IV. Auxiliary Programs

a) BSMOOTH2: BSMOOTH2 takes the binary output  (for normalized intensity - either integrated disk or normalized specific intensity) from SPECTRUM and smooths the output to a specified resolution by convolving with a Gaussian profile.

BSMOOTH2 can be used either in a command-line format (which makes inclusion in batch files easy), or will prompt the user for the necessary information.  In the prompt mode, BSMOOTH2 gives the following prompts:

Enter name of input file >

Enter name of output file >

Enter output resolution in Angstroms >

Enter spacing in Angstroms of the spectrum in the output file >

All other information, such as the spacing in the original file is kept in a header in the binary file.

Use the following format for the command-line mode:

bsmooth2  infile outfile resolution spacing

BFLXSM2 is a very similar program which may be used on the binary output of SPECTRUM (for absolute flux, or non-normalized specific intensity) and  has identical prompts.  Please note: the spacing in the output file must be an integer multiple of the spacing in the input file!!!  For instance, if you compute the spectrum with a spacing of 0.02A, an appropriate output spacing might be 0.1A, but not 0.05A!  This goes for BSMOOTH2, BFLXSM2 and  SMOOTH2 (which works on ascii files).  If this rule is not obeyed, bizarre results can be expected!

A program, quite similar to BSMOOTH2, called CUSTOMSM, will smooth the binary output of SPECTRUM using a mixed (i.e. pseudo-Voigt) profile instead of a Gaussian profile.  This is quite useful when it is necessary to duplicate the more complicated line-spread profile of  a real spectrograph.  It is available upon special request.

The following are programs which are related to SPECTRUM, and may be of some use to the user. These programs, except for XMK16 and XMKF are included in the UNIX distribution of SPECTRUM.   However, XMK16 and XMKF (warning - still an experimental program!) are available on the same ftp site as the UNIX/Linux version of SPECTRUM.  Check under the xmk directory under the directory for SPECTRUM.

b) BLACKWEL: BLACKWEL computes "Blackwell diagrams" (cf. Blackwell and Shallis 1979) for a given set of spectral lines with measured equivalent widths. Blackwell diagrams are used to determine the elemental abundances and microturbulent velocity for a star, given an appropriate atmosphere model. That is to say, Blackwell diagrams give the same information as the classical curve of growth method, but without the debilitating effects of the many invalid assumptions.  The set of spectral lines used to construct the Blackwell diagram should be placed in a file in which each line has the following format (very similar format to the lines in luke.lst):
wavelength   code    E(lower)  E(higher)  log(gf)   df    TT    EqW    source

Here the wavelength, code, E(lower) and E(higher) - both in inverse cm, log(gf), the damping factor, df and the transition type, TT, are in exactly the same format as in luke.lst.  The equivalent width, EqW, measured in a stellar spectrum (such as the solar flux atlas), is expressed in milliangstroms.  It is required to place a "source", a short string of 5 or fewer characters at the end of each line.  An example file feI.eqw is given in the distribution.  It can be used in conjunction with the model 67244k2p00.mod to test BLACKWEL.

The output from BLACKWEL has the format:

wavelength  code  vt  Atot  AH

where vt is the microturbulent velocity, Atot is the abundance of the element represented by code on the abundance scale used by SPECTRUM, whereas AH is the abundance scale based on Hydrogen.  See  section II  for more details.  How does one use this output?  BLACKWEL calculates the abundance (Atot, AH)
of the element in question which yields the equivalent width for the given line for a range of microturbulent velocities (which you input into BLACKWEL).  If
you plot  the abundance versus the microturbulent velocity the result is a sloping line going from high abundances for low microturbulent velocity to low
abundances for high microturbulent velocities.  Different spectral lines will give differently sloping lines; the intersection (or the region of least confusion)
will give the optimal microturbulent velocity for the star and the abundance of the element in question.  For best results, use isolated lines with well-determined log(gf)'s with strengths ranging from weak to medium.  Strong lines will not give satisfactory results.

In releases of SPECTRUM with version numbers 2.65 and later, a Tcl/Tk graphical interface for BLACKWEL is now available.  See the section on Graphical User Interfaces for more details and for some tips on how to use and interpret the results from BLACKWEL.

c) GFADJUST: GFADJUST is the program which was used (see section II.a) to "adjust" some of the gf values in the line file, luke.lst, by inverting the Solar atlas. GFADJUST is not limited to inverting the Solar atlas, but can be used with any other star which can be represented by an LTE stellar atmosphere model.

To compute the log(gf) for a given line, measure the equivalent width of the line in the spectrum of a star with well-determined physical parameters (the sun!) and an appropriate Kurucz model. Make a line list file with exactly the format used by BLACKWEL;  GFADJUST will calculate log(gf) values for all of the lines in the file.

d) LINES: will run in exactly the same way as SPECTRUM, but instead of outputting a synthetic spectrum, will print out a file containing the equivalent width and the central depth of the lines in the spectral range being computed.  This can be very useful for identifying specific lines in the synthetic spectrum.

e) VSINI: will rotationally broaden the binary output of SPECTRUM and output the rotationally broadened spectrum to a binary file of the same format.  This binary file can then be further smoothed by BSMOOTH2 or BFLXSM2.  The program AVSINI will handle ascii output from SPECTRUM.  For VSINI, supply the projected rotational velocity for the spectrum (vsini) and an appropriate value of the limb-darkening parameter.  You can read more about the limb-darkening parameter in David Gray's stellar atmosphere book, but a value of 0.6 will suffice for most cases.  For AVSINI, you will also have to supply the spacing in the input ascii spectrum and the spacing in the output file.  It is a good idea to use the same value for both.  Both VSINI and AVSINI can be run on the command line; the order of the input parameters is the same as in the prompts.

f) MACTURB: can be used to introduce macroturbulent broadening into the spectrum.  This program uses the radial-tangential formulation of David Gray.  MACTURB works only on ascii spectra.  It should be used before smoothing the spectrum with SMOOTH2.

f) XMK16/18/19:  XMK16/18/19 are  X-windows based spectral analysis programs.  They are the descendent of the DOS programs MK and MK32, both written originally for performing MK spectral classification on digital spectra. XMK16 has a number of other capabilities, however, including a rectification capability, and the ability to measure equivalent widths. Equivalent widths are measured by direct trapezoidal integration of the line profile.  A sister program (XMKF, available by special request) can measure equivalent widths  by performing a nonlinear least squares fit of a number of mixed (i.e. pseudo Voigt) profiles to a blended line profile (see section IIa above for details).