Multiwavelength optical observations of two chromospherically active binary systems: V789 Mon and GZ Leo

0.94 MB·English

by  M. C. Gálvez

Multiwavelength optical observations of two chromospherically
active binary systems: V789 Mon and GZ Leo∗
M.C. G´alvez1,2, D. Montes2, M.J. Fern´andez-Figueroa2, E. De Castro2 and M. Cornide2
9
0 [email protected]
0
2

n
a
J ABSTRACT

2
This paper describes a multiwavelengh optical study of chromospheres in two X-ray/EUV
1
selected active binary stars with strong Hα emission, V789 Mon (2RE J0725-002) and GZ Leo

] (2RE J1101+223). The goal of the study is to determine radial velocities and fundamental
R
stellar parameters in chromospherically active binary systems in order to include them in the
S
activity-rotationand activity-age relations. We carried out high resolution echelle spectroscopic
.
h observationsandappliedspectralsubtractiontechniqueinordertomeasureemissionexcessesdue
p tochromosphere. Thedetailedstudyofactivityindicatorsallowedustocharacterizethepresence
-
of different chromospheric features in these systems and enabled to include them in a larger
o
r activity-rotationsurvey. Wecomputedradialvelocitiesofthesystemsusingcrosscorrelationwith
t
s the radial velocity standards. The double-line spectral binarity was confirmed and the orbital
a solutionsimprovedforbothsystems. Inaddition,otherstellarparameterssuchas: spectraltypes,
[
projected rotational velocities (vsini), and the equivalent width of the lithium Li i λ6707.8 ˚A

1 absorption line were determined.
v
4 Subject headings: binaries: spectroscopic – stars: activity – stars: individual (V789 Mon, GZ Leo)
3
6 1. Introduction tation and atmospheric properties of companions.
1
In order to understand the physical origin of
. Whyeachparticularstarreachesacertainlevel
1 activity, the time-scale of variability, and to be
0 of activity, and why seemingly similar stars may
able to extrapolate to other systems, we initiated
9 havedrasticallydifferentchromospheres? Itis be-
a multiwavelenghoptical survey of a large sample
0 lievedthatdifferentialrotationandconvectionare
: of binaries with various levels of activity. We car-
v the keyfactorsresponsibleforthe magneticactiv-
ried out the study of the chromosphere of active
i ityinthestar. Asignificantfractionofstellarpop-
X binary systems using the informationprovidedby
ulation, especially at later evolutionary stages, is
r severalopticalandnearinfraredspectroscopicfea-
a foundinclosebinaries,wheretidalforcesandmass
tures that are formed at different heights in the
transfer become additional factors controlling ro-
chromosphere(seeMontesetal. 1997,1998,2000;
G´alvez et al. 2002, 2007), and applied it to sev-
1Centre for Astrophysics Research, Science and Tech-
eralknownandnewlydiscoveredbinaries,mostof
nologyResearchInstitute,UniversityofHertfordshire,Hat-
fieldAL109AB,UK which are BY Dra or RS CVn type with various
2Departamento de Astrof´ısica, Facultad de Ciencias levels of activity.
F´ısicas, Universidad Complutense de Madrid, E-28040 The ultimate goal of the project is to establish
Madrid,Spain
areliableactivity-rotationrelationinbinarystars.
*Based on observations collected with the 2.2 m tele-
To achieve high accuracy one needs to investigate
scopeattheCentroAstrono´micoHispanoAlema´n(CAHA)
at Calar Alto (Almer´ıa, Spain), operated jointly by the all possible causes of activity for each particular
Max-Planck Institut fu¨r Astronomie and the Instituto de system and for each component in the system. It
Astrof´ısicade Andaluc´ıa (CSIC) and withthe 2.1 m Otto
is essential to take into consideration dispersion
StruveTelescopeatMcDonaldObservatoryoftheUniver-
and signal-to-noise of the spectral data obtained
sityofTexasatAustin(USA)
1
with different instrumental set-ups. Furthermore, D & D lines. The latter is closer to the K1V
1 2
thestudyofbinariesascomparedtosinglestarsis type determined by Fekel & Henry (2000) from
complicated by the interaction between compan- the temperature and luminosity-sensitive line ra-
ions, line blending (SB2 case), reflected light, etc. tios in the 6430-6465˚A region. Popper (1996) re-
Active stars, which are recognized by the pres- portedphotometricvariability,butcouldnotcon-
ence of emission lines in one or both components, firm whether it is due to eclipses. Strassmeier
present the most challenging case and need to be et al. (2000) determined a period of 2.944 days
studied on the individual basis. To obtain an ac- from their photometric observations, but noted
curate orbital solution and to study the behav- that a period of 1.514 days was also possible. Fi-
ior of activity,one needs to separateradialveloci- nally, Fekel & Henry (2000) and Pandey et al.
ties,equivalentwidths(EW hereafter)ofemission (2002, 2005) confirmed that the photometric pe-
lines, and fluxes between two components. It is riod (1.526 days) is nearly identical to its orbital
also vital to carry out a precise stellar classifica- period. This binary system shows the typical sig-
tionduetotheimportanceofthesubgiantandgi- natures of chromospheric activity. A filled-in Hα
ant presence in the activity-rotation relation (e.g. absorption line in both components was reported
in RS CVn case). by Jeffries et al. (1995), while Ca ii H & K emis-
Here we concentrate on two recently discov- sions were detected in the spectra by Strassmeier
ered, X-ray/EUV selected, chromospherically ac- et al. (2000). In addition, GZ Leo is a young sys-
tive double-line spectroscopic binaries included in tem as indicated by its large lithium abundance
this survey: V789 Mon (2RE J0725-002, BD-00 (Strassmeier et al. 2000; Fekel & Henry 2000).
1712) and GZ Leo (2RE J1101+223, HD 95559, A preliminary analysis of these two systems
BD+23 2297, HIP 53923). can be found in G´alvez et al. (2003) and G´alvez
V789 Mon, V = 9.33, was classified by Jef- (2005).
fries et al. (1995) as a double-line spectroscopic In§2wesummarizeourobservations. In§3we
binary (SB2) with almost identical K5V compo- derivestellarparametersandtheorbitelementsof
nents. They gave a rotational velocity, vsini, the systems. The behavior of the different chro-
of 25 kms−1 and an orbital period of 1.40 days. mospheric activity indicators is described in § 4.
The photometric period (1.412 days) reported by Finally § 5 gives a summary.
Cutispoto et al. (1999) and Robb & Gladders
(1996) indicates synchronous rotation. Jeffries et 2. Observations
al. (1995)suggestthatthissystemcouldbeeclips-
In this paper we analyzed high resolution
ing,but Robb& Gladders(1996)didnotfindany
echelle spectra of both systems that we obtained
evidence of eclipses in their photometric observa-
during two observing runs detailed in Table 1.
tions. Cutispoto et al. (1999) derived a K3V +
The spectrawereextractedusingthestandardre-
M0:Vclassificationforthissystemfromcolorsand
duction procedures in the IRAF1 echelle package
spectral signatures. Regarding activity, Jeffries et
(bias subtraction, flat-field division and optimal
al. (1995) detected the Hα line in emission above
extractionofthe spectra). We obtainedthe wave-
the continuum level for both components.
length calibration by taking spectra of a Th-Ar
GZLeoisaSB2binarywithV =8.8mag. Jef-
lamp. Finally, we normalized the spectra by a
fries et al. (1995) obtained a first orbital solution
polynomial fit to the observed continuum.
witha periodof1.528daysusingelevenradialve-
We obtained nine spectra of V789 Mon in one
locitymeasurements. Later,Fekel&Henry(2000)
observingrunandfifteenspectraofGZLeointwo.
reported an improved orbital solution using nine
Thesignaltonoiseratio(S/N)achievedintheHα
new radial velocity determinations and one addi-
lineregioninallthespectrais≈60forV789Mon
tional value given by Strassmeier et al. (2000).
and ≈ 100 for GZ Leo.
The system consists of two similar rapidly rotat-
ing stars (vsini ≈ 30 kms−1). The spectral type
1IRAF is distributed by the National Optical Observatory,
of the components is not well established. Jef-
whichisoperatedbytheAssociationofUniversitiesforRe-
fries et al. (1995) classified it as G5 while Popper searchinAstronomy,Inc.,undercontractwiththeNational
(1996) as K2, based on the strength of the Na i ScienceFoundation.
2
3. Stellar parameters of the binary sys- withtheequalcontributionfromeachcomponent.
tems Similarlyto V789Mon,the fits wereperformedin
allthephaseswherethecomponentscouldbesep-
We list the adopted stellar parameters of both
arated, and several spectral types and luminosity
systems in Table 2. Spectral type and the photo-
class reference stars were used. Our spectral clas-
metricdata(B−V,V,P )aretakenfromSIM-
phot sification is very close to the K2V type reported
BAD and Cutispoto et al. (1999) for V789 Mon
by Popper (1996) and the K1V type by Fekel &
and from SIMBAD and Fekel & Henry (2000) for
Henry (2000).
GZLeo. Theastrometricdata(parallax,π;proper
motions, µαcosδ and µδ) are from Hipparcos 3.2. Rotational velocity
(ESA 1997) and Tycho−2 (Høg et al. 2000) cat-
Jeffries et al. (1995) estimated the projected
alogues. For V789 Mon no parallax value is avail-
rotational velocity (vsini) for V789 Mon as
able, so we give the spectroscopic parallax deter-
25kms−1. For GZ Leo they found 27 kms−1,
mined here (see § 3.4). Finally, the orbital pe-
while Fekel & Henry (2000) reported values of
riod (P ) and the projected rotational velocities
orb 32.4 and 31.6 kms−1 for each component. Fekel
(vsini)havebeendeterminedinthispaperaswell
& Henry (2000) noted that the actual vsini may
(see below).
slightly smaller, because spots cause the lines to
3.1. Spectral types be shallower than in an unspotted star, making
FWHM appear larger. Strassmeier et al. (2000)
In order to obtain the spectral type of these report similar values of 31 and 26 kms−1 respec-
binary systems we compared our high resolution
tively.
echelle spectra with spectra of inactive reference
Using starmod we obtained for each observ-
stars of different spectral types and luminosity ing run vsini values about 30 kms−1 for both
classes observed during the same observing run.
components of V789 Mon and GZ Leo. In or-
To avoid contamination by chromospheric emis-
der to determine a more accurate rotational ve-
sion, we only used spectral orders that are free
locities we applied a cross-correlation technique
of lines sensitive to chromospheric activity. This
toourhighresolutionechellespectraofthesestars
analysis makes use of the program starmod de-
usingIRAFtaskfxcor. Themethodisdescribed
veloped at Penn State University (Barden 1985)
in detail in our previous papers (see G´alvez et
and modified by us. starmod allows to rotation-
al. 2002; Lo´pez-Santiago et al. 2003). In short,
allybroadenspectraandshifttheminthevelocity
the method is based on the fact that when a stel-
space. We ran it on the spectra of the reference
lar spectrum with rotationally broadened lines is
stars and combined with the appropriate weights
cross-correlated against a narrow-line spectrum,
tocreatecompositespectrathatwereusedastem-
the width of the cross-correlationfunction (CCF)
plate for comparisonwith our targets.
will depend on the amount of rotational broaden-
ForV789Monweobtainedthebestfitbetween
ing of the former spectrum.
observed and synthetic spectra using a K5V ref-
For V789 Mon, we used as a template a slowly
erence star (HR 1614) for both components, con-
rotating K5Vstar HR 1614. We obtaineda vsini
tributing respectively 60% and 40% of the total of 28.28±1.59 and 25.09±2.06 kms−1 for the pri-
fluxinthecontinuum. Thefitswerecarriedoutin
mary and the secondary component respectively.
allthephaseswherethespectraofthecomponents
In the case of GZ Leo, we used a K0V star
could be separated. Our spectral classification is
HR 166 and a K2V star HD 166620as templates,
in agreement with a K5V type reported by Jef-
in McDonald and FOCES observing runs respec-
fries et al. (1995), while the relative contribution
tively. Weobtainedaveragedvaluesof26.23±1.13
of the companions is compatible with Cutispoto
and 26.92±1.14 kms−1 for each component (see
et al. (1999) result that assigned a later spectral
Table 2).
type for the secondary component.
In the case of GZ Leo, we obtained the best
fitbetweenobservedandsyntheticspectrausinga
K0Vreferencestar(HR166)forbothcomponents,
3
3.3. Radial velocities and orbital solution ponents are very similar. The obtained parame-
ters are in agreement with Jeffries et al. (1995).
Heliocentricradialvelocitieswerealsoobtained
See Table 5 for details.
by using the cross-correlation technique (see e.g.
WearrivedatasimilarconclusionsforGZ Leo:
G´alvezetal. 2007). Thespectraofthetargetwere
a nearly circular orbit (e = 0.0073) with an or-
cross-correlated order by order, using the routine
fxcor in IRAF, againstspectra ofradialvelocity bital period of 1.5260 days, which is very sim-
ilar to the photometric rotational period (P
standards with similar spectral type taken from phot
≈ 1.5264 days, Fekel & Henry 2000), indicating
Beavers et al. (1979). We derived the radial ve-
a synchronized rotation. The obtained minimum
locity for each order from the position of peak of
masses (Msin3i) and the mass ratio (q =1.0139)
the cross-correlation function (CCF) and calcu-
are in agreement with Fekel & Henry (2000). See
lated the uncertainties based on the fitted peak
Table 6 for computed parameters.
height and the antisymmetric noise as described
by Tonry & Davis (1979).
3.4. Physical Parameters
Since both systems are SB2, we could see two
peaks in the CCF associated with two compo- Adopting spectral types derived in § 3.1 for
nents; we fitted each peak separately. When the primary components and adopting masses
the component peaks were too close, we used de- and temperatures from Landolt-B¨ornstein tables
blending technique. (Schmidt-Kaler 1982), we derived masses and
spectral types for the secondaries.
IntheTables3and4welist,foreachspectrum,
the heliocentric radial velocities (V ) and their We also estimated minimum radii (Rsini), lu-
hel
associatederrors(σ ). The latter areobtained as minosities and minimum masses (Msini) of the
V
weighted means of the individual values deduced components using photometric periods (from the
for each order in V789 Mon and GZ Leo spec- literature,see§3.) androtationalvelocities(from
tra. We note that the uncertainties returned by § 3.2). The results, given in Table 7, confirm our
fxcor for SB2 binaries are overestimated; when previous classification.
fitting each component, the presence of the other Furthermore, since Hipparcos parallax of
component will increase the antisymmetric noise, V789 Mon is not known, we calculated a spec-
thereby biasing the error. troscopic parallax. Using bolometric magnitude
To compute the orbital solution for these sys- of a K5V star (+6.7m), the bolometric correction
tems we combined our radial velocity data (for (BC = −0.72m) corresponding to a K5V, and
both components) with the data of Jeffries et al. the luminosity ratio, we obtain the bolometric
(1995), Fekel & Henry (2000) and Strassmeier et and the visual total absolute magnitudes for the
al. (2000), see Tables 3 and 4. The radial ve- system. Comparingthetotalvisualabsolutemag-
locity data are plotted in Fig. 1 for V789 Mon nitude (6.76m), with the measured V magnitude
and GZ Leo on left and right side respectively. (9.33m), we derive a distance of 32.72 ±0.13 pc
Solid symbols represent the primary and open (π = 30.56 ±0.22 mas) for V789 Mon. Due to
symbols represent the secondary. The curves rep- possible activity-induced variations in the V, we
resent a minimum χ2 fit orbit solution. The or- expect a 10% additional error in this distance
bit fitting code uses the Numerical Recipes (Press measure. Since the system is relatively close, the
et al. 1986) implementation of the Levenberg- interstellar reddening is neglected.
Marquardtmethodoffittinganon-linearfunction Using the total luminosity based on the mea-
to the data and weights each datum according to sureddistance(54.26pc)forGZLeo,weobtained
its associated uncertainty (see e.g. G´alvez et al. the total magnitude for the system V = 8.77.
2002, 2007 for further details). ThedifferencebetweentheobservedV magnitude
We found a nearly circular (e = 0.0129) for (8.92)and the calculatedone canbe explainedby
V789Monsystem,withanorbitalperiodof1.4021 the presence of photospheric cool spots over the
days which confirms it as a synchronous system stellar surfaces of both components.
(P ≈ 1.412 days, Cutispoto et al. 1999). The
phot
mass ratio (q = 1.0542) indicates that both com-
4
3.5. Kinematics (1993) resulting EW(Li i) = 40.9 m˚A. One addi-
tionalestimationoftheEW(Lii)inthisspectrum
We computed the galactic space-velocity com-
can be obtained by using the spectral subtrac-
ponents (U, V, W) of these systems using as ra-
tion technique. In the subtracted spectrum the
dial velocity the center of mass velocity (γ) and
contribution from the photospheric lines is elimi-
accurate proper motions and parallax taken from
nated, therefore the EW = 51.2 m˚A measured in
Hipparcos (ESA 1997)and Tycho−2 (Høg et al.
this spectrum is the total EW(Li i) and, as both
2000)catalogues(seeTable2),exceptforspectro-
components are equal, also the EW(Li i) of each
scopic parallax of V789 Mon that has been calcu-
one. ThedifferenceintheEW obtainedfromboth
lated here.
methodsislikelyduetotheinfluenceofthestellar
The obtained values and associated errors are
metallicity in the calibration used in the first one
given in Table 8. The velocity components lie in to obtain the EW(Fe i) and in the standard star
the (U, V) diagram obviously inside the young
usedasreferenceinthesubtractiontechniquethat
disk population boundaries (Eggen 1984, 1989; are not taken into account here. The EW(Li i)
Montes et al. 2001a, 2001b) indicating that both of each component we have determined (40.9 m˚A
systems are young disk stars. In addition, the with the first method and 51.2 m˚A with the sec-
EggenkinematiccriteriapredictthatGZLeomay ond one) is smaller than the 63 m˚A estimated by
be a member of Hyades supercluster (see Montes Fekel & Henry (2000) and the 60 m˚A given by
et al. 2001a and G´alvez 2005).
Strassmeier et al. (2000). The small differences
between these values can be attributed to the dif-
3.6. The Li i λ6707.8 line
ferent method used by these authors to estimate
The resonancedoubletof Lii atλ6707.8˚A is a the Fe i contributionand the errorinthe EW de-
spectroscopic feature very important in the diag- termination.
nosticofageinlate-typestars,sinceitisdestroyed This small EW(Li i) is in agreement with the
easilybythermonuclearreactionsinthestellarin- values observed in Hyades cluster stars of this
terior. spectral type (see Montes et al. 2001b) and with
Thespectralregionofthislineisincludedinall the membership to the Hyades supercluster (see
our spectra for both systems. § 3.5). That situated the system age in about
600 Myr.
At the spectral resolution of our spectra and
with the rotational velocity, we have determined
for the components of both binaries, the Li i line 4. Chromospheric activity indicators
is blended with the nearby Fe i λ6707.41 ˚A line.
The echelle spectra analyzed in this work al-
In the case of V789 Mon the Li i line is very
lowed us to study the behavior of the different
weak and we could not measure the EW with the chromospheric activity indicators, from the Ca ii
required precision. For GZ Leo, we clearly see H&KtotheCaiiIRTlines,whichareformedat
the Li i+Fe i absorption feature in all the spec-
different atmospheric heights. The chromospheric
tra (see Fig. 2 left side), but due to the blending
contribution in these features were determined by
withotherphotosphericlinesofbothcomponents,
usingthespectralsubtractiontechniquedescribed
we have only measured the EW in the spectrum
indetailbyMontesetal. (1995,1997,1998,2000)
of the third night of the FOCES02 observing run
and G´alvez et al. (2002, 2007).
(see Fig. 2 right) which is very close to conjunc-
tion (ϕ = 0.02). The obtained value EW(Li i + The excess emission in EW is measured in the
Fe i) = 60.7 m˚A contain the contribution from subtractedspectra,aftercorrectingfortherelative
contribution of each component in the total con-
both components, but taken into account that
tinuum. This contribution is determined by using
they have the same spectral type and contribu-
the radii and temperatures adopted in § 3. For
tion to the continuum we can assume that this is
also the EW(Li i+Fe i) of each component. We instance, in the Hα line regionthe relative contri-
butions were S = 0.61 for the primary compo-
have corrected this total EW by subtracting the P
EW(Fe i) = 19.8 m˚A calculated from the empiri- nent and SS = 0.39 for the secondary component
inV789MonandforGZLeowehadthe0.50con-
calrelationshipwith(B–V)givenbyFavataetal.
5
tributions for both components. spectra indicates that it is a very active binary
In Table 9 we give the corrected EW for the system similar to some RS CVn and BY Dra sys-
Hα and Ca ii IRT lines for V789 Mon. In Ta- tems, which always show Hα emission above the
ble 10 we list corrected EW for the Ca ii H & continuum. The EW measured in the subtracted
K, Hǫ, Hδ, Hγ, Hβ, Hα, and Ca ii IRT (λ8498, spectra by Gaussianfitting gives anaveragevalue
λ8542, λ8662 ˚A) lines for GZ Leo in FOCES run of EW(Hα) = 0.65/0.62 ˚A for primary and sec-
and for Hα and Ca ii IRT in McDonald run. The ondarycomponents(whichishigherthanreported
corrected EWs is given separately for each com- byJeffriesetal. (1995),EW(Hα)=0.09/0.25˚A).
ponent(P/S),unlessthelinesofbothcomponents In GZ Leo, we analyzed 15 spectra including
overlap,in whichcase the combined EW is given. theHαlineregion. Inthefirstrun(McDonald98),
The finalestimatederrorsinEW areintherange the Hα line appears completely filled by emission
10-20% (see e.g. G´alvez et al. 2007 for details). in both components (see Fig. 3, left panel of the
Finally, corrected EWs were converted to the right side). As in the case of V789 Mon, the
absolutesurfacefluxesbyusingtheempiricalstel- emission from both components is nearly identi-
lar flux scales calibratedby Hall (1996)as a func- calandwasmeasuredbyapplyingatwo-Gaussian
tion of the star color index. In our case, we used fit to the subtracted spectra (see Fig. 3, right
theB−V indexandthecorrespondingcoefficients panel of the right side). The spectrum at phase
forCaiiH&K,HαandCaiiIRT.ForHǫweused 0.46 is close to conjunction, so the measured EW
the same coefficients as for Ca ii H& K,while for given in Table 10 is the combined EW. The
Hδ, Hγ and Hβ the coefficients were obtained by mean values of EW are EWP(Hα) = 0.32 ˚A and
interpolating between the values for Ca ii H & K EWS(Hα)=0.36˚A.Wenotehoweverthatavari-
andHα. InTables9and11,wegivethelogarithm ation with phase is observed in both components,
of the calculated absolute flux at the stellar sur- with a remarkable tendency to behave in the op-
face (logF ) for different chromospheric activity posite way for Ca ii IRT EWs; unfortunately the
S
indicators. phasecoverageisnotenoughforareliableconclu-
sion (Table 10).
Figs.3-5showarepresentativeregionsaround
the Hα, Ca ii H & K and Ca ii IRT λ8498, λ8542 When we analyze the FOCES02 run, we see
lines. The observed (solid line) and the synthe- again that the Hα line is completely filled-in by
sized spectra (dashed line) are shown in the left emission for both components. After fitting the
panel, while subtracted spectrum (dotted line) in subtracted spectra we measured the excess emis-
theright. Theobservingrunandtheorbitalphase sion EWP(Hα) = 0.46 ˚A and EWS(Hα) = 0.47
(ϕ) of each spectrum are also shown in these fig- ˚A, i.e., a 35% increase compared to McDonald98
ures. InFig.6weplotarepresentativesubtracted run.
spectra in Hβ, Hγ and Hδ region for GZ Leo. In GZ Leo system the secondary star usually
showsslightlylargervaluesofHαEW andnovari-
4.1. The Hα line ation with phase was detected (see Table 10).
InV789 MonHα is observedin emissionabove
4.2. The Hβ, Hγ and Hδ lines
the continuum in all the observed spectra (see
Fig. 3, left panel of the left side). After apply- The other three Balmer lines (Hβ, Hγ and
ingthespectralsubtractiontechnique,onecansee Hδ) were included only in some of our spectra
thatHαemissionfromtheprimaryandsecondary for GZ Leo during the FOCES02. The absorp-
components looks very much alike. tion lines were clearly filled-in with emission in
Inallthespectra,exceptonewhichisveryclose the observed spectra. After applying the spectral
to conjunction, we measured the emission coming subtraction, we could extract the excess emission
frombothcomponentsbyusingatwo-Gaussianfit fromboth components (see representativespectra
tothesubtractedspectra(seeFig.3,rightpanelof in Fig. 6). When the S/N was high enough we
the left side). The combinedEW is givenwhen it deblended the emissioncoming from both compo-
was not possible to deblend the two components. nents by applying a two-Gaussian fit to the sub-
tracted spectra (see Table 10). These three lines
The persistence of Hα emission in V789 Mon
6
show similar variations with orbital phase to the 4.4. Ca ii IRT lines (λ8498, λ8542, and
Hαlineinbothcomponents,theirmeanvaluesare λ8662)
EW(Hβ) = 0.13/0.13˚A, EW(Hγ) = 0.09/0.09˚A
The three lines of the Ca ii infrared triplet
and EW(Hδ) = 0.07/0.09 ˚A.
(IRT) were included in all our echelle spectra for
We also measured the ratio of excess emission
both systems. In all of them we could observe a
EW inthe Hα andHβ lines, EW(Hα)/EW(Hβ),
clear emission above the continuum in the core of
and the ratio of excess emission EHα/EHβ with the Ca ii IRT absorption lines (see Figs. 5) from
the correction:
both components.
E EW(Hα) InV789Mon,(Fig.5leftside),wemeasuredan
Hα = ∗0.2444∗2.512(B−R) averageEW of≈0.4˚Aforthethreelines,andwe
E EW(Hβ)
Hβ
found smallvariationswith the orbitalphase that
fromHall& Ramsey(1992). Ittakesinto account were anti-correlated with the Balmer lines EW
the absolute flux density in these lines and the variations in both primary and secondary compo-
color difference in the components (B −R=1.45 nents.
for a K0V star). We obtained a mean values of For GZ Leo binary, (Fig. 5 right side), we
EHα/EHβ ≈3.4fortheprimaryand≈3.6forsec- found that in FOCES run there were no signifi-
ondary. Thesevaluesindicate,accordingtoBuzasi cant variations with the orbital phase, but in Mc-
(1989) and Hall & Ramsey (1992) the presence of Donald’soneanevidentorbitalvariationwereob-
prominence-like material at the stellar surface in served which shows an anti-correlation with the
both components of the system. EW(Hα) variation. We measured mean EWs of
≈ 0.25/0.31 ˚A for each component in McDonald
4.3. Ca ii H & K and Hǫ runand≈0.36˚Afor bothcomponentsinFOCES
The Ca ii H & K line region was included only run.
in the spectra of the FOCES run for GZ Leo. In In addition, we calculated the ratio of excess
these spectra we observed strong emission in the emission EW, E8542/E8498, which is also an in-
Ca ii H & K lines and a clear emission in the Hǫ dicator of the type of chromospheric structure
line arising from both components (see Fig. 4). that produces the observed emission. In solar
In our spectra, the Ca ii H & K lines are lo- plages, values of E8542/E8498 ≈ 1.5-3 are mea-
sured, while in solar prominences the values are
cated at the end of the echellogram,where the ef-
≈9,thelimitofanopticallythinemittingplasma
ficiencyofthespectrographandtheCCDdecrease
(Chester 1991).
veryrapidly,andthereforetheS/N ratioobtained
is very low, and the normalization of the spectra In V789 Mon, small values of the E8542/E8498
are very difficult. In spite of this we could apply ratiowerefound,≈1.5fortheprimarycomponent
the spectral subtraction in this region, see Fig. 4. and 1.4 for the secondary (see Table 9). This
As we can see in this figure, the Hǫ line arising indicates that the Ca ii IRT emission of this star
from one of the component overlaps at some or- arises from plage-like regions.
bital phases with the Ca ii H line from the other In GZ Leo we found a E /E value of
8542 8498
component, so their EW were measured with a ≈2.4and≈1.4forprimaryandsecondaryinMc-
Gaussian fit when it was possible (see comments Donald run, but ≈ 0.9 for primary and secondary
for each spectrum in the footnotes of Table 10). in FOCES run. These values indicate that Ca ii
The mean measured EWs in these spectra are IRTemissioncomefromplage-likeregions,incon-
EW=1.13/0.84˚AforeachcomponentintheCaii trasttotheBalmerlineswhichseemtocomefrom
Kline andEW=0.82/0.65˚Aforeachcomponent prominences.
in the Ca ii H line. An increase of the Ca ii H & This markedly different behavior of the Ca ii
K emissions is observed with the orbital phase. IRT and the Hα emission has also been found
ThepresenceofanexcessemissionfromHǫline in other chromospherically active binaries (see
(EW= 0.30/0.43˚A) indicates that it is a very ac- Ar´evalo&La´zaro1999;Montesetal. 2000;G´alvez
tive system. et al. 2003; G´alvez 2005 and references therein).
The different behavior from one epoch to another
7
is consistent with the different activity level mea- sion in the Hα line is above the continuum and
suredin eachobservingrun. The lackofvariation it is very similar in the two components, showing
with orbitalphase ofthe GZ Leoemissionlines in no variations with phase. GZ Leo shows a high
FOCES run could be due to that the hight level level of chromospheric activity too. After apply-
of activity is reflecting a large number of active ing spectral subtraction the chromospheric excess
regionsmaybedistributedalongallsurface. Such, emission from both components are certainly de-
couldproduceanevenspotted-plagedsurfacethat tected in all the activity indicators during the ob-
would avoid to see the modulation clearly. serving runs. The Ca ii H & K lines are in emis-
sion and Hǫ is also detected in emission for both
5. Summary systems.
The level of chromospheric emission change,
In this paper we present a detailed spectro-
i.e.,wehaveobserveda35%incrementintheEW
scopic analysis of two X-ray/EUV selected chro-
of Hα emission line and a nearly 30% increment
mospherically active binary systems V789 Mon
in the EW of Ca ii IRT emission lines from one
(2RE J0725-002) and GZ Leo (2RE J1101+223).
epoch to another.
We analyzed high resolution echelle spectra that
Furthermore,thevariationofHαandCaiiIRT
include the optical chromospheric activity indica-
tors from the Ca ii H & K to Ca ii IRT lines, as emission with phase is anti-correlated and the ra-
wellastheLiiλ6707.8lineandotherphotospheric tios EHα/EHβ and E8542/E8498 indicate that the
emission of the Balmer lines would arise from
lines of interest.
prominence-like material, and that the emission
Usingalargenumberofradialvelocitiesweim-
of Ca ii IRT lines arise from plage-like regions.
proved the orbital parameters of these systems.
We obtained that both systems have nearly cir-
We would like to thank Dr. L.W. Ramsey
cular orbits and have an orbital period very close
(Pennsylvania State University) for collaborating
to photometric period, indicating that they have
in the McDonald observing run (2.1 m telescope,
a synchronous rotation.
Texas, USA) and the staff of McDonald obser-
The spectral classification derived using the
vatory for their allocation of observing time and
comparison with spectra of reference stars, two
their assistance with our observations. Thanks to
K5V components for V789 Mon and two K0V
Dr. Nadya Gorlovaand Dr. JoernRossa for their
componentsforGZLeoisingoodagreementwith
great help on the English redaction and correc-
the classification obtained from physical parame-
tions. ThisworkwassupportedtheSpanish”Pro-
ters (Rsini). As both components are main se-
grama Nacional de Astronom´ıay Astrof´ısica” un-
quence stars, we can classify these chromospheri-
der grants AYA2005-02750 and AYA2008-00695,
cally active binaries as BY Dra systems (Fekel et
and the ”Comunidad de Madrid” under PRICIT
al. 1986).
project S-0505/ESP-0237(ASTROCAM).
Byusingtheinformationprovidedbythewidth
of the CCF we determined a projected rotational REFERENCES
velocity, vsini, of 28.28 and 25.09 kms−1 for the
primaryandsecondarycomponents ofV789Mon, Ar´evalo, M. J., & La´zaro, C. 1999,AJ, 118, 1015
and26.23and26.98kms−1 forthecomponentsof
Barden, S. C. 1985, ApJ, 295, 162
GZ Leo.
The presence of the Li i line in both systems is Beavers, W. I., Eitter, J. J., Ketelsen, D. A., &
inagreementwiththekinematicsresults,i.e.,they Oesper, D. A. 1979,PASP, 91, 698
belong to the youngdisk population. Also,witha
EW(Li i) ≈ 40 m˚A for each component, GZ Leo Buzasi, D. L. 1989, PhD Thesis, Pennsylvania
State Univ.
could be a member of the Hyades supercluster.
We analyzed, using the spectral subtraction Chester, M. M. 1991, PhD Thesis, Pennsylvania
technique, all the optical chromospheric activity State Univ.
indicators. In V789 Mon, both components show
high levels of chromospheric activity. The emis-
8
Cutispoto, G., Pastori, L., Tagliaferri, G., Jeffries, R. D., Bertram, D., & Spurgeon, B. R.
Messina,S.,&Pallavicini,R.1999,A&AS,138, 1995,MNRAS, 276, 397
87
La´zaro, C. & Ar´evalo,M.J. 1997, AJ, 113, 2283
Eggen, O. J. 1984, ApJS, 55, 597
Lo´pez-Santiago, J., Montes, D., Ferna´ndez-
Eggen, O. J. 1989, PASP, 101, 366 Figueroa, M. J., & Ramsey, L. W. 2003, A&A,
411, 489
ESA 1997, The Hipparcos and Tycho Catalogues,
ESA SP-1200 Montes,D.,DeCastro,E.,Ferna´dez-Figueroa,M.
J., & Cornide, M. 1995, A&AS, 114, 287
Favata, F., Barbera, M., Micela, G., & Sciortino,
S. 1993, A&A, 277, 428 Montes, D., Ferna´ndez-Figueroa, M. J., De Cas-
tro, E., & Sanz-Forcada, J. 1997, A&AS, 125,
Fekel, F. C., Bopp, B. W., Africano, J. L.,
263 (Paper I)
Goodrich, B. D., Palmer, L. H., Quigley, R.,
& Simon, T. 1986, AJ, 92, 1150 Montes, D., Sanz-Forcada, J., Ferna´ndez-
Figueroa, M. J., De Castro, E., & Poncet, A.
Fekel,F.C.,&Henry,G.W.2000,ApJ,120,3265
1998,A&A, 330, 155 (Paper II)
G´alvez, M. C., Montes, D., Ferna´ndez-Figueroa,
Montes, D., Ferna´ndez-Figueroa, M. J., De Cas-
M. J., Lo´pez-Santiago, J., De Castro, E., &
tro, E., Cornide, M., Latorre, A., & Sanz-
Cornide, M. 2002, A&A, 389, 524 (Paper IV)
Forcada J. 2000, A&AS, 146, 103 (Paper III)
G´alvez, M. C., Montes, D., Ramsey, L. W.,
Montes, D., Lo´pez-Santiago, J., G´alvez, M. C.,
Ferna´ndez-Figueroa, M. J., De Castro, E., &
Ferna´ndez-Figueroa, M. J., De Castro, E. &
Cornide, M. 2003, Proceedings of the V Scien-
Cornide, M. 2001a, MNRAS, 328, 45
tificMeetingofthe SpanishAstronomicalSoci-
ety(SEA),”HighlightsofSpanishAstrophysics Montes, D., Lo´pez-Santiago, J., Ferna´ndez-
III” Kluwer Academic Publishers (Printed in Figueroa,M. J., & G´alvez, M. C. 2001b,A&A,
Netherlands), ed J. Gallego, J. Zamorano, N. 379, 976
Cardiel, 474
Pandey, J. C., Singh, K. P., Sagar, R., & Drake,
G´alvez, M. C. 2005, PhD Thesis, Universidad S.A.2002,JournalofAstrophysicsandAstron-
Complutense de Madrid omy, 23, 9
G´alvez, M. C., Montes, D., Ferna´ndez-Figueroa, Pandey,J.C.,Singh,K.P.,Drake,S.A.,&Sagar,
M. J., De Castro, E., & Cornide, M. 2007, R. 2005, AJ, 130, 1231
A&A, 472, 587 (Paper V)
Pfeiffer, M. J., Frank, C., Baumueller, D., et al.
Gray,D.F.1992,TheObservationandAnalysisof 1998,A&AS, 130, 381
Stellar Photospheres,2nd ed. (Cambridge Uni-
Popper, D. M. 1996, ApJSS, 106, 133
versity Press)
Press, W. H., Flannery, B. P., & Teukolsky, S. A.
Gunn, A. G. 1996, IAJ, 23, 33
1986, Numerical recipes. The art of scientific
Hall, J. C., & Ramsey, L. W. 1992,AJ, 104, 1942 computing, Cambridge: University Press
Hall, J. C. 1996, PASP, 108, 313 Rhode,K.L.,Herbst,W.,&Mathieu,R.D.2001,
AJ, 122, 3258
Hartmann,L.,Hewett,R.,Stahler,S.,&Mathieu,
R. D. 1986,ApJ, 309, 275 Robb, R. M., & Gladders, M. D. 1996,IAU, 4412
Høg,E.,Fabricius,C.,Makarov,V.V.,etal.2000, Schmidt-Kaler, T. 1982, in Landolt-B¨ornstein,
A&A, 355, L27 Vol. 2b, ed K. Schaifers & H. H. Voig (Heidel-
berg: Springer)
9
Strassmeier, K. G., Hall, D. S., Fekel, F. C., &
Scheck, M. 1993, A&AS, 100, 17
Strassmeier, K. G., Washuettl, A., Granzer, Th.,
Scheck,M.,&Weber,M.2000,A&AS,142,275
Tonry, J., & Davis, M. 1979, AJ, 84, 1511
This2-columnpreprintwaspreparedwiththeAASLATEX
macrosv5.2.
10