Mem. S.A.It. Suppl. Vol.

8, 130
Memorie della
c SAIt 2005

Supplementi

T eff and log g determinations

Barry Smalley

Astrophysics Group, Keele University, Staffordshire, ST5 5BG, United Kingdom

e-mail: bs@astro.keele.ac.uk

Abstract. A discussion on the determination of effective temperature (T eff ) and surface

gravity (log g) is presented. The observational requirements for model-independent funda-
mental parameters are summarized, including an assessment of the accuracy of these values
for the Sun and Vega. The use of various model-dependent techniques for determining T eff
and log g are outlined, including photometry, flux fitting, and spectral line ratios. A combi-
nation of several of these techniques allows for the assessment of the quality of our param-
eter determinations. While some techniques can give precise parameter determinations, the
overall accuracy of the values is significantly less and sometimes difficult to quantify.

Key words. Stars: atmospheres, Stars: fundamental parameters, Techniques: photometric,

Techniques: spectroscopic, Line: profiles

1. Introduction in the cases of a few bright stars and certain bi-

nary systems. We have to rely on model atmo-
The stellar atmospheric parameters of effective sphere analyses of spectra in order to deduce
temperature (T eff ) and surface gravity (log g) the atmospheric parameters.
are of fundamental astrophysical importance. We need to be confident in the atmospheric
They are the prerequisites to any detailed abun- parameters before we start any detailed analy-
dance analysis. As well as defining the physical ses. This is especially important when compar-
conditions in the stellar atmosphere, the atmo- ing stars with peculiar abundances to normal
spheric parameters are directly related to the ones.
physical properties of the star; mass (M), ra-
dius (R) and luminosity (L).
1.1. Effective Temperature
Model atmospheres are our analytical link
between the physical properties of the star (M, The effective temperature of a star is physically
R and L) and the observed flux distribution and related to the total radiant power per unit area
spectral line profiles. These observations can at stellar surface (F∗ ):
be used to obtain values for the atmospheric Z ∞
4 L
parameters, assuming of course that the models σT eff ≡ Fν dν = F∗ = 2
0 4πR
used are adequate and appropriate. The values
of T eff and log g obtained must necessarily be It is the temperature of an equivalent black
consistent with the actual values of M, R and L. body that gives the same total power per unit
Unfortunately, the physical properties of stars area, and is directly given by stellar luminosity
are not generally directly ascertainable, except and radius.
 Smalley: T eff and log g determinations 131

Since there is no true ‘surface’ to a star, 2.1. Sources of Stellar Fluxes

the stellar radius can vary with the wavelength
of observation and nature of the star. Radius Ultraviolet fluxes have been obtained by
is taken as the depth of formation of the con- various space-based observatories: TD1
tinuum, which in the visible region is approxi- (Thompson et al. 1978; Jamar et al. 1976;
mately constant for most stars (Gray 1992). Macau-Hercot et al. 1978), OAO-2 (Code et
Providing there is no interstellar reddening al. 1980), and the IUE final archive. HST is
(or due allowance for it is made), the total ob- also another potential source of flux-calibrated
served flux at the earth ( f⊕ ) can be used to de- ultraviolet spectra.
termine the total flux at the star: Optical spectrophotometry can be obtained
from various sources, such as Breger (1976);
θ2 Adelman et al. (1989); Burnashev (1985);
F∗ = f⊕
4 Glushneva et al. (1998). The ASTRA spec-
The only additional requirement is a de- trophotometer should soon provide a large
termination of the stellar angular diameter amount of high-precision stellar flux measure-
(θ). This can be obtained directly using tech- ments (Adelman et al. 2005). In the absence
niques such as speckle photometry, interfer- of suitable spectrophotometry, optical fluxes
ometry, and lunar occultations, and indirectly can be estimated from photometry (Smalley &
from eclipsing binary systems with known dis- Dworetsky 1995; Smalley et al. 2002).
tances. We must, however, be aware that some Infrared flux points can be obtained from
of these methods require the (not always ex- the 2MASS, DENIS and IRAS surveys, as well
plicit) use of limb-darkening corrections. as the compilation by Gezari et al. (1999).

1.2. Surface Gravity 2.2. Sources of Angular Diameters

The surface gravity of a star is directly given Useful catalogues of angular diameter mea-
by the stellar mass and radius (in solar units): surements are CADARS (Pasinetti-Fracassini
M et al. 2001) and CHARM2 (Richichi et al.
g = g 2005). But, beware, not all are direct measure-
R2
ments!
or, logarithmically, Incidentally, Kervella et al. (2004a) have
log g = log M − 2 log R + 4.437 produced a good surface brightness relation-
ship for main-sequence and sub-giants. While
Surface gravity is a measure of the pho-
it is obviously not for use of fundamental pa-
tospheric pressure of the stellar atmosphere.
rameters, it can be used in the determination of
Direct measurements are possible from eclips-
stellar distances (See for example Southworth
ing spectroscopic binaries, but again be aware
et al. 2005).
of hidden model atmosphere dependences.

2. Fundamental Stars 2.3. Source of Masses and Radii

A fundamental star has at least one of its at- Detached eclipsing binary systems are our
mospheric parameters obtained without refer- source of stellar masses and radii. These are
ence to model atmospheres. An ideal funda- often accurate to 1∼2%, and give us our direct
mental star will have both parameters mea- log g determinations. Useful sources include
sured. These stars are vital for the quality as- Popper (1980), Andersen (1991), Perevozkina
surance of model predictions. Unfortunately, & Svechnikov (1999), Lastennet & Valls-
the number of fundamental stars is relatively Gabaud (2002).
limited by the lack of suitable measurements. For use in T eff determinations, radii need to
There now follows a non-exhaustive summary be converted into angular diameters, which re-
of the main sources of observational data. quires an accurate distance determination. For
132 Smalley: T eff and log g determinations

example, the HIPPARCOS parallax catalogue ibration with uses Vega as a zero-point must
(ESA 1997), or the membership of a cluster assume a value for log g. However, detailed
with a known distance, provided that distance model atmosphere analyses give a value of
has not been obtained using model-dependent log g= 3.95 ± 0.05 (Castelli & Kurucz 1994).
methods. An interesting discussion on the accuracy
Single star mass determination is exceed- of the visible and near-infrared absolute flux
ingly difficult, with microlensing the only calibrations is given by Mégessier (1995).
known direct method (Alcock et al. 2001; These uncertainties place a limit on our cur-
Jiang et al. 2004). This relies on chance align- rent direct determinations of stellar fundamen-
ments and is considerably less accurate than tal parameters.
that possible with eclipsing binary systems.
3. Indirect Methods
2.4. Accuracy of Direct Measurements The direct determination of T eff and log g is not
2.4.1. Sun possible for most stars. Hence, we have to use
indirect methods. In this section we discuss the
Our nearly stellar companion, the Sun, has the use of various techniques used to determine the
most accurately known stellar parameters. The atmospheric parameters.
measured total solar flux at the earth, the Solar When determining T eff and log g, using
Constant, is f = 1367 ± 4 W m−2 (Mendoza model-dependent techniques, we must not ne-
2005). Variations due to the Solar Cycle and glect metallicity ([M/H]). An incorrect metal-
rotation, contribute 0.1% and 0.2%, respec- licity can have a significant effect on perceived
tively (Zahid et al. 2004). This equates to ± values of these parameters.
4 K in the Solar effective temperature. A value
of T eff = 5777 ± 10 K is obtained from the
Solar Constant and the measured Solar radius, 3.1. Photometric Grid Calibrations
including calibration uncertainties. The Solar There have been many photometric systems
surface gravity is exceedingly well known; developed to describe the shape of stellar
log g = 4.4374 ± 0.0005 (Gray 1992). flux distributions via magnitude (colour) dif-
ferences. Since they use wide band passes ob-
2.4.2. Vega servations can be obtained in a fraction of the
time required by spectrophotometry and can be
The bright star Vega is our primary stellar flux extended to much fainter magnitudes. The use
calibrator (Hayes & Latham 1975; Bohlin & of standardized filter sets allows for the quan-
Gilliland 2004). The measured total flux at titative analysis of stars over a wide magnitude
the earth is f⊕ = 29.83 ± 1.20 × 10−9 W m−2 range.
(Alonso et al. 1994), which is an uncertainty Theoretical photometric indices from
of some 4%. There have been reports that Vega ATLAS flux calculations are normalized using
may be variable (Fernie 1981; Vasil’yev et al. the observed colours and known atmospheric
1989), but these have not been substantiated, parameters of Vega. Vega was originally
and may well be spurious. Nevertheless, this is chosen because it is the primary flux standard
something that ought to be investigated. Using with the highest accuracy spectrophotometry.
the interferometric angular diameter of Ciardi An alternative, semi-empirical approach, is
et al. (2001), θ = 3.223 ± 0.008 mas, we ob- to adjust the theoretical photometry to min-
tain T eff = 9640 ± 100 K. Most of the uncer- imize discrepancy with observations of stars
tainty (∼95K) is due to the uncertainties in the with known parameters. Moon & Dworetsky
measured fluxes, while the error in the angular (1985) used stars with fundamental values
diameter only contributes ∼10K. to shift the grids in order to reduce the dis-
Since Vega is a single star, there is no direct crepancy between the observed and predicted
fundamental log g measurement. Thus any cal- colours. In contrast, Lester, Gray & Kurucz
 Smalley: T eff and log g determinations 133

(1986) treated the raw model colours in the calculations (Blackwell & Lynas-Gray 1994;
same manner as raw stellar photometry. The Mégessier 1994):
model colours were placed on the standard
system using the usual relations of photomet- f⊕ F∗ σT eff 4
ric transformation. However, both these ap- ≡ =
fλ0 F λ0 φ(T eff , log g, λ0 )
proached have the potential to mask physical
problems with models. The method requires a complete flux dis-
Overall, photometry can give very good tribution in order to obtain the total integrated
first estimates of atmospheric parameters. In ( f⊕ ) stellar flux. In practice, however, all of
the absence of any other suitable observations, the flux is not observable, especially in the far-
the values obtainable from photometry are of ultraviolet. But, this is only a serious problem
sufficient accuracy for most purposes, with in the hottest stars, where model atmospheres
typical uncertainties of ±200 K and ±0.2 dex can be used to insert the missing flux, in or-
in T eff and log g, respectively. der to obtain the total integrated flux. Accurate
infrared fluxes are, of course, essential for this
method to produce reliable results.
3.2. T eff –colour Relationships The method is sensitive to the presence of
Effective temperatures can be estimated from any cooler companion stars. The effect of the
photometric colour indices. Empirical calibra- companion is to lower the T eff derived for the
tions are based on stars with known tem- primary. A modified method was proposed and
peratures, often obtained using the IRFM. discussed by Smalley (1993). This method re-
There are many examples in the literature, lies on the relative radii of the two components
for example, (Alonso et al. 1996; Houdashelt in the binary system. The effect of allowing for
et al. 2000; Sekiguchi & Fukugita 2000; the companion can be dramatic; the T eff de-
VandenBerg & Clem 2003; Clem et al. 2004; termined for the primary can be increased by
200 K or more.
Ramírez & Meléndez 2005b).
A very useful by-product of the IRFM is
Particularly useful are V − K calibrations, that it also gives the angular diameter (θ) of the
since this index is much less sensitive to metal- star.
licity than B − V (Alonso et al. 1996; Kinman Given good spectrophotometry, the IRFM
& Castelli 2002; Ramírez & Meléndez should give estimates of T eff , which are clos-
2005b). However, this index is more sensitive est to the ‘true’ fundamental value. In fact it
to the presence of a cool companion. has been used as the basis of other calibrations
Often, there are several steps involved in (e.g. Ramírez & Meléndez 2005a). Typically
obtaining the calibrations. The uncertainties we can obtain temperatures to an accuracy of
and final error on the parameters obtained are 1∼2% (Blackwell et al. 1990). The IRFM re-
not always immediately obvious. sults for Vega have an uncertainty of ∼150K.
Uncertainties in absolute calibration of IR
photometry are important. For example, for
3.3. InfraRed Flux Method 2MASS an error of ∼50K, for a T eff of 6500K,
The InfraRed Flux Method (IRFM), devel- arises from the uncertainty in the absolute cal-
oped by Blackwell & Shallis (1977) and ibration alone.
Blackwell, Petford & Shallis (1980), can be
used to determine T eff . The method relies on 3.4. Flux Fitting
the fact that the stellar surface flux at an in-
frared wavelength (λ0 ) is relatively insensitive The emergent flux distribution of a star is re-
to temperature. The method is almost model lated to its atmospheric parameters. We can
independent (hence near fundamental), with use spectrophotometry to determine values for
only the infrared flux at the stellar surface, these parameters, by fitting model atmosphere
φ(T eff , log g, λ0 ), requiring the use of model fluxes to the observations. Figure 2 shows the
134 Smalley: T eff and log g determinations

Fig. 1. The Smalley & Dworetsky (1995) uvbyβ photometry grids

 Smalley: T eff and log g determinations 135

sensitivity of the flux distribution to the var- agree. This gives a line in a T eff – log g di-
ious atmospheric parameters. However, inter- agram.
stellar reddening must be allowed for, since it Excitation Potential
can have a significant effect on the observed Abundances from the same element and
flux distribution and derived parameters. ionization stage should agree for all exci-
The currently available optical flux dis- tation potentials
tributions need are in need of revision. This Microturbulence
is something that will be done by ASTRA The same abundance of an element should
(Adelman et al. 2005). be obtained irrespective of the equivalent
width of the lines. This is the technique
used to obtain the mictroturbulence param-
3.5. Balmer Profiles eter (ξturb ). See Magain (1984) for a dis-
The Balmer lines provide an excellent T eff di- cussion of the systematic errors in micro-
agnostic for stars cooler than about 8000 K due turbulence determinations. Typically, can
to their virtually nil gravity dependence (Gray expect to get ξturb to no better than ± 0.1
1992; Heiter et al. 2002). By fitting these the- km s−1 .
oretical profiles to observations, we can deter-
mine T eff . For stars hotter than 8000 K, how-
ever, the profiles are sensitive to both temper- Using these techniques it is possible to get
ature and gravity. For these stars, the Balmer a self-consistent determination of a star’s at-
lines can be used to obtain values of log g, pro- mospheric parameters.
vided that the T eff can be determined from a
different method.
3.8. Global Spectral Fitting
3.6. Spectral Line Ratios
An alternative to a detailed analysis of individ-
Spectral lines are sensitive to temperature vari- ual spectral line measurements, is to use the
ations within the line-forming regions. Line whole of the observed stellar spectrum and find
strength ratios can be used as temperature di- the best-fitting synthetic spectrum. The normal
agnostics, similar to their use in spectral clas- procedure is to take a large multi-dimensional
sification. Gray & Johanson (1991) used line grid of synthetic spectra computed with var-
depth ratios to determine stellar effective tem- ious combinations of T eff , log g, ξturb , [M/H]
peratures with a precision of ±10 K. While and locate the best-fitting solution by least
this method can yield very precise relative squares techniques.
temperatures, the absolute calibration on to
The benefit of this method is that it can be
the T eff scale is much less well determined
automated for vast quantities of stellar obser-
(Gray 1994). This method is ideal for inves-
vations and that it can be used for spectra that
tigating stellar temperature variations (Gray &
are severely blended due to low resolution or
Livingston 1997).
rapid rotation.
Naturally, the final parameters are model
3.7. Metal Line Diagnostics dependent and only as good as the quality of
In a detailed spectral analysis, the equivalent the model atmospheres used. The internal fit-
width of many lines are often measured. These ting error only gives a measure of the precision
can be used to determine the atmospheric pa- of the result and is thus a lower limit uncer-
rameters via metal line diagnostics. tainty of the parameters on the absolute scale.
Determination of the accuracy of the parame-
Ionization Balance ters requires the assessment of the results of fit-
The abundances obtained from differing ting, using the exact same methods, to spectra
ionization stages of the same element must of fundamental stars.
136 Smalley: T eff and log g determinations

Fig. 2. The sensitivity of flux distributions to T eff , log g and [M/H]. The base model (T eff = 7500, log g =
4.0, [M/H] = 0.0) is represented by a solid line. The dotted and dashed lines indicate models with one of
the parameters adjusted, as indicated.
 Smalley: T eff and log g determinations 137

Fig. 3. Sensitivity of Balmer profiles to [M/H] and v sin i. The synthetic spectra (solid line) have been
calculated with T eff = 7500 and log g = 4.0 and a simulated resolution of around 0.4Å. The true shape of
the Hγ profile (with no metal lines) is shown as the dotted line.
138 Smalley: T eff and log g determinations

4. Parameters of Individual Stars later star or a compact object (Patience et al.

1998).
In this section the atmospheric parameters of
some individual stars are presented. Figure 4 shows a T eff –log g diagram for 63
Tau. This is a great visualization tool, since it
allows you to view the relative positions of so-
4.1. Procyon lutions from the different methods. Using such
a diagram it is easy to see how varying var-
Procyon is a spectroscopic binary, with a pe- ious other parameters, such as [M/H], affects
riod of 40 years. The companion is a white the relative positions of the various solutions.
dwarf. This bright F5IV-V star is a very use-
In theory all diagnostics should give a
ful fundamental star. Using f⊕ = 18.0 ± 0.9
unique T eff and log g solution. However, in
× 10−9 W m−2 (Steffen 1985) and θ = 5.448
practice there is a region in T eff and log g space
± 0.053 mas (Kervella et al. 2004b), we get
that contains the solution and its uncertainty.
T eff = 6530 ± 90K. Accurate masses of the
In the case of 63 Tau, the best fitting solution
two components were obtained by Girard et al.
is T eff = 7400 ± 200K and log g = 4.2 ± 0.1 for
(2000), who gave M = 1.497 ± 0.037 M for
[M/H] = +0.5.
the primary. Kervella et al. (2004b), however,
used the HIPPARCOS parallax to revise the
mass to M = 1.42 ± 0.04 M . The radius is ob- 4.4. 53 Cam
tained from the angular diameter and distance:
R = 2.048 ± 0.025 R (Kervella et al. 2004b). The magnetic Ap star 53 Cam has a rotation
These give log g = 3.96 ± 0.02 (Kervella et al. period of 8 days and spectroscopic binary or-
2004b). bital period of 6 12 years (Hoffleit & Warren
1991).
4.2. Arcturus Photometric calibrations give discrepant
results: uvbyβ grid of Moon & Dworetsky
The cool K1.5III giant Arcturus is another (1985) gives 10610 ± 130 K and 4.06 ± 0.05,
important fundamental star. The total flux at while the uvby grid of Smalley & Kupka
the earth was determined by Griffin & Lynas- (1997) gives 8720 ± 250 K and 4.76 ± 0.13
Gray (1999) to be f⊕ = 49.8 ± 0.2 × 10−9 for [M/H] = +1.0 and the Kunzli et al. (1997)
W m−2 , which implies an uncertainty of <1%! Geneva calibration gives 8740 ± 90 K, 4.44 ±
Using θ = 21.373 ± 0.247 mas obtained by 0.10.
Mozurkewich et al. (2003), we get T eff = 4250 Available flux measurements yield f⊕ =
± 25K. (Griffin & Lynas-Gray 1999) 9.19 ± 0.73 × 10−11 W m−2 . Using the IRFM
The model atmosphere analysis by Decin T eff = 8200 ± 250 K is obtained for a single star
et al. (2003) gave T eff = 4320 ± 140K and solution. A binary solution would give 8600K
log g = 1.50 ± 0.15. Their T eff is consistent for a 6500 K main-sequence secondary, which
with the fundamental value. Griffin & Lynas- is in agreement with some of the photometric
Gray (1999) found log g = 1.94 ± 0.05. results.
Verhoelst et al. (2005) presented a discus-
The analysis by Kochukhov et al. (2004)
sion on the possible presence of a binary com-
gave T eff = 8400 ± 150 K and log g = 3.70 ±
panion (see also Griffin 1998).
0.10, but they found that the results from spec-
trophotometry and Balmer lines discordant.
4.3. 63 Tau This demonstrates the important difference be-
tween effective temperature as indicated by the
Situated in the Hyades open cluster, 63 Tau emergent fluxes and that obtained from line-
is a classical Am star with a spectroscopic bi- forming regions. If the model used is not ap-
nary period of 8.4 days. The companion has not propriate to the physical structure of the star’s
been detected, and it is either a cool G-type or atmosphere, then the results will disagree.
 Smalley: T eff and log g determinations 139

Fig. 4. A T eff –log g diagram for 63 Tau. The results from four methods are shown as follows: the filled
square is from the Moon & Dworetsky (1985) uvbyβ girds, the filled circle is from spectrophotometric flux
fitting, the dashed line that from fitting Hβ profiles and the dotted line the IRFM result. Photometry and
Balmer lines agree very well, but are significantly hotter than the results from Spectrophotometry and the
IRFM. The solid arrows indicate the effect of using [M/H] = +0.5 models. Now spectrophotometry is in
good agreement with photometry and Balmer line, but the IRFM is still significantly lower. However, by
introducing a cool companion (5000 K) the IRFM can be brought into agreement with the other methods
(dotted arrow). The solid line is the Hyades isochrone, based on the evolutionary calculations of Schaller et
al. (1992). (Adapted from Smalley 1996)

5. Conclusions Naturally, the exact size of the uncertainty will

depend upon the sensitivity of the lines used in
The atmospheric parameters of a star can be the analysis.
obtained by several techniques. By using a It may appear strange, but the effective
combination of these techniques we can as- temperature of a star is not important; it is the
sess the quality of our parameter determina- T(τ0 ) relationship that determines the spectral
tions. While some techniques can give precise characteristics (Gray 1992). Hence, the pa-
parameter determinations, the overall accuracy rameters obtained from spectroscopic methods
of the values is significantly less and some- alone may not be consistent with the true val-
times difficult to evaluate. Realistically, the ues as obtained by model-independent meth-
typical errors on the atmospheric parameters of ods. This is not necessarily important for abun-
a star, will be T eff ±100 K (1∼2%) for T eff and dance analyses of stars, but it is an issue when
±0.2 dex (∼20%) for log g. For a typical mic- using the parameters to compare with funda-
troturbulence uncertainty of ± 0.1km s−1 , these mental values or to infer the physical proper-
uncertainties give rise to errors of the order of ties of stars.
0.05 ∼ 0.1 dex in abundance determinations.
140 Smalley: T eff and log g determinations

References Glushneva I.N., Kharitonov A.V., Knyazeva

L.N., Shenavrin V.I., 1992, A&AS, 92, 1
Adelman S.J., Pyper D.M., Shore S.N., White Gray D.F., 1992, The observation and analysis
R.E., Warren W.H., 1989, A&AS, 81, 221 of stellar photospheres, CUP.
Adelman S.J., Gulliver A.F., Smalley B., Gray D.F., 1994, PASP, 106, 1248
Pazder J.S., Younger P.F. et al., 2004, IAUS, Gray D.F., Johanson H.L., 1991, PASP, 103,
224, 911 439
Adelman S.J., Gulliver A.F., Smalley B., Gray D.F., Livingston W.C., 1997, ApJ, 474,
Pazder J.S., Younger P.F. et al., 2005, these 802
proceedings. Griffin R.F., 1998, Obs, 118, 299
Alcock C., Allsman R.A., Alves D.R., Axelrod Griffin R.F., Lynas-Gray A.E., 1999, A&A,
T.S., Becker A., et al., 2001, Nature, 414, 282, 899
617 Hayes D.S., Latham D.W., 1975, ApJ, 197, 593
Alonso A., Arribas S., Martínez-Roger C., Heiter U., Kupka F., van’t Veer-Menneret C.,
1994, A&A, 282, 684 Barban C., Weiss W.W., et al. 2002, A&A,
Alonso A., Arribas S., Martínez-Roger C., 392, 619
1996, A&A, 313, 873 Hoffleit D., Warren Jr W.H, 1991, The Bright
Andersen J., 1991, A&A Rev., 3, 91 Star Catalogue, 5th Revised Ed. [V/50]
Blackwell D.E., Lynas-Gray A.E., 1994, A&A, Houdashelt M.L., Bell R.A., Sweigart A.V.,
282, 899 2000, AJ, 119, 1448
Blackwell D.E., Shallis M.J., 1977, MNRAS, Jamar C., Macau-Hercot D., Monfils
180, 177 A., Thompson G.I., Houziaux L.,
Blackwell D.E., Petford A.D., Shallis M.J., Wilson R., 1976, Ultraviolet Bright-
1980, A&A, 82, 249 Star Spectrophotometric Catalogue, ESA
Blackwell D.E., Petford A.D., Arribas S., SR-27 [III/39A]
Haddock D.J., Selby M.J., 1990, A&A, 232, Jiang G., DePoy D.L., Gal-Yam A., Gaudi
396 B.S., Gould A., et al., 2004, ApJ, 617, 1307
Bohlin R.C., Gilliland R.L., 2004, AJ, 127, Kervella P., Thévenin F., Di Folco E.,
3508 Ségransan D., 2004a, A&A, 426, 297
Breger M., 1976, ApJS, 32, 7 Kervella P., Thévenin F., Morel P., Berthomieu
Burnashev V.I., 1985, Abastumanskaya G., Bordé, Provost J., 2004b, A&A, 413, 251
Astrofiz. Obs. Bull. 59, 83 [III/126] Kinman T., Castelli F., 2002, A&A, 391, 1039
Castelli F., Kurucz R.L., 1994, A&A, 281, 817 Kochukhov O., Bagnulo S., Wade G.A.,
Ciardi D.R., van Belle G.T., Akeson R.L., Sangalli L., Piskunov N., et al., 2004, A&A,
Thompson R.R., Lada E.A., Howell S.B., 414, 613
2001, ApJ, 559, 1147 Kunzli M., North P., Kurucz R.L., Nicolet B.,
Clem J.L., VandenBerg D.A., Grundahl F., Bell 1997, A&AS, 122, 51
R.A., 2004, AJ, 127, 1227 Lastennet E., Valls-Gabaud D., 2002, A&A,
Code A.D., Holm A.V., Bottemiller R.L., 1980, 396, 551
ApJS, 43, 501 [II/83] Lester J.B., Gray R.O., Kurucz R.L., 1986,
Decin L., Vandenbussche B., Waelkens C., ApJS, 61, 509
Eriksson K., Gustafsson B., et al., 2003, Macau-Hercot D., Jamar C., Monfils A.,
A&A, 400, 679 Thompson G.I., Houziaux L., Wilson R.,
ESA, 1997, The Hipparcos and Tycho 1978, Supplement to the Ultraviolet Bright-
Catalogues, ESA SP-1200 Star Spectrophotometric Catalogue, ESA
Fernie J.D., 1981, PASP, 93, 333 SR-28 [II/86]
Gezari D.Y., Pitts P.S., Schmitz M., 1999, Magain P., 1984, A&A, 134, 189
Unpublished [II/225] Mégessier C., 1994, A&A, 289, 202
Girard T.M., Wu H., Lee J.T., Dyson S.E., van Mégessier C., 1995, A&A, 296, 771
Altena W.F., et al., 2000, AJ, 119, 2428 Mendoza B., 2005, AdSpR, 35, 882
 Smalley: T eff and log g determinations 141

Moon T.T., Dworetsky M.M., 1985, MNRAS, Smalley B., 1996, ASP Conf, 108, 43
217, 305 Smalley B., Dworetsky M.M., 1995, A&A,
Mozurkewich D., Armstrong J.T., Hindsley 293, 446
R.B., Quirrenbach A., Hummel C.A. et al., Smalley B., Kupka F., 1997, A&A, 328, 349
2003, AJ, 126, 2502 Smalley B., Gardiner R.B., Kupka F., Bessell
Pasinetti-Fracassini L.E., Pastori L., Covino S., M.S., 2002, A&A, 395, 601
Pozzi A., 2001, A&A, 367, 521 [II/224] Southworth J., Maxted P.F.L., Smalley B.,
Patience J., Ghez A.M., Reid I.N., Weinberger 2005, A&A, 429, 645
A.J., Matthews, K., 1998, AJ, 115, 1972 Steffen M., 1985, A&AS, 59, 403
Perevozkina E.L., Svechnikov M.A., 1999 Thompson G.I., Nandy K., Jamar C., Monfils
[V/118] A., Houziaux L., Carnochan D.J., Wilson R.,
Popper D.M., 1980, ARA&A, 18, 115 1978, The Science Research Council, U.K.
Ramírez I., Meléndez J., 2005a, ApJ, 626, [II/59B]
4446 VandenBerg D.A, Clem J.L., 2003, AJ, 126,
Ramírez I., Meléndez J., 2005b, ApJ, 626, 465 778
Richichi A., Percheron I., Khristoforova M., Vasil’Yev I.A., Merezhin V.P., Nalimov V.N.,
A&A, 431, 773 Novosyolov V.A., 1989, IBVS, 3308, 1
Schaller G., Schaerer D., Meynet G., Maeder Verhoelst T., Bordé P.J., Perrin G., Decin L.,
A., 1992, A&AS, 96, 269 Eriksson K., et al., 2005, A&A, 435, 289
Sekiguchi M., Fukugita M., 2000, AJ, 120, Zahid H.J., Hudson H.S., Fröhlich C., 2004,
1072 Solar Physics, 222, 1
Smalley B., 1993, MNRAS, 265, 1035