. : http://star.arm.ac.uk/~ambn/mzbpap.ps
: Fri May 31 18:25:41 1996
: Mon Oct 1 21:12:08 2012
:
: south pole
Versions of CCP7 SYNSPEC code
M. Zboril
Armagh Observatory, College Hill, Armagh, BT61 9DG, N. Ireland
Abstract. A general spectrum synthesis CCP7 program SYNSPEC was
modified for the purposes to analyse the atmospheres of CP stars as well
as very late type stars. Modifications result into two versions of the code:
for PC's and computers basically with UNIXbased operating systems.
1. Introduction
Program SYNSPEC is a general spectrum synthesis program. CCP7 version
assumes that the model atmosphere is given. Subsequently, under the stellar at
mosphere approximation by planeparallel geometry and radiative, hydrostatic,
statistical equilibria it operates with differential equation form of radiative trans
fer equation to produce normal synthetic spectrum, detailed line profiles of a few
individual lines, emergent flux in continuum, or simply identification table only,
but all these outputs are optional. The more detailed description and a guide
how to work with can be found in Hubeny, Lanz and Jeffery (1994).
The program is very flexible and basic set up is for early type stars but unfortu
nately not primarily for study CP stars. Consequently, since having dealt with
CP stellar atmospheres a time, some modifications were necessary and are to be
described. Recently, another modifications regarding late type stars have been
in progress. To give an overview a previous status will be mentioned altogether
with latest upgrades labelled by an ''+'' symbol for clarity.
2. Input models and opacity sources
For a number of applications both various input models and opacity sources are
necessary. In comparison with previous set up, recent one reads:
1. Models
ffl NLTE model, published or NLTE code Tlusty
ffl LTE Kurucz (1992) models
ffl Allard and Hauschildt (1995) model grid +
2. Opacity sources
ffl standard and some nonstandard opacity sources, Hubeny, Lanz and
Jeffery (1994)
33

Figure 1. A portion of low resolution spectrum for M1 dwarf. Solid
lineAllard and Hauschildt (1995).
ffl photoionization cross sections +
ffl He nonstandard opacity +
ffl an ODF for Allard and Hauschildt (1995) model grid +
NLTE program Tlusty (Hubeny & Lanz 1995) is another CCP7 library code,
the model atmosphere outputs being fully consistently read by Synspec code.
Since boundfree opacities from explicit levels are treated and understand as
standard opacity sources resulting thus to the total absorption coefficient we
have to describe new photoionization cross sections in this section. Apart from
some of them already coded, Opacity Project crosssections, we contributed here
coding the following sources: Reilman and Manson (1979), Hofsaess (1979),
Vernazza, Avrett and Loeser (1976) and SIMBAD Topbase database outputs.
At a present stage, ODFs for late type stars require to switch off all previous
opacities and since many of them have 2 A spacing only low resolution spectra can
be evaluated properly (Figure 1) and identification table for particular segments
of wavelengths.
3. Other treatments
The program allows a quite general choice of chemical species, ionization degrees,
energy levels and atomic transitions to be dealt with, nevertheless, up to atomic
number 30. Owing to properties of (CP) stellar spectra and handling with them,
some other treatments to the program were naturally required. Basically, though
34

these additional modifications were initialized by study CP stellar spectra, they
have quite general applications.
ffl chemical elements up to Z = 82 +
ffl vertical stratification of stellar atmosphere chemical elements +
ffl iterative abundance process +
ffl horizontal stratification +
ffl partition function option (Mn and Ga species) +
There is no reason for the cutoff of chemical species at Z=82. On the con
trary, it is an intermediate step and further implementation is trivial and should
follow. Vertical stratification of chemical elements is expected from diffusion
theory (Michaud 1970) as a competitive action of gravitational and radiative
forces. Set up for vertical stratification of all chemical elements is 3 layers and
once again, including a different number of layers is extremely simple. Vertical
stratification is involved by means of depth points (not optical depths). Figure 2
illustrates vertical stratification of helium.
I have developed an iterative process to find the abundance corresponding to
a given equivalent width and given accuracy. In practice, the original version
is designed for reverse process, for a given abundance the equivalent width of
transition can be computed properly. In an extreme situation, e.g. flat part of
curve of growth and very high accuracy required, we might encounter converg
ing problems when an abundance is computed from given equivalent width. To
accelerate, a Ng (1974) acceleration can be switched on at a moment, however,
interval dividing method may (or may not) be more efficient. Ng acceleration
saves information from previous iterations, the accelerated nth iterate is taken
as a linear combination of the three previous ones and, in fact, may accelerate
iterative abundance process since original procedure keeps only the previous it
erate. We note that at this stage to switch both abundance iterative process and
vertical stratification options is not recommended.
Since vertical stratification of elements is available, there is natural need to in
troduce horizontal stratification. Therefore, disk integration process is ready. In
more detail, horizontal stratification is treated as intensity (or flux) variations
through stellar sphere and transformed to stellar disk to be integrated and to
obtain usual observed residual flux. Thus, horizontal stratification is related to
the brightness and not to physical processes generated by the presence of 'abun
dance' or 'cool' spot at stellar surface.
Partition functions for manganese and gallium are available from the following
sources: Irwin (1981), Gray (1992) and Traving, Bashek and Holweger (1966),
the latter being coded already, as one of the contributions to minimize the de
pendence on atomic physics and since vertical stratification of these elements in
the atmospheres of CP3 stars is expected.
There are still some other minor modifications done in program generated
by a long term getting touch with it. We may mention wavelength Stark shifts
for all elements, rectification process if mode detailed line profile is switched on,
35

Figure 2. HeI 4471 A line profile. Kurucz (1992) model atmo
sphere used, solid linesolar helium abundance (n(He)=0.1),dots
nonsolar (0.06) abundance, dashed line0.06 abundance up to depth
point 48, solar abundance up to depth point 54, abundance 0.11 for
rest depth points.
Zeeman splitting as a pseudoturbulence term in Doppler width calculation and
some others. More specifically, as concerns Stark constants, there are several
options coded for helium, for rest elements the relevant quantities are read.
4. Concluding writeup
Generally, the original philosophy adopted in program was to include as many
atomic data as possible and to give the user flexibility in modifying and changing
some or even all of the atomic parameters. This philosophy is maintained and
only several control switches are added to input data files. However, there
are two versions of the code, PC version running under DBOS environment
since Salford University compiler was used and a version suitable for computers
with UNIXbased operating systems. Unfortunately, input data files are quite
different as well as modifications. PC version includes only iterative abundance
process and treatment of new chemical elements.
Potentional user is recommended to use version running under UNIX systems
as it is very close to the version in CCP7 library and described in Hubeny, Lanz
and Jeffery (1994).
36

Figure 3. Convergence of iterative abundance process. Percentual
accuracy 0.1 required and reached after 8 iterations.
Acknowledgments. I am deeply indebted to Armagh Observatory for travel
grant and the Austrian Academy of Sciences for financial support provided. My
thanks go also to Ivan Hubeny for remarks on manuscript.
References
Allard, F. & Hauschildt, P. 1995, ApJ, 445, 433
Gray, D. F. 1992, in The Observations and Analysis of Stellar Photospheres:
Cambridge Univ. Press
Hofsaess, D. 1979, Atomic Data and Nuclear Tables, 24
Hubeny, I. & Lanz, T. 1995, ApJ, 439, 875
Hubeny, I., Lanz, T. & Jeffery, C. S. 1994, CCP7 Users Guide
Irwin, A. W. 1981, ApJS, 45, 621
Kurucz, R. L. 1992, Rev. Mex. Astron. Astrofiz., 23, 45
Michaud, G. 1970 ApJ, 160, 641
Ng, K. C. 1974, J. Chem. Phys., 61, 2680
Reilman, R. E. & Manson S. T. 1979, ApJS, 40, 815
Traving, G., Bashek, B., Holweger, H. 1966, Tabellen fur die Berechnung von
Zustandsummen, Abhandlungen aus der Hamburger Sternwarte, Band
VIII, No.1, p. 3
Vernazza, J. E., Avrett, E. H. & Loeser R. 1976, ApJS, 30, 1
37