V. Using SPECTRUM synthetic spectra to interpret observed stellar spectra.

SPECTRUM is a powerful program, but it is still imperfect, and should not be trusted completely. Please keep in mind the following points.

a) SPECTRUM computes the spectrum using the assumption of LTE.

b) SPECTRUM does not include effects of chromospheres.

c) In addition to all of its own inadequacies, SPECTRUM can compute no better of a spectrum than is allowed by the input model.  Stellar atmosphere models are not perfect!

d) The log(gf) values used in luke.lst are not necessarily to be trusted. I am continually trying to improve them - I am now starting on the long, long process of comparing the observed solar spectrum with the computed solar spectrum, and there is much work to be done on oscillator strengths, damping coefficients, etc. Please critically evaluate all of these parameters before you base any conclusions (abundances, etc) on a log(gf) in luke.lst

e) If you use SPECTRUM to identify interesting features in a stellar spectrum (such as line ratios, etc) which are, say, luminosity or temperature sensitive, please identify those lines in the luke.lst file before any conclusions are made! Some of the lines in luke.lst are fictitious, and may be acting in a contrary way to nature!!

f) The synthetic spectrum, as it comes out of SPECTRUM can, in only a very few cases, be directly compared with an observed spectrum. An observed spectrum has been passed through a spectrograph, and has been convolved with the line spread function characteristic of that spectrograph. Quite often, this line spread function is close to a Gaussian, and so the use of , BSMOOTH2, etc. (which convolve the synthetic spectrum with a Gaussian of your choice) may be quite adequate. In other cases, you may wish to determine the line spread function of the spectrograph directly, and write a program which will perform the convolution. This is a very important point! Please also bear in mind that your spectrograph may suffer from scattered light as well. Scattered light will tend to "fill in" the bottoms of strong lines.

g) Please remember to compute the synthetic spectrum with a sufficiently small wavelength step - I recommend 0.02 A/step, or smaller if a really critical comparison of line strengths is required. I have seen people use a wavelength step of 0.1 A and then expect the computed spectrum to give good results!!

Below find references to papers refered to in this documentation

REFERENCES

Barnard, A.J., Cooper, J. and Smith, E.W. 1974, J. Quant. Spectrosc. Radiat.
   Transfer, 14, 1025-1077.

Blackwell, D.E. and Shallis, M.J. 1979, MNRAS 186, 673.

Corliss, C.H. and Bowman, W.R. 1962, "Experimental Transition Probabilities
   for Spectral Lines of Seventy Elements", National Bureau of Standards
   Monograph 53.

Danylewych & Nicholls 1977, Proceedings of the Royal Society of London.

Fuhr, J.R., Martin, G.A., Wiese, W.L. and Younger, S.M. 1981, "Atomic
   Transition Probabilities for Iron, Cobalt and Nickel (A Critical Data
   Compilation of Allowed Lines)", Journal of Physical and Chemical
   Reference Data, Vol 10, No. 2, pp 305-565 (part of reference J).

Grevesse, N. and Sauval, A.J. 1973, A&A 27, 29.

Griffin, R. and Griffin, R. 1979, "A Photometric Atlas of Procyon",
   Cambridge, England.

Kato, K. and Sadakane, K. 1982, A&A 113, 135.

Kovacs, I. 1969, "Rotational Structure in the Spectra of Diatomic Molecules",
   (New York: American Elsevier).

Kurucz, R.L. 1991, in IAU Symposium 149, The Stellar Populations of Galaxies,
   Angra do Reis, Brazil.

Kurucz, R.L. and Peytremann 1975, Smithsonian Astrophysical Observatory
   Special Report No. 362.

Moore, C.E., Minnaert, M.G.J. and Houtgast, J. 1966, "The Solar Spectrum,
   2935 A to 8770 A", NBS Monograph #61, U.S. Government Printing Office,
   Washington D.C.

Wiese, W.L. and Fuhr, J.R. 1975, "Atomic Transition Probabilities for
   Scandium and Titanium (A Critical Data Compilation of Allowed Lines)",
   Journal of Physical and Chemical Reference Data, Vol 4, No. 2, pp 263-352.
   (part of reference J)

Wiese, W.L., Smith, M.W. and Glennon, B.M. 1966, "Atomic Transition
   Probabilities, Vol I: Hydrogen through Neon", National Standard Reference
   Data Series, NSRDS-NBS 4.  (reference N1)

Wiese, W.L., Smith, M.W. and Miles, B.M. 1969, "Atomic Transition
   Probabilities, Vol II: Sodium through Calcium", National Standard
   Reference Data Series, NSRDS-NBS 22.  (reference N2)

Younger, S.M., Fuhr, J.R., Martin, G.A. and Wiese, W.L. 1978, "Atomic
   Transition Probabilities for Vanadium, Chromium, and Manganese (A Critical
   Data Compilation of Allowed Lines)", Journal of Physical and Chemical
   Reference Data, Vol 7, No. 2, pp 495-629. (part of reference J)