Smooth time variation spectra as a tool to study line profile variability in spectra of hot stars
Line profiles in spectra of almost all OB stars are variable. In many cases the line profile variations are extremely weak (not more than 1% of the continuum level). For detection of such micro line profile variations of various nature we used a smooth Time Variation Spectra method (smTVS). This met...
Збережено в:
| Дата: | 2012 |
|---|---|
| Автор: | |
| Формат: | Стаття |
| Мова: | English |
| Опубліковано: |
Головна астрономічна обсерваторія НАН України
2012
|
| Назва видання: | Advances in Astronomy and Space Physics |
| Онлайн доступ: | http://dspace.nbuv.gov.ua/handle/123456789/119105 |
| Теги: |
Додати тег
Немає тегів, Будьте першим, хто поставить тег для цього запису!
|
| Назва журналу: | Digital Library of Periodicals of National Academy of Sciences of Ukraine |
| Цитувати: | Smooth time variation spectra as a tool to study line profile variability in spectra of hot stars / N.P. Sudnik // Advances in Astronomy and Space Physics. — 2012. — Т. 2., вип. 1. — С. 5-8. — Бібліогр.: 10 назв. — англ. |
Репозитарії
Digital Library of Periodicals of National Academy of Sciences of Ukraine| id |
irk-123456789-119105 |
|---|---|
| record_format |
dspace |
| spelling |
irk-123456789-1191052017-06-05T03:03:59Z Smooth time variation spectra as a tool to study line profile variability in spectra of hot stars Sudnik, N.P. Line profiles in spectra of almost all OB stars are variable. In many cases the line profile variations are extremely weak (not more than 1% of the continuum level). For detection of such micro line profile variations of various nature we used a smooth Time Variation Spectra method (smTVS). This method appeared to be very sensitive and can be used to detect ultra weak line profile variations which can not be detected by other methods. We applied the smTVS method to detect the micro LPV in spectra of bright O stars λ Cep, λ Ori A and ζ Ori A. Spectra of these stars were obtained by using the 6-m and 1-m telescopes of the Northern Caucasus Special Astrophysical Observatory (Russia) and the 1.8-m telescope of Bohyunsan Astronomical Observatory (Korea). 2012 Article Smooth time variation spectra as a tool to study line profile variability in spectra of hot stars / N.P. Sudnik // Advances in Astronomy and Space Physics. — 2012. — Т. 2., вип. 1. — С. 5-8. — Бібліогр.: 10 назв. — англ. 2227-1481 http://dspace.nbuv.gov.ua/handle/123456789/119105 en Advances in Astronomy and Space Physics Головна астрономічна обсерваторія НАН України |
| institution |
Digital Library of Periodicals of National Academy of Sciences of Ukraine |
| collection |
DSpace DC |
| language |
English |
| description |
Line profiles in spectra of almost all OB stars are variable. In many cases the line profile variations are extremely weak (not more than 1% of the continuum level). For detection of such micro line profile variations of various nature we used a smooth Time Variation Spectra method (smTVS). This method appeared to be very sensitive and can be used to detect ultra weak line profile variations which can not be detected by other methods. We applied the smTVS method to detect the micro LPV in spectra of bright O stars λ Cep, λ Ori A and ζ Ori A. Spectra of these stars were obtained by using the 6-m and 1-m telescopes of the Northern Caucasus Special Astrophysical Observatory (Russia) and the 1.8-m telescope of Bohyunsan Astronomical Observatory (Korea). |
| format |
Article |
| author |
Sudnik, N.P. |
| spellingShingle |
Sudnik, N.P. Smooth time variation spectra as a tool to study line profile variability in spectra of hot stars Advances in Astronomy and Space Physics |
| author_facet |
Sudnik, N.P. |
| author_sort |
Sudnik, N.P. |
| title |
Smooth time variation spectra as a tool to study line profile variability in spectra of hot stars |
| title_short |
Smooth time variation spectra as a tool to study line profile variability in spectra of hot stars |
| title_full |
Smooth time variation spectra as a tool to study line profile variability in spectra of hot stars |
| title_fullStr |
Smooth time variation spectra as a tool to study line profile variability in spectra of hot stars |
| title_full_unstemmed |
Smooth time variation spectra as a tool to study line profile variability in spectra of hot stars |
| title_sort |
smooth time variation spectra as a tool to study line profile variability in spectra of hot stars |
| publisher |
Головна астрономічна обсерваторія НАН України |
| publishDate |
2012 |
| url |
http://dspace.nbuv.gov.ua/handle/123456789/119105 |
| citation_txt |
Smooth time variation spectra as a tool to study line profile variability in spectra of hot stars / N.P. Sudnik // Advances in Astronomy and Space Physics. — 2012. — Т. 2., вип. 1. — С. 5-8. — Бібліогр.: 10 назв. — англ. |
| series |
Advances in Astronomy and Space Physics |
| work_keys_str_mv |
AT sudniknp smoothtimevariationspectraasatooltostudylineprofilevariabilityinspectraofhotstars |
| first_indexed |
2025-07-08T15:13:59Z |
| last_indexed |
2025-07-08T15:13:59Z |
| _version_ |
1837092192159531008 |
| fulltext |
Smooth time variation spectra as a tool to study line pro�le
variability in spectra of hot stars
N.P. Sudnik∗
Advances in Astronomy and Space Physics, 2, 5-8 (2012)
© N.P. Sudnik, 2012
Saint-Petersburg State University, 28 Universitetskiy ave., 198504, Saint-Petersburg, Russia
Line pro�les in spectra of almost all OB stars are variable. In many cases the line pro�le variations are extremely
weak (not more than 1% of the continuum level). For detection of such micro line pro�le variations of various nature
we used a smooth Time Variation Spectra method (smTVS). This method appeared to be very sensitive and can be
used to detect ultra weak line pro�le variations which can not be detected by other methods. We applied the smTVS
method to detect the micro LPV in spectra of bright O stars λCep, λOri A and ζ Ori A. Spectra of these stars
were obtained by using the 6-m and 1-m telescopes of the Northern Caucasus Special Astrophysical Observatory
(Russia) and the 1.8-m telescope of Bohyunsan Astronomical Observatory (Korea).
Key words: stars: early-type, stars: massive, stars: individual: λ Cep, λ OriA, ζ OriA; line: pro�le variations;
methods: data analysis, methods: statistical
introduction
An investigation of the line pro�le variations
(LPV) in spectra of hot stars started in 70th of
the XX century. To detect variability researchers
usually overplotted individual or residual (individ-
ual pro�le minus the mean one) pro�les of line or
computed dynamical spectrum. It was easy to rec-
ognize the LPV by using these procedures if spec-
tral deviations are large compared with noise and
data are well sampled in time. With improve-
ment of observational and computational facilities
for LPF study di�erent methods, intended to ob-
tain periods, scales and other characteristics of LPV
and their changes in time, have come to use. At
present the most widely used methods are Temporal
Variance Spectrum (TVS) analysis, Fourier analysis,
wavelet analysis, Doppler imaging technique. Here
we describe a smoothing Time Variation Spectrum
(smTVS) method proposed by Kholtygin et al. [5]
and its application for detection of micro LPV in
spectra of O stars.
observations
We analysed the observations that were made in
1997�2009 at the Special Astrophysical Observatory
(SAO) of the Russian Academy of Science and at Bo-
hyunsan Optical Astronomy Observatory (BOAO),
Korea. The stars were observed in SAO with the 6-
m telescope and LYNX spectrograph1 with spectral
resolution 60,000 and CCD 512×512 pixels and with
the 1-m telescope and CEGS spectrograph [9] with
spectral resolution 45,000 and CCD 1242×1152 pix-
els. The reduction of SAO spectra was made with the
help of the MIDAS software package. The stars were
also observed at BOAO with the 1.8-m telescope and
the BOES spectrograph [6] with spectral resolution
45,000 and large CCD (2048×4096 pixels). All spec-
tra in BOAO were obtained within the 3830−8260 Å
region. Preliminary reduction of CCD frames was
made with the help of IRAF software package. A log
of observations is given in Table 1. All spectra were
normalized to the continuum level by the same way.
The procedure of �nding the continuum level is de-
scribed in [5].
smoothing time variation
spectrum (smTVS) analysis
For detection of micro LPV of various nature we
used the smooth Time Variation Spectra (smTVS)
technique. That is substantially modi�ed version of
Temporal Variance Spectrum (TVS) analysis intro-
duced by Fullerton et al. [2]. In this method the
standard deviations of variations in both lines and
continuum are compared. If the amplitude of de-
viations within a line is larger than in the contin-
uum then detection at a selected signi�cance level
∗snata.astro@gmail.com
1http://www.sao.ru/hq/ssl/LYNX.html
5
Advances in Astronomy and Space Physics N. P. Sudnik
can be claimed. If the observed standard deviations
are smaller than expected for the noise contribution
into LPV at the selected signi�cance level, then only
an upper limit of the amplitude of the LPV can be
estimated.
In the case of line pro�le variability with a su�-
ciently high amplitude the value of TV S(λ) in the
vicinity of the corresponding line substantially ex-
ceeds its value in the adjacent continuum. But in
many cases the LPV are extremely weak (not more
than 1% of the continuum level for many of O stars).
In the absence of any visible variations we can deter-
mine whether a line pro�le was indeed variable using
a following procedure.
Before the standard deviation spectrum was ob-
tained, di�erential spectra were smoothed using a
wide Gauss �lter (S). After smoothing the ampli-
tude of a noise component of the di�erential pro�les
decreases by a factor of ≈
√
S/δλ , where δλ is a
pixel width. If the width of the variable component
is not smaller than that of the �lter, then smooth-
ing will not signi�cantly change the amplitude of the
variable component, and peak in the standard devia-
tion spectrum that corresponds to the variable com-
ponent can be detected.
This procedure, called smoothing Time Varia-
tion Spectra (smTVS), was introduced by Kholty-
gin et al. [5]. The smTVS is described by the follow-
ing equation:
smTV S(λ, S) =
n∑
i=1
gi
[
Fi(λ, S)− F (λ, S)
]2
(N − 1)
n∑
i=1
gi
, (1)
where N is a number of spectra, Fi(λ, S) is a �ux
in the spectrum with number i at wavelength λ
smoothed with Gauss �lter (S is a �lter width) and
normalized to the continuum level, F (λ, S) is a �ux
at wavelength λ averaged over all smoothed �uxes,
gi is a relative weight of the ith observation.
To ensure that LPV are real we de�ned a small
signi�cance level α ¿ 1 for the hypothesis that the
line pro�le variations are due to a random variations
of the noise component of the pro�les only. The
smTVS obeys a χ2/d.o.f. statistics with N − 1 de-
grees of freedom. Let χ2
α be speci�ed so that the
probability P
(
χ2/d.o.f. > χ2
α
)
= α. If the calcu-
lated smTV S(λ) value exceeds χ2
α , the hypothesis
that the line pro�le is variable can be accepted.
It was shown empirically that the best results cor-
respond to smoothing with a Gaussian �lter with
width S = 0.7 − 1.3Å (it means usually 15�30 pix-
els). For S = 0 the value of smTVS(λ,0) corresponds
to the TVS value introduced by Fullerton et al. [2].
Note that the e�ciency of the method is sensitive
to the number of obtained spectra, and it enhances
substantially as this number increases.
Here we present an application of smTVS anal-
ysis to the spectra of bright massive O stars λCep,
λOri A and ζ Ori A. Parameters of these stars are
presented in Table 2.
The Figure 1 presents a density plot of the
smTVS for the Heii λ 4200Å line pro�les. Darker ar-
eas correspond to higher amplitudes of the smTVS.
The density plot shows that the variability of the
Heii λ 4200Å line is clearly present at all �lter
widths. The smTVS with the �lter width more
then 1Å also indicates variability of the weak
Ciii λ 4187Å and Siii λ 4212Å lines pro�les, which
cannot be detected using ordinary methods.
Fig. 1: Density diagram for smTVS (top) and mean
line pro�le of Heii λ 4200Å line (bottom) in spectra of
λCep. Darker areas correspond to higher amplitudes of
the smTVS. The maximum smTVS value is taken as 1.
In Figure 2 TVS (top panel) and smTVS (bot-
tom panel) functions are shown for comparison. The
variability of Ciii and Siiv line undetected by simple
TVS analysis is clearly seen.
Fig. 2: TVS (top) and smTVS (bottom) of the
Heii λ 4200Å line in spectra of λCep. Filter width is
1.5 Å. The horizontal line corresponds to the signi�cance
level 0.001.
6
Advances in Astronomy and Space Physics N. P. Sudnik
The smTVS of the line pro�les in spectra of ζ Ori
A presented in Figure 3 is obtained for observations
made in BOAO. The total duration of observation is
2 hours. In spite of this the variability of the line Hγ
and weak lines in its wings can be seen.
In spectra of λOri A a weak variability (no more
than 1% of the continuum level) is also detected. In
Figure 4 one can see that the variability of lines Hγ ,
and lines of Si, O, N ions occurs at all �lter width
although it can not be found by simple TVS analysis.
Fig. 3: The same as Fig. 1 but for the Hγ line in spectra
of ζ Ori A.
Fig. 4: The same as Fig. 1 but for the Hγ line in spectra
of λOri A.
Note that, in spite of the fact that a lot of lines
are extremely weak and their residual intensities do
not exceed the noise level of the adjacent continuum
their variability is clearly detected.
Unfortunately, our technique cannot be used for
accurate localization of weak lines with variable pro-
�les for large �lter widths.
The procedure of searching the LPV described
above can only answer a question whether the pro�le
of a speci�c line is variable or not. At the same time,
to determine a mechanism of LPV we need to know
whether the pro�le variations are regular (cyclical),
irregular (stochastic) or both. To de�ne this and also
to reveal characteristic features of variability and its
changes in time we have to use the Fourier or wavelet
analysis.
results and conclusions
We can see that the smTVS analysis is a very
sensitive method to detect micro LPV that can not
be found in a usual way. It can be used when ampli-
tude of variation is small (less then 1% in continuum
units) and do not exceed the noise level, a number
of spectra is small, time grid is uneven.
Weak variability of the lines of the Si, C, O, N
ions in spectra of O stars λCep, λOri A and ζ Ori A
was revealed by smTVS only.
acknowledgement
The author acknowledges Saint-Petersburg State
University for a research grant 6.38.73.2011.
references
[1] Bouret J.-C., Donati J.-F. & Martins F. et al. 2008,
MNRAS, 389, 75
[2] Fullerton A.W., Gies D.R. & Bolton C.T. 1996, ApJS,
103, 475
[3] Fullerton A.W., Massa D. L. & Prinja R.K. 2006, ApJ,
637, 1025
[4] Howarth I.D., Siebert K.W., Hussain G.A. J. &
Prinja R.K. 1997, MNRAS, 284, 265
[5] Kholtygin A. F., Burlakova T.E., Fabrika S.N.,
Valyavin G.G. & Yushkin M.V. 2006, Astron. Rep., 50,
887
[6] Kim K.-M., Jang B.-H., Han I. et al. 2002, J. of the
Korean Astron. Soc., 35, 221
[7] Lamers H. J.G. L.M. & Leitherer C. 1993, ApJ, 412, 771
[8] Maíz-Apellániz J., Walborn N.R., Galué H.Á. &
Wei L.H. 2004, ApJS, 151, 103
[9] Musaev F.A. 1996, Astron. Lett., 22, 715
[10] Repolust T., Puls J. & Herrero A. 2004, A&A, 415, 349
7
Advances in Astronomy and Space Physics N. P. Sudnik
Table 1: Log of observations. ∆T is duration of observations, texp is exposition time.
object date number of spectra texp, min ∆T, h telescope spectrograph, CCD
λCep 14.08�16.08 1997 70 2�5 7.8 SAO, BTA Lynx, 512 x 512
HD 210839 20.11�16.12 2007 30 3�5 7.2 BOAO, 1.8m BOES, 2048 x 4096
λOri A 29.11�04.11 2001 75 10 12.5 SAO, 1m CEGS, 1242 x 1152
HD 37742 20.11�22.11 2007 18 1�3 2.4 BOAO, 1.8m BOES, 2048 x 4096
ζ Ori A 29.11�04.11 2001 36 2 1.2 SAO, 1m CEGS, 1242 x 1152
HD 36861 17.12 2007 34 7 2.9 BOAO, 1.8m BOES, 2048 x 4096
Table 2: Parameters of program stars
λCep ( O6 I (n)fp ) λOri A ( O8 III ((f)) ) ζ Ori A ( O9.7 Ib )
parameter value reference value reference value reference
m
V
5.04 [8] 3.41 [8] 3.78 [8]
lg L/L¯ 5.83 [10] 5.38 [7] 5.64 [1]
Teff , K 36 000 [3] 33 600 [3] 29 500 [1]
M/M¯ 62 [10] 31 [7] 40 [1]
R/R¯ 21 [3] 15 [3] 25 [1]
lg g 3.58 [10] 3.61 [7] 3.25 [1]
Ṁ/M¯ · 10−6 2.7 [3] ≤ 1.4 [3] 1.4�1.9 [1]
V sin i, km/s 200 [10] 74 [4] 110 [1]
V∞, km/s 2250 [3] 2400 [3] 2100 [1]
8
|