Baltic
Astronomy,
vol. 5, 589-590,
1996.
ON T H E A C C U R A C Y OF QUANTITATIVE S P E C T R A L
CLASSIFICATION OF STARS
V. Malyuto 1 and Th. Schmidt-Kaler 2
1
Tartu Astrophysical
Astronomisches
Germany
2
Observatory,
Institut
Toravere, EE2444>
der Ruhr-Universität
Bochum,
Estonia
D-44780
Bochum,
Received October 15, 1995.
Abstract. We describe a scheme of automated quantitative spectral classification of F-K stars which may be applicable to studies of
galactic structure and evolution. The accuracy of our spectral classification is compared with the corresponding estimates in the case of
a photometric classification.
K e y words: methods: observational - techniques: spectroscopic stars: fundamental parameters (classification)
The automated quantitative classification technique has been developed by us on the base of spectra obtained with the photoelectric
spectrum scanner designed at Astronomisches Institut at Bochum
(resolution about 1 nm, the wavelength range about 350-850 nm).
The scanner has been attached to the 61 cm Cassegrain telescope of
the University of Bochum which is located at the European Southern
Observatory at La Silla, Chile. The used sample containes 53 F - G K M K standard stars, only a few metal-deficient stars (of moderate
deficiency) are available among these. We used an approach based on
criteria evaluation in spectra; ratios of narrow-band measurements
(4-12 nm) in selected intervals are used as classification criteria. The
technique includes the method of stepwise linear regression to calibrate criteria and the method of the steepest decent to classify stars
(to determine their spectral types and absolute magnitudes). The
description of the technique will be published elsewhere.
Unauthenticated
Download Date | 6/16/17 6:30 AM
590
V. Malyuto and Th.
Schmidt-Kaler
To estimate the internal acuracy of classification, we classified
standard stars using the technique based on the same stars. The
data are characterized by the following rms differences (published
minus calculated values):
0.10 for spectral types, 28 F5-K5 V-III stars,
0.12 for spectral types, 18 F5-K5 II—I stars,
0.67 mag for absolute magnitudes, 28 F5-K5 V-III stars,
1.16 mag for absolute magnitudes, 18 F5-K5 II—I stars.
For a comparison, we list here the corresponding estimates in the
case of a photometric classification based on Vilnius photoelectric
photometry (Straizys 1977, 1992):
0.075 for spectral types, F8-G8 V-I stars,
0.73 mag for absolute magnitudes, F8-G8 V-III stars,
1.10 mag for absolute magnitudes, F8-G8 I—II stars.
We see that there is a good coincidence of the estimates obtained
with these different classification techniques. These rms estimates
include the rms errors of published MK values and they are of the
same order as those. It means that the real accuracy of quantitative
classification is much better. Direct estimates of the rms differences
(our spectrophotometric minus Vilnius photometric classification results) for the stars in common could allow to obtain direct estimates
of accuracies of both classification techniques.
REFERENCES
Straizys V. 1977, Multicolor Stellar Photometry, Mokslas Publishers,
Vilnius, Lithuania
Straizys V. 1992, Multicolor Stellar Photometry, Pachart Publishing
House, Tucson, Arizona
Unauthenticated
Download Date | 6/16/17 6:30 AM