Baltic
Astronomy,
vol. 8, 491-498,
1999.
S Y N T H E T I C P H O T O M E T R Y E X P E R I M E N T S IN T H E
V I C I N I T Y O F T H E P A S C H E N J U M P . II.
V. Straizys
Institute of Theoretical Physics
Vilnius 2600,
Lithuania
and Astronomy,
Gostauto
12,
Received D e c e m b e r 15, 1999, revised February 19, 2000.
Abstract. T h e p a p e r continues selection of the o p t i m u m response
f u n c t i o n s of the photometric system for the Gaia orbiting observatory.
It is shown t h a t the p a s s b a n d s at 810, 875 a n d 938 nm, with the
widths of 40, 30 a n d 20 nm respectively, are the best compromise
for t h e two p u r p o s e s : (1) t e m p e r a t u r e a n d gravity determination for
B - A - F s t a r s a n d ( 2 ) photometric identification of M-type s t a r s a n d
carbon-rich s t a r s .
Key words:
m e t h o d s : synthetic photometry - stars: f u n d a m e n t a l
p a r a m e t e r s , classification - orbiting observatories: G a i a
1. INTRODUCTION
Stellar photometry in the near infrared spectrum, including the
Paschenjump, is important both for the classification of B - A - F stars
and for the identification of oxygen-rich and carbon-rich stars of low
temperatures. Early-type stars exhibit lines of the Paschen series
and the Paschen jump in this region. The height of the Paschen
jump, as well as the absorption strengths of the Paschen lines, are
functions of temperature and gravity. Stars of spectral classes K, M,
S, R and N are near maximum intensity in the infrared wavelengths,
and a number of strong molecular bands are present there.
In the first part of this study (Straizys 1998, hereafter Paper
I) it was demonstrated that this spectral region is sufficiently informative to be used in photometric two-dimensional classification of
early-type stars. It was shown that the medium width passbands
at 800, 875 and 940 nm are optimum for measuring the height of
Unauthenticated
Download Date | 6/18/17 12:57 AM
492
V.
Straizys
the Paschen j u m p and the absorption strength of the lines - high
members of the Paschen series ( P l l , P12, P13 and P14). Later on,
the 940 passband was shifted to 938 nm to avoid strong emission
in [S III] line observed in emission nebulae. These three passbands,
combined with some passbands in the optical spectral region, make
it possible to classify B - A - F stars in temperatures and gravities
using interstellar reddening-free Q, Q diagrams. Consequently, the
Paschen diagrams can be used in the case when stars are too faint to
be measured in the ultraviolet and violet passbands near the Balmer
jump. Also it was shown that due to TiO and CN molecular bands
the same passbands are useful for identification of M-type and carbon stars. These three infrared passbands were proposed as a part of
the photometric system of the Gaia orbiting observatory planned by
the European Space Agency (see Merat et al. 1999 and other papers
in the Proceedings of the Gaia Workshop at Leiden).
Later on, Grenon, Jordi, Figueras and Torra (1999, hereafter
G J F T ) proposed for Gaia a somewhat different set of passbands in
the same spectral region with the prime intention to recognize and
classify photometrically M-type and carbon stars. The set consists of
F75, F78, F82 and F89 passbands at central wavelengths of 747, 778,
816 and 894 n m and half-widths of 30-50 nm. Two of these passbands at 816 and 894 n m were supposed to be useful for measuring
the Paschen j u m p and the Paschen line absorption.
In the present paper an attempt is made to compare classification
possibilities of the two sets of passbands proposed in these papers.
The intention is to optimize infrared passbands for the two tasks at
the same time: for measuring the Paschen jump and Paschen line
absorption and for separating the oxygen- and carbon-rich stars of
low temperatures.
2. SYNTHETIC P H O T O M E T R Y
At the beginning we will estimate the quality of the G J F T passbands for two-dimensional classification of B - A - F stars. For this the
synthetic color indices and Q-parameters will be calculated for the
set of Kurucz (1995) models of stellar atmospheres of solar metallicity.
Synthetic color indices were calculated by the following equation:
f
F(X)Tx(\)R1(\)d\
J
F(\)T*(\)R2(\)d\
m,\ — rri2 = —2.5 log
+ const,
Unauthenticated
Download Date | 6/18/17 12:57 AM
(1)
Photometry
in the vicinity
of the Paschen jump.
II
493
where F( A) are the synthetic flux distribution functions of the models, t i ( A ) is the t r a n s m i t t a n c e function of x units of interstellar dust
(from Straizys 1992), i?i(A) and i?2(A) are the response functions of
the passbands 1 and 2. T h e value of the constant makes all color
indices of the model with T e = 35000 K and log g = 4.0 to be zeros.
T h e integration step was always 2 nm.
T h e following response functions were used: (1) the F82 and
F89 response curves from G J F T , cut to zero from both sides at the
2 % response, (2) the response functions of the 800, 875 and 9S8
passbands defined in P a p e r I, (3) the response functions of the passb a n d s at 405, 460 and 545 n m proposed for Gaia by H0g, Knude and
Straizys (1999). T h e only difference between the response functions
used in P a p e r I and the response functions used here is in the form
of rising a n d falling edges: in the present paper the functions were
taken of the form of trapezium to make them more similar to the
t r a n s m i t t a n c e functions of dichroic interference filters.
Color indices are used to calculate the reddening-free Q-parameters:
Q1234 = ( m i - m2) - ( E i 2 / E 3 4 ) ( r n 3 - m 4 ) ,
(2)
here
Ek,t = (mk ~ "^) r eddened ~
~ mf)i ntr insic-
(3)
Trying to u n d e r s t a n d reasons of the positions of sequences in
Q, Q diagrams,e have m a d e shifts of the passbands in different directions a n d changed their widths.
3. RESULTS
We have calculated the same types of color indices, as in P a p e r
I. They include the blanketing-free color temperature index 460-545,
the Paschen j u m p color indices 800-938, 810-938, F82-938, etc. a n d
the indices measuring the absorption of the high member lines of the
Paschen series 800-875, 810-F89, etc. We used the same types of
Q-parameters, as in P a p e r I.
Figs, l a a n d l b show a comparison of the Q (800, 875, 460,
545) vs. Q(4 05, 4 60, 545) diagram from paper I with the Q (F82,
F89, 4 60, 545) vs. Q(405, 460, 545) diagram with the two infrared
passbands f r o m G J F T . Fig. l c gives the diagram identical to Fig. l a
except of the shift of the 800 passband to 810. T h e diagram in panel
Unauthenticated
Download Date | 6/18/17 12:57 AM
494
V. Straizys
Fig. 1. The Q,Q diagrams for the Kurucz model atmospheres of
solar metallicity for log g = 4.0, 2.5 and 1.0. (a) Q(800, 875, 460, 545)
vs. Q(405, 460, 545); (b) Q(F82, F89, 460, 545) vs. Q(405, 460, 545)]
(c) Q(810, 875, 460, 545) vs. Q(405, 460, 545).
Unauthenticated
Download Date | 6/18/17 12:57 AM
Photometry in the vicinity of the Paschen jump. II
495
0.4 —i—i—i—i—I—i—i—I—i—|—I—i—i—I—I—i—I—i—r
S 0.2 -
Q£
0.0
1 1 1 1
i
1 1 1
i I i i i i i i i i 1
0.5 n
1.0
1*405, 460, 545
0.4 —i—i—i—i—i—i—i—i—i—|—i—i—i—i—i—i—i—i—r
3 0.2 -
—0 2 —'—1—LjJ—1—1—1—1 ' 1
0.0
0.5
1 1 1
'
1
1.0
*1405, 460, 545
n
Fig. 2. The Q,Q diagrams for the Kurucz model atmospheres of solar
metallicity for log g = 4.0, 2.5 and 1.0. (a) Q(800, 875, 938) vs. Q(405,
460, 545); (b) Q(F82, F89, 938) vs. Q(405, 46O, 545)\ (c) Q(810, 875,
938) vs. Q(405, 460, 545).
Unauthenticated
Download Date | 6/18/17 12:57 AM
496
V. Straizys
(b) gives the maximum separation of the sequences for log g = 1.0
and 4.0 which is by 0.05 mag smaller than in panel (a). At the same
time, the difference between panels (a) and (c) is insignificant.
Figs. 2 (a,b,c) give a comparison of other set of Q, Q diagrams.
The X-axis everywhere is the same as in Fig. 1. On the F-axis
the following Q-parameters are plotted: (a) Q{800, 875, 938), (b)
Q(F82, F89, 938) and (c) Q(810, 875, 938). Again, the maximum
separation of log <7 = 1.0 and 4.0 sequences in (b) is smaller t h a n in
(a) by 0.06 mag. At the same time, (a) and (c) it is twice smaller.
This means t h a t the 800 passband can be replaced by 810 passband with almost no h a r m for the classification accuracy.
On the other hand, the shift of the whole response function F89
to 880 n m and 870 n m gives almost no improvement in the sequence
separation. This means that the F89 curve is too broad (half-width
48 nm) to be a measure of absorption of the Paschen lines near the
series limit.
4. DISCUSSION AND CONCLUSIONS
T h e F82 passband in the G J F T system is intended for separation
of M-type and carbon stars, since in it the TiO absorption is faint
and the CN absorption is strong. The passband 800, proposed in
Paper I for the measurement of the Paschen continuum shortwards
of the j u m p , is slightly too short to be the most sensitive criterium
of the presence of TiO or CN absorption bands. However, we are
safe to shift the 800 passband from 800 nm to 810 nm. In this case
there is a compromise: the passband 810 is a good measure of the
Paschen continuum (Fig. 3) and is close to the position with no TiO
bands and strong CN band (Fig. 4).
T h e F89 passband with the width 48 nm in the G J F T system is
used as a measure of pseudocontinuum both in oxygen- and carbonrich cool stars. T h e same passband can be used for measurement of
the blocking effect of the high members of the Paschen series which
is nice criterium of gravity. However, as it was shown in Section 3,
the best sensitivity of Q-parameters is observed when this passband
is at 875 n m and has a width of 30 nm. Fig. 4 shows that the F89
passband may be replaced by the 875 passband which remains quite
close to the peak of pseudocontinuum both in M-type and carbon
stars.
Unauthenticated
Download Date | 6/18/17 12:57 AM
Photometry
RW
1
in the vicinity of the Paschen jump.
1
1.0
1
p1
1
1
1
497
II
1
p3
P2
0.8
0.6
-
0.4
P13
V
P12
0.2
0.0
1
1
800
850
\
V
P11
\P8
P10
\V
\ I
WJ
P9|
900
X(nm)
950
F i g . 3 . Response functions of the 810, 875 and 035passbands plotted
together with the energy distribution curve of the model of Vega. Energy
fluxes are in a relative scale.
R«
1.0
0.8
0.6
0.4
0.2
0.0
700
750
800
850
900
X. (nm)
950
F i g . 4 . Response functions of the 810, 875, 938, F82 and F89 passbands plotted together with the energy distribution curves of a M5 III star
and a carbon star. Part of the Figure is taken from Grenon et al. (1999).
Unauthenticated
Download Date | 6/18/17 12:57 AM
498
V.
Straizys
The general conclusion: the three passbands at 810, 875 and 938
n m with widths of 40, 30 and 20 nm respectively, may be successfully used for the two purposes: (1) photometric determination of
temperatures and gravities for B - A - F type stars in the presence of
interstellar reddening and (2) photometric identification of oxygenrich and carbon-rich stars. The passbands will be most important in
case of heavily reddened stars which are too faint to be measurable
in the ultraviolet.
A C K N O W L E D G M E N T . The author is grateful to Erik H0g for
the numerical d a t a of the response functions of the G J F T system.
REFERENCES
Grenon M., Jordi C., Figueras F., Torra J. 1999, An Intermediate Band
System for Gaia, Manuscript submitted at the Leiden meeting of the
Photometric Working Group for Gaia, April 20-21, 1999
H0g E., Knude J., Straizys V. 1999, Specification of Gaia Photometric
Systems, Document SAG-CUO-58 of February 11, 1999
Straizys V. 1992, Multicolor Stellar Photometry,
Pachart Publishing
House, Tucson, Arizona
Straizys V. 1998, Baltic Astronomy, 7, 571 (Paper I)
Unauthenticated
Download Date | 6/18/17 12:57 AM