following columns

same as columns 9 and 10, but for other ion colliders (other corresponding subscripts)


Some widths and shifts appear at medium and not at low densities. This means that they are proportional with the density. Thus data at low densities can be deduced from those at medium densities by linear interpolation with the perturber density.

Fitting coefficients table

This supplementary table gives fitting coefficients with the temperature using the following equations :

  • log(w)=a0 + a1 log(T) + a2 (log(T))2
  • d/w = b0 + b1 log(T) + b2 (log(T))2

The fitting coeficients are obtained through a least-square method. These fitting formulae are quite accurate(1), especially for the width, N.B. This fitting must be used inside the temperature interval of the table of widths and shifts.

(1) : Sahal-Bréchot, S., Dimitrijević, M.S., & Ben Nessib, N. 2011, Baltic Astronomy, 20, pp. 523-530