SPECTRUM v1 - Examples
Equivalent widths and line profiles for individual linesIn order to calculate the line profile of a given absorption line for atomic species Z with a wavelength LAMBDA (in Angstroms) and a microturbulent velocity VT (in km/s). The following input illustrates the options currently available within SPECTRUM
1 0 0 3 ! nmod, nskip, wavmax, nsource (1) 0 ! mode 14 5056 8.5 10 ! z, lambda, [Z], vt 14 5056 1.E-4 10 ! z, lambda, n_z, vt (2) -14 5056 1.E-4 10 ! -z, lambda, n_z, vt (4) 0 0 0 0 ! to finish input
(1) The input model may contain more than one model (i.e. it may be an entire grid of models). Spectrum will cycle through this grid NMOD times, omitting NSKIP-1 models from the
input for each model treated. However, the input data following MODE must be repeated NMOD times in order to obtain equivalent results for each model in the grid. WARNING CSJ has never used multiple model inputs and regards this feature as obsolescent.
(2) The elemental abundance is entered as a number fraction, either as a logarithm normalized such that [H] = 12.00 in hydrogen-normal stars, or as a fraction such that n_H = 1.0 in hydrogen-normal stars (an additional note on what happens for stars in which hydrogen is not normal will follow).
(4) SPECTRUM searches the linelist for the line belonging to species Z which lies closest to and within WTOL of lambda. It computes the total equivalent width due to this line and to any blending line belonging to the same 'multiplet' that lies within WTOL of the principle line. Normally WTOL=1.0 Angstrom, but if Z is negative, then WTOL=0.1 Angstrom.
Elemental abundances from line equivalent widthsIn order to calculate the abundance of an atom giving rise to a given absorption line for atomic species Z with a wavelength LAMBDA (in Angstroms) and a microturbulent velocity VT (in km/s).
Note (3). If the abundance is entered as a negative number, it is assumed to be the line equivalent width. Spectrum will apply a Newton-Raphson iteration to derive the elemental abundance which will reproduce this equivalent width for the current model atmosphere.
1 0 0 3 ! nmod, nskip, wavmax, nsource (1) 0 ! mode 14 5056 -120 10 ! z, lambda, ew, vt (3) 0 0 0 0 ! to finish input
The following replicates the above except that each line calculation will be repeated NAB times using abundances AB(1:NAB) and NVT times using microturbulent velocity VT(1:NVT). Line profiles are not saved and only a summary of the equivalent widths is provided in the output file.
Curves of growth for individual lines
1 0 0 3 ! nmod, nskip, wavmax, nsource (1) 20 ! mode 3 5 ! nab, nvt 14 5056 ! z, lambda 7. 8. 9. ! ab(1:nab) 0 2.5 5 7.5 10 ! vt(1:nvt)
This option may be extended to cycle over a number of lines NL of atoms Z(1:NL) with wavelengths LAMBDA(1:NL)
1 0 0 3 ! nmod, nskip, wavmax, nsource (1) 21 ! mode 3 5 ! nl, nab, nvt 14 5056 ! z(1), lambda(1) ... 14 4575 ! z(nl), lambda(nl) 7. 8. 9. ! ab(1:nab) 0 2.5 5 7.5 10 ! vt(1:nvt)
If specific intensities are required, the number of values of cosine angles at which they will be calculated must be specified immediately after initial options. For example
1 0 0 7 ! nmod, nskip, wavmax, nsource (1) 11 ! number of cosine angles 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.001 ! cosine angles 0 ! mode 14 5056 7. 5. ! z, lambda, ab, vt 0 0 0 0
It is probably a good ide to keep the requirments simple, i.e. a single line profile or a single spectral
synthesis - I have no idea what would happen in more complicated cases! The output is written as
an ascii file "spectrum.000" in terms of normalized total flux and normalized specific intensities, followed by
actual fluxes and intensities (F_nu and I_nu). The normalized emergent flux continues to be written to
"dipso.001...." as before.
Synthetic spectra over substantial wavelength intervalsThis is the state of the art option (which usually means it is the most bug-ridden), and is comparatively liable to be upgraded, debugged and generally messed around. For a given model atmosphere, the user may either define a new wavelength grid, given a linelist, or may use a previously defined wavelength grid and linelist.
In the original version (MODE=10) the newly-generated grid assumes solar abundances for all species. These can be modified in the stored grid for subsequent use. This grid can also be edited to compute spectra for multiples sets of abundances, but the i/o preparation is done line by line.
With MODE=11, the abundances of all species can be defined when the grid is setup (solar values are still the default). The wavelength grid definition is sufficiently improved (and computers are now so much faster) that I generally always set up a new grid each time and the following dialogue is pretty well standard.
1 0 0 3 ! nmod, nskip, wavmax, nsource (1) 11 ! mode 1 ! define grid (y/n) 4000 4500 ! wstart, wend 14 8.0 ! z, [Z] ! set individual abundances 16 7.5 ! z, [Z] 0 0 ! ....... ok 2 ! nvt 0 10 ! vt(1:nvt)
WARNING: because of the manual overheads involved in maintaining the code and in using these optios, MODE=10 and the practice of saving the wavelength grid and linelist for manual editing should be regarded as obsolescent.