Upload
warwick
View
1
Download
0
Embed Size (px)
Citation preview
Mon. Not. R. Astron. Soc. 000, 1–8 (2010) Printed 24 December 2010 (MN LATEX style file v2.2)
Qatar-1b: a hot Jupiter orbiting a metal-rich K dwarf star
K. A. Alsubai1?, N. R. Parley2, D. M. Bramich3, R. G. West4, P. M. Sorensen5,A. Collier Cameron3, D. W. Latham6, K. Horne3, D. R. Anderson7, D. J. A. Brown2,L. A. Buchhave8, G. A. Esquerdo6, M. E. Everett9, G. Furesz6, C. Hellier7,G. M. Miller2, D. Pollacco10, S. N. Quinn6, J. C. Smith11, R. P. Stefanik6,and A. Szentgyorgyi6.1Qatar Foundation, P.O.BOX 5825, Doha, Qatar.2SUPA, School of Physics and Astronomy, University of St Andrews, North Haugh, St Andrews, Fife KY16 9SS, UK.3European Southern Observatory, Karl-Schwarzschild-Straβe 2, 85748 Garching bei Munchen, Germany4Department of Physics and Astronomy, University of Leicester, LE1 7RH, UK.5Nordic Optical Telescope, Apartado 474, E-38700 Santa Cruz de la Palma, Santa Cruz de Tenerife, Spain6Harvard-Smithsonian Center for Astrophysics, 60 Garden Street, Cambridge, MA 02138, USA.7Astrophysics Group, Keele University, Staffordshire, ST5 5BG, UK.8Niels Bohr Institute, Copenhagen University, DK-2100 Copenhagen, Denmark.9Planetary Science Institute, 1700 East Fort Lowell Road, Suite 106, Tucson, AZ 85719, USA.10Astrophysics Research Centre, School of Mathematics & Physics, Queens University, University Road, Belfast, BT7 1NN, UK.11Hidden Loft Observatory, Tucson, AZ 85755, USA.
Accepted 0000 Dec 00. Received 0000 Dec 00; in original form 0000 October 00
ABSTRACTWe report the discovery and initial characterisation of Qatar-1b, a hot Jupiter orbitinga metal-rich K dwarf star, the first planet discovered by the Alsubai Project exoplanettransit survey. We describe the strategy used to select candidate transiting planetsfrom photometry generated by the Alsubai Project instrument. We examine the rateof astrophysical and other false positives found during the spectroscopic reconnaissanceof the initial batch of candidates. A simultaneous fit to the follow-up radial velocitiesand photometry of Qatar-1b yield a planetary mass of 1.09 ± 0.08MJ and a radius of1.16±0.05RJ. The orbital period and separation are 1.420033 days and 0.0234 AU foran orbit assumed to be circular. The stellar density, effective temperature and rotationrate indicate an age greater than 4 Gyr for the system.
Key words: Planetary systems – stars: individual: Qatar-1 – techniques: photometry– techniques: spectroscopy – techniques: radial velocities
1 INTRODUCTION
Transiting extrasolar planets are important because mea-surements of the planetary transits as well as the stellar re-flex velocity provide both the mass and radius, and hence thedensity of the planet. In contrast to the relatively tight mass-radius relationship of main-sequence stars, the hot Jupitersfound in transit surveys exhibit a wide range of radii at eachmass, thus additional parameters affect the radii of close-ingas giants. With over 100 transiting extrasolar planets nowsecurely characterised1 statistics are beginning to supportcomparative studies to unravel the factors that determine
? E-mail:[email protected] see http://exoplanet.eu
their radii and orbit parameters (Mordasini et al. 2009). To-ward this goal it is important to extend the statistics of hotJupiters to smaller planets.
The Kepler satellite has delivered over 700 transitingobjects, providing good statistics on the orbital periods andradii of transiting bodies (Borucki et al. 2010). At typicaldistances 300-1000 pc, follow-up spectroscopic studies forhigh-precision radial velocities to measure the stellar wob-ble and hence the masses of the smaller transiting bodiespresents a considerable challenge for present-generation in-struments. The wide-angle ground-based exoplanet transitsurveys, in contrast to Kepler, survey brighter stars typicallywithin 100 pc. For these relatively nearby stars, follow-upobservations are considerably easier.
c© 2010 RAS
arX
iv:1
012.
3027
v2 [
astr
o-ph
.EP]
23
Dec
201
0
2 K. Alsubai et al.
The Alsubai Project2 (Alsubai et al. 2011) has initi-ated a wide-field transit search programme deploying ini-tially a 5-camera CCD imaging system designed to go deeperthan most current wide-angle survey systems such as Super-WASP3, HATNet4, TrES5 and XO6. The Alsubai Project’sfirst site in New Mexico was chosen to complement the Su-perWASP sites in the Canary Islands and South Africa,where suites of eight 200mm f/2.0 Canon lenses image16◦×32◦ degree fields at 15 arcsec pix−1. Combining Alsubaiand SuperWASP data should increase the exoplanet discov-ery rate by enabling multiple transits of candidate systemsto be obtained more quickly. The Alsubai camera’s higherangular resolution and larger aperture should also extendthe search for transiting planets to fainter stars with smallerradii, and hence to the discovery of smaller transiting plan-ets.
This paper reports the discovery and initial characteri-sation of the first confirmed transiting exoplanet to emergefrom the Alsubai Project exoplanet transit survey, orbitingthe star 3UC311-087990 (Qatar-1, α2000 = 20h13m31s.61,δ2000 = +65◦09′43′′.4). The survey observations and candi-date selection procedures that led to the discovery of Qatar-1b are presented in Section 2, together with our analysis ofthe stellar spectrum and follow-up photometry. We discussin Section 3 our determination of the stellar and planetarysystem parameters, leading to the summary and conclusionsin Section 4.
2 OBSERVATIONS
The Alsubai camera system images an 11◦ × 11◦ field ofview simultaneously at two pixel scales. A single 200mmf/2.0 Canon lens covers the full 11◦ × 11◦ degree field at9.26 arcsec pix−1. Four 400mm f/2.8 Canon lenses each cover5.5◦ × 5.5◦ fields arranged in a 2 × 2 mosaic to cover thesame 11◦ × 11◦ field. All cameras use FL1 ProLine PL6801KAF-1680E 4K×4K CCD detectors. The robotic mount cy-cles through four pointings, taking 100s exposures with the400mm cameras and 60s exposures with the 200mm camera,thereby covering a ∼ 400 square degree field with a cadenceof 8 minutes. The 400mm lenses target stars between V = 11and 15 mag with 100-s exposures, while the 200mm lens cov-ers the magnitude range from V = 8 to 12 mag with 60-sexposures. We thus obtain photometry of all stars in thefield in the range from V = 8 to 15 mag.
The data are reduced at the University of St Andrewsusing pipeline software based on the image-subtraction al-gorithm of Bramich (2008). A detailed description of thepipeline is given by Alsubai et al. (2011). The pipeline dataproducts are ingested into a data archive at the Universityof Leicester, which uses the same architecture as the WASParchive (Pollacco et al. 2006).
2 http://www.alsubaiproject.org/default.aspx3 see http://www.superwasp.org/4 http://www.hatnet.hu/5 http://dl.dropbox.com/u/502281/Sites/solas/tres/tres.html6 http://www-int.stsci.edu/∼pmcc/xo/
Figure 1. Alsubai discovery light curve of Qatar-1b. The data
are phase folded using the ephemeris HJD = 2455518.4102 +
1.420033E. The solid line represents the best fitting model tran-sit light curve derived from these observations and the follow-up
radial-velocities and photometry.
2.1 Discovery photometry
An automated transit search was conducted on the archivedata using the box least-squares (BLS) algorithm of Kovacs,Zucker, & Mazeh (2002) as modified for the SuperWASPproject by Collier Cameron et al. (2006). Systematic pat-terns of correlated noise were modelled and removed fromthe archive light curves using a combination of the Sys-Rem algorithm of Tamuz, Mazeh, & Zucker (2005) and theTrend Filtering Algorithm (TFA) of Kovacs, Bakos, & Noyes(2005). The light curves of all candidates were subjected tothe candidate screening tests described by Collier Cameronet al. (2007) to ensure that the depths and durations of thetransits were consistent with expectation for objects of plan-etary dimensions transiting main-sequence stars. In caseswhere the same star had been observed by the SuperWASPsurvey, we ran periodogram tests to seek evidence of a tran-sit signal in the SuperWASP data at the same period.
The star 3UC311-087990 was found to exhibit transit-like events at 1.42-day intervals in three fields of the AlsubaiProject instrument. Field 073723+532023 was observed 2239times from 2010 Jun 16 to Sep 7, while fields 074307+520936and 074440+552202 were both observed on 2916 occasionsbetween 2010 Jun 16 and Sep 24. These fields are to thenorth of the declination limit of the SuperWASP survey,so no SuperWASP light curve was available. In all threefields, 3UC311-087990 exhibited a clear transit signal withsignal detection efficiencies (as defined by Kovacs, Zucker,& Mazeh 2002) SDE = 11, 22 and 16 respectively forthe three fields. The corresponding signal-to-red noise ra-tios were Sred = 11, 12 and 13 using the definition of CollierCameron et al. (2006).
The transit duration and J − H colour of 3UC311-087990 were found to be consistent with the radius and massof a main-sequence K dwarf host. For such a star, the 0.02-mag transit depth suggests a companion radius close to thatof Jupiter.
c© 2010 RAS, MNRAS 000, 1–8
Qatar-1b: a hot Jupiter orbiting a metal-rich K dwarf star 3
Table
1.
Als
ub
ai
tran
siti
ng
pla
net
can
did
ate
s.T
he
colu
mn
sare
lab
elle
das
follow
s:
colu
mn
1:
UC
AC
3id
enti
fica
tion
for
the
targ
et,
colu
mn
s2
an
d3:
J2000
Rig
ht
Asc
ensi
on
an
dD
eclin
ati
on
mea
sure
dby
this
pro
ject
,
colu
mn
4:
Vm
agn
itu
de
from
Als
ub
ai
pip
elin
eca
lib
rati
on
again
stU
CA
C3
magn
itu
de,
colu
mn
s5
an
d6:
Ph
oto
met
ric
per
iod
an
dep
och
,co
lum
ns
7to
9:
Eff
ecti
ve
tem
per
atu
re,
log
surf
ace
gra
vit
y,an
dro
tati
on
al
lin
eb
road
enin
gof
the
synth
etic
tem
pla
tesp
ectr
um
that
gave
the
bes
tm
atc
hto
the
ob
serv
edsp
ectr
a,
ass
um
ing
sola
rm
etallic
ity,
colu
mn
s10
an
d11:
Ab
solu
tera
dia
lvel
oci
tyon
the
IAU
syst
eman
drm
svari
ati
on
wh
entw
oor
more
mea
sure
men
tsw
ere
ob
tain
ed,
colu
mn
s12
an
d13:
Nu
mb
erof
TR
ES
ob
serv
ati
on
san
dto
tal
exp
osu
reti
me,
colu
mn
14:
Det
ecti
on
sum
mary
:An
mea
ns
the
tran
sits
wer
ed
etec
ted
byn
of
the
Als
ub
ai
cam
eras,
W1
mea
ns
the
tran
sits
wer
eals
od
etec
ted
wit
hth
esa
me
per
iod
by
Su
per
WA
SP
,W
0m
ean
sth
at
Su
per
WA
SP
did
not
con
firm
the
det
ecti
on
,an
dn
oW
mea
ns
the
targ
etfi
eld
was
not
obse
rved
by
Su
per
WA
SP
,
colu
mn
15:
Dis
posi
tion
of
the
can
did
ate
:
D=
the
spec
tru
msh
ow
stw
ose
tsof
lin
es;FA
?=
alikel
yp
hoto
met
ric
fals
eala
rm;F
R=
ara
pid
lyro
tati
ng
star
for
wh
ich
pre
cise
rad
ialvel
oci
ties
are
not
feasi
ble
;G
=a
gia
nt,
pre
sum
ab
lyin
ab
len
ded
syst
emw
ith
an
eclip
sin
gb
inary
;G
?=
alikel
ygia
nt;
H=
ah
ot
star;
P?
=th
evel
oci
tyvari
ati
on
sare
small
an
dn
ot
inco
nsi
sten
tw
ith
ap
lanet
ary
com
pan
ion
;P
!=
a
con
firm
edp
lan
etary
com
pan
ion
;S
=th
esp
ectr
um
issi
ngle
-lin
edan
dsh
ow
sa
larg
evel
oci
tyvari
ati
on
du
eto
ast
ella
rco
mp
an
ion
;T
?=
alikel
ytr
iple
syst
em.
Can
did
ate
α2000
δ 2000
VP
erio
dE
poch
Teff
log(g
)V
rot
Vra
dσV
Nobs
Exp
TF
ield
Typ
e
mag
days
HJD
Kkm
s−1
km
s−1
km
s−1
min
215-0
18836
04:2
0:2
9.3
017:0
3:2
9.0
11.7
2.6
539
2455182.0
3187
5000
3.0
06
32.0
50
...
130
A1
G210-0
17759
04:2
3:1
8.2
214:5
8:1
9.2
14.2
1.6
343
2455170.0
2498
7750
5.0
06
0.7
37
0.8
60
2105
A1W
0H
214-0
19086
04:2
5:1
0.2
616:4
9:5
9.9
14.8
2.1
799
2455186.8
9730
6250
4.5
060
−5.8
23
...
180
A1W
1F
R
214-0
19418
04:2
7:5
8.8
616:5
5:2
1.8
13.3
1.2
068
2455510.8
3400
6125
3.7
57
−41.8
41
0.0
75
270
A1W
0FA
?214-0
19710
04:3
0:2
2.4
416:5
4:2
7.1
14.0
2.3
042
2455180.0
3781
6250
5.0
025
−5.8
79
...
145
A1W
0..
.
216-0
21225
04:3
5:0
4.3
017:3
5:2
7.9
13.3
1.2
131
2455154.5
8890
5875
3.7
57
13.8
23
0.0
78
254
A1W
0FA
?212-0
19540
04:3
5:1
5.2
015:3
0:5
7.9
12.6
5.2
482
2455155.2
4280
6125
3.5
045
−1.8
58
1.2
81
280
A1
FR
197-0
17651
04:3
8:5
8.3
608:1
3:1
5.0
13.3
1.3
681
2455511.6
0200
6750
4.0
09
86.0
48
0.0
61
376
A1W
0FA
?183-0
15913
04:4
1:5
7.4
901:2
4:4
8.7
12.5
1.5
418
2455178.8
6996
6250
3.5
020
64.4
20
...
124
A1
D210-0
19969
04:4
5:4
1.0
714:4
3:2
1.6
13.1
1.9
023
2455154.8
6690
5250
4.5
025
26.3
03
...
124
A1W
0D
217-0
22884
04:4
7:3
1.6
418:2
9:0
8.8
13.1
2.1
060
2455155.6
2150
6875
4.0
028
−6.5
08
0.4
34
247
A1W
0H
184-0
18774
05:0
3:2
2.9
301:4
5:0
7.2
13.4
1.8
672
2455154.9
5550
6750
3.5
020
−22.1
10
...
130
A2W
1D
194-0
20871
05:0
3:5
9.8
406:4
8:3
6.8
13.2
2.3
520
2455184.5
9328
6000
5.0
016
−19.1
36
...
130
A1W
1D
197-0
23150
05:0
9:1
4.3
508:1
6:5
2.8
12.2
4.2
357
2455156.6
3930
6000
4.0
08
−8.2
43
0.1
69
260
A1W
0FA
?199-0
22618
05:0
9:3
5.2
909:1
7:4
3.2
14.1
1.7
068
2455182.4
1670
6500
4.0
030
40.0
73
54.2
18
284
A1W
1S
196-0
23727
05:1
6:0
2.9
907:3
7:0
3.4
13.7
1.8
731
2455154.9
3060
4750
4.5
01
−11.4
94
0.1
18
3150
A1W
0FA
?205-0
23986
05:2
5:0
6.4
712:0
4:3
3.7
11.3
5.2
372
2455175.6
1072
5000
3.0
04
27.9
00
...
115
A1
G190-0
24393
05:2
8:1
3.2
504:4
2:2
4.4
13.3
1.3
150
2455180.8
8287
5417
4.3
33
29.7
85
0.2
67
6289
A1W
0FA
?210-0
28815
05:3
0:4
1.8
014:5
4:4
9.0
12.9
2.8
459
2455155.2
3150
6000
3.7
59
28.8
85
0.1
21
287
A1
P?
308-1
05347
18:5
1:2
1.1
963:5
9:0
2.5
12.8
1.5
758
2455365.0
0960
6500
3.5
045
−113.2
03
...
130
A2W
1D
284-1
37549
18:5
9:3
0.3
951:3
4:5
0.0
12.4
2.4
046
2455365.5
3430
6000
3.0
045
19.9
63
...
124
A1W
1T
?304-1
11233
19:2
2:5
2.6
561:5
3:4
8.6
13.3
2.0
262
2455510.6
1000
5250
4.5
0110
−14.7
07
...
130
A2W
0F
R309-1
00007
19:2
5:1
5.3
464:0
0:1
7.9
13.2
2.1
804
2455407.4
4653
5625
4.3
83
−36.8
59
0.1
61
4175
A1W
0FA
?
301-1
28216
19:3
2:0
7.3
360:2
7:0
7.0
13.7
1.8
227
2455364.6
8300
5250
4.0
02
1.5
29
0.1
04
274
A2W
0P
?301-1
28743
19:3
5:4
2.5
660:2
7:0
5.4
12.5
1.0
851
2455413.4
0090
6083
4.0
06
−28.5
14
13.8
03
372
A2W
1T
?294-1
43877
19:4
5:1
1.2
556:3
5:1
6.3
12.2
1.3
284
2455424.5
3773
6250
4.5
04
−95.6
71
...
130
A2W
1D
311-0
87990
20:1
3:3
1.6
165:0
9:4
3.4
12.6
1.4
201
2455407.6
4900
4861
4.4
42
−37.9
00
0.1
89
9472
A3
P!
294-1
52358
20:2
0:2
2.7
456:3
2:2
5.0
13.0
2.3
267
2455365.3
2740
5000
3.5
03
−72.1
59
0.1
26
260
A1W
0G
?
c© 2010 RAS, MNRAS 000, 1–8
4 K. Alsubai et al.
2.2 Spectroscopic reconnaissance
A list of 28 candidates was provided by the Alsubai Projectto the CfA team in November 2010. The usual first stepat the Harvard-Smithsonian Center for Astrophysics (CfA)for vetting transiting-planet candidates from wide-angleground-based photometric surveys is to obtain reconnais-sance spectra with the Tillinghast Reflector Echelle Spectro-graph (TRES) on the 1.5-m Tillinghast Reflector at the FredL. Whipple Observatory operated by the Smithsonian Astro-physical Observatory (SAO) on Mount Hopkins in SouthernArizona. These spectra are used to look for evidence of stel-lar systems that are the source of the transit-like light curves(e.g. see Latham et al. 2009) and also to provide refinedstellar parameters for the targets. An initial spectroscopicreconnaissance was carried out for all 28 candidates, basedon 60 spectroscopic observations with TRES over a span of17 nights, using the medium fiber at a spectral resolutionR = 44000. A summary of the candidates and the results ofthe reconnaissance are reported in Table 1.
For the initial analysis, the observed spectra are cor-related against a grid of synthetic spectra drawn from alibrary calculated by John Laird using Kurucz models (Ku-rucz 1992) and a line list prepared by Jon Morse. The syn-thetic spectra cover a window of 300A centered near thegravity-sensitive Mg b features. The template that gives thebest match to the observed spectrum is then used to derivethe rotational and radial velocities. To establish the zeropoint of the radial velocities on the IAU system, a smallcorrection is derived using nightly observations of standardstars.
The initial analysis of the TRES spectra against the li-brary of synthetic spectra provides useful information aboutthe characteristics of the target star, such as effective tem-perature, surface gravity, and rotational and absolute radialvelocity, but it only uses a small fraction of the full 390to 900 nm spectral range of TRES. To look for evidence oflow-amplitude orbital motion, we take advantage of the widewavelength coverage by correlating the individual observa-tions of a star against a template derived from observationsof the same star, either a single observation that has espe-cially strong signal-to-noise, or a master observed templateconstructed by shifting and co-adding all the observationsof the star.
The radial velocities reported in Table 1 are calibratedusing an absolute velocity zero point based on observationsof the IAU radial-velocity standard star HD 182488. TheKepler team has agreed to use this as the standard for thereconnaissance spectroscopy of Kepler Objects of Interest.We have adopted the velocity of −21.508 km s−1 as the valueon the IAU system for HD 182488. During the November runwe accumulated 15 strong observations of HD 182488, givingan observed mean velocity on the TRES native system of−20.807±0.057 km s−1, where the error is the rms residualsfrom the mean. This is an offset of +0.701 km s−1 from theadopted IAU velocity for HD 182488. The radial velocitiesreported in Table 1 have had this offset applied, to bringthem as closely as possible into line with the radial velocityof HD 182488 on the IAU system.
When many candidates are being vetted, it is efficientto schedule the initial observations near times of quadrature,as predicted from the photometric ephemerides. In the case
Table 2. Relative radial velocities for Qatar-1.
BJD Phase Radial velocity Bisector span(d) (km s−1 ) (km s−1)
2455518.69628 0.2014 −0.1297 ± 0.0605 0.0068 ± 0.0443
2455519.59174 0.8320 0.3515 ± 0.0292 −0.0227 ± 0.0131
2455520.63231 0.5648 0.2411 ± 0.0462 0.0164 ± 0.02962455521.58409 0.2350 −0.0663 ± 0.0407 0.0127 ± 0.0166
2455523.60908 0.6610 0.2611 ± 0.0261 −0.0045 ± 0.0143
2455525.59298 0.0581 0.0 ± 0.0221 −0.0065 ± 0.00882455526.58711 0.7582 0.3618 ± 0.0458 0.0037 ± 0.0219
2455527.59121 0.4653 0.0854 ± 0.0221 −0.0179 ± 0.0128
2455528.58427 0.1646 −0.0403 ± 0.0305 0.0120 ± 0.0117
of double-lined binaries that are undergoing grazing eclipses,this guarantees that the velocities of the two stars will benear their maximum separation, and thus a single spectrumwill be sufficient for rejecting the target if it shows a compos-ite spectrum. If the first observation reveals the spectrum ofjust one star, a second observation near the opposite quadra-ture is optimum for disclosing orbital motion due to unseenstellar companions. Of course, this assumes that the photo-metric ephemeris has the correct period and is up to date.
Nine of the candidates showed clear evidence of stellarcompanions. In eight cases the initial reconnaissance obser-vation revealed a composite spectrum with at least two setsof lines, with evidence that three stars were likely to be in-volved for two of the targets. For the ninth case the spectrumwas single-lined, but with a very large change in velocity be-tween the two observations. Three of the candidates showedbroad lines corresponding to equatorial rotational velocitiesof tens of km s−1, which would render impractical the mea-surement of very precise radial velocities, and two of thecandidates proved to have temperatures corresponding tolate A stars. In three cases the classification of the TRESspectra indicated that the targets are giants, presumablybright stars diluting the light of eclipsing binaries, either inhierarchical triple systems or accidental alignments. Seven ofthe candidates showed no significant velocity variations, butcareful inspection of the available light curves supported theinterpretation that the transit detections were false alarms.Two of the candidates with secure transit detections showedvery small velocity variations that could be consistent withorbital motion due to planetary companions, and one can-didate that was observed only once showed a spectrum suit-able for very precise velocity measurements. These three tar-gets deserve additional follow up. Finally, one of the candi-dates (3UC311-087990, hereinafter referred to as Qatar-1)was confirmed as a system with a transiting planet, as de-scribed in the following sections.
2.3 Radial-velocity follow-up
The first two radial velocities of Qatar-1, obtained near op-posite quadratures, showed a small but significant differenceconsistent with the photometric ephemeris and with the in-terpretation of a planetary mass for the companion. Subse-quently this star was observed every clear night with a longer
c© 2010 RAS, MNRAS 000, 1–8
Qatar-1b: a hot Jupiter orbiting a metal-rich K dwarf star 5
Figure 2. Upper panel: Relative radial-velocity data for Qatar-
1, with 1 σ error bars, phase folded on the ephemeris given inTable 3. The best fitting circular orbit model (solid line) is also
shown. Lower panel: Residuals of the fit to the best fitting circular
orbit model.
exposure time of 54 minutes, with the goal of deriving an or-bital solution. The multi-order relative radial velocities fromall nine observations of Qatar-1 are reported in Table 2 andare plotted in Fig. 2 together with a circular orbit, phased tothe period and epoch of the photometric ephemeris. No sig-nificant correlation was found between the variation in theline bisectors and the relative radial velocities, as shown inFig. 3. Thus there is no evidence in the TRES spectra thatthe radial-velocity variations are due to phenomena otherthan orbital motion, such as an unresolved blend with afaint eclipsing binary or spots coupled with stellar rotation(Queloz et al. 2001).
2.4 Spectroscopic parameters of Qatar-1
As mentioned in the Introduction, a transiting planet allowsus to determine both the radius and the mass of the planet,if a spectroscopic orbit for the host star is available to com-plement the transit light curve. This in turn provides keyinformation about the bulk properties of the planet, such asdensity. However, the planetary mass and radius values arerelative to the mass and radius of the host star, and the accu-racy with which the planetary properties can be determinedis often limited by the uncertainties in the characteristicsof the host star. Nearby stars with accurate parallaxes havethe advantage that the observed luminosity of the star helpspin down key stellar parameters such as radius and effectivetemperature. For more distant stars, such as Qatar-1, thealternative is to use stellar models together with values forthe effective temperature and metallicity derived from thespectra.
We have used our library of synthetic spectra and acorrelation analysis of our TRES spectra similar to that de-scribed by Torres, Neuhauser, & Guenther (2002), to de-rive the following results for Qatar-1: effective temperatureTeff? = 4861±125 K, surface gravity log g? = 4.40±0.1 (log
Figure 3. Line bisector span versus radial velocity for Qatar-1.
Table 3. Stellar parameters for Qatar-1 derived from spectro-
scopic reconnaissance, photometric catalogues and model fitting.
Spectroscopic parameter Value Source
Teff? (K) 4861 ± 125 K TRES
[Fe/H] 0.20 ± 0.10 TRESv sin I (km s−1) 2.1 ± 0.8 TRES
γRV (km s−1) −37.835 ± 0.063 TRES
Photometric parameter Value SourceV (mag) 12.843 ± 0.137 TASS4
J (mag) 10.999 ± 0.021 2MASS
H (mag) 10.527 ± 0.019 2MASSKs (mag) 10.409 ± 0.017 2MASS
Model parameter Value SourceM? (M�) 0.85 ± 0.03 MCMC
R? (R�) 0.823 ± 0.025 MCMC
ρ? (ρ�) 1.52 ± 0.12 MCMClog g (cgs) 4.536 ± 0.024 MCMCAge (Gyr) > 6 MCMC+YY
cgs), projected rotational velocity7 v sin I = 2.1 ± 0.8 kms−1 and metallicity [Fe/H] = +0.20±0.1 dex. Spectroscopicdeterminations of stellar surface gravity are notoriously dif-ficult, so it is fortunate that log g? can be determined in-dependently from a joint analysis of the transit light curveand spectroscopic orbit. For Qatar-1 that analysis yieldedlog g? = 4.53 ± 0.02, as described in the next section.
The stellar parameters adopted for the host star arelisted in Table 3 together with catalogue magnitudes andadditional stellar dimensions obtained by fitting a model tothe transit profiles and spectroscopic orbit.
7 The symbol I represents the inclination of the stellar rotation
axis to the line of sight, whereas i is used elsewhere in this paper
to denote the orbital inclination.
c© 2010 RAS, MNRAS 000, 1–8
6 K. Alsubai et al.
2.5 Follow-up photometry
A full transit of the planet Qatar-1b was observable fromhigh northern latitudes on the night of 2010 Nov 27. R-band photometry was obtained of the ingress with the CCDcamera on the 0.95m James Gregory Telescope (JGT) lo-cated at St Andrews, Scotland, during the transit of 2010Nov 27. A total of 25 180-s exposures was obtained in clearconditions, but the sequence was terminated early by snowclouds. The egress of the same Nov 27 transit was observedin clear conditions with the 60-cm telescope and CCD cam-era of the University of Keele. A sequence of 535 20-s R-bandmeasurements was obtained.
Using the refined ephemeris from these observations weidentified an opportunity to observe a complete transit usingthe KeplerCam CCD on the FLWO 1.2m telescope on theevening of 2010 Dec 2. The transit was observed using 90-s exposures in the Sloan i filter; the resulting high-qualityphotometry was decorrelated against external parametersas described by Bakos et al. (2010). A third transit wasobserved in its entirety on 2010 Dec 7, again using the JGTwith an R-band filter.
The four follow-up light curves are shown, together withthe best-fitting model, in Fig. 4. The transit is seen tobe slightly more than 0.02 magnitudes deep in both wave-length bands. The i-band light curve in particular shows fourwell-defined contacts but rather lengthy ingress and egressphases, suggesting a moderately high impact parameter.
3 STELLAR AND PLANETARY DIMENSIONS
The dimensions of the planet and its host star were deter-mined from a simultaneous model fit to the radial veloci-ties and the combined photometry from the Alsubai Projectcameras and follow-up transit observations. The transit lightcurve was modelled using the formulation of Mandel & Agol(2002) in the small-planet approximation. A 4-coefficientnonlinear limb darkening model was used, employing fixedcoefficients appropriate to the R band for the Alsubai, JGTand Keele photometry, and to the Sloan i band for the Ke-plerCam photometry. These were interpolated to the appro-priate effective temperature and metallicity from the tabu-lation of Claret (2004).
The parameter optimization was performed using thecurrent version of the Markov-Chain Monte Carlo (MCMC)code described by Collier Cameron et al. (2007) and Pollaccoet al. (2008). The transit light curve is modelled in terms ofthe epoch T0 of mid-transit, the orbital period P , the ratioof radii d = (Rp/R?)2, the approximate duration tT of thetransit from initial to final contact and the impact param-eter b = a cos i/R?. The radial-velocity orbit is defined bythe stellar orbital velocity semi-amplitude K? and the offset∆γ of the centre-of-mass velocity from the zero-point of therelative velocities listed in Table 2. Where the eccentricity isallowed to float, the two additional fitting parameters e cosωand e sinω are introduced, as recommended by Ford (2006).
The linear scale of the system depends on the orbitalseparation a, which through Kepler’s third law depends onthe stellar mass M?. The stellar mass was estimated ateach step in the Markov chain as a function of the effec-tive temperature, metallicity and density of the star (Enoch
Figure 4. Photometric follow up light curves, offset from each
other by an arbitrary amount for clarity. From top to bottom:R-band photometry of the planetary egress obtained using the
Keele 60-cm telescope, 2010 Nov 27; R-band photometry of the
Nov 27 planetary ingress obtained using the 0.95m JGT; R-bandJGT photometry of the Dec 7 transit; Sloan i-band photometry
of the Dec 2 planetary transit obtained using KeplerCam on the
1.2-m telescope at FLWO. All data have been phase folded on theephemeris given in Table 3. The best fitting model transit light
curve is overplotted in all three cases.
et al. 2010). The effective temperature and metallicity weretreated as additional MCMC model parameters, constrainedby Gaussian priors with mean values and variances deriveddirectly from the stellar spectra, as listed in Table 3.
A model fit for an eccentric orbit yields an orbitaleccentricity e = 0.23 ± 0.11. The uncertainty in the ec-centricity and the orientation of the orbit yields a morehighly inflated and uncertain value for the stellar radius,R? = 1.04±0.11R�. The planet’s radius and density are sim-ilarly affected. The stellar mass increases to 0.87± 0.03M�.A star of this mass would have to be among the oldest inthe galactic disk population to have evolved to such a largeradius. A more likely explanation is that the best-fit valueof the eccentricity is spurious, and that the orbit is close tocircular. The improvement in the fit resulting from the addi-tion of e cosω and e sinω as fitting parameters is insufficientto justify adoption of anything other than a circular orbit.The F -test approach of Lucy & Sweeney (1971) indicatesthat there is a 16.8% probability that the improvement inthe fit could have arisen by chance if the underlying orbitwere circular. In the absence of conclusive evidence to thecontrary, we adopted the circular orbit model.
The orbital and planetary parameters derived from theMCMC model fit are summarized in Table 4.
c© 2010 RAS, MNRAS 000, 1–8
Qatar-1b: a hot Jupiter orbiting a metal-rich K dwarf star 7
Table 4. System parameters and 1σ error limits derived from the MCMC analysis. Although for reasons given in the text we adopt the
circular orbit solution, we include the eccentric solution here to show its influence on other system parameters.
Parameter Symbol Circular Eccentric Units
Transit epoch T0 5518.4102 ± 0.0002 5518.4103 ± 0.0003 days
Orbital period P 1.420033 ± 0.000016 1.420033 ± 0.000015 days
Planet/star area ratio (Rp/R∗)2 0.02117 ± 0.00045 0.02114 ± 0.00045Transit duration tT 0.06716 ± 0.00077 0.06632 ± 0.00095 days
Impact parameter b 0.696+0.021−0.024 0.695+0.020
−0.024 R∗Stellar reflex velocity K1 0.218+0.015
−0.016 0.220+0.015−0.016 km s−1
Centre-of-mass velocity offset ∆γ 0.118794+0.000052−0.000053 0.1336+0.0081
−0.0080 km s−1
Orbital eccentricity e 0.0 (fixed) 0.24+0.10−0.12
Longitude of periastron ω . . . 84.6+11.9−10.1 degrees
Orbital inclination i 83.47+0.40−0.36 79.4+2.3
−2.9 degrees
Orbital semi-major axis a 0.02343+0.00026−0.00025 0.02363+0.00030
−0.00029 AU
Planet radius Rp 1.164 ± 0.045 1.47 ± 0.16 RJ
Planet mass Mp 1.090+0.084−0.081 1.132+0.096
−0.087 MJ
Planet surface gravity log gp 3.265+0.044−0.045 3.078+0.091
−0.076 [cgs]
Planet density ρp 0.690+0.098−0.084 0.355+0.139
−0.086 ρJ
Planet temperature Teq 1399 ± 42 1564 ± 94 K
4 DISCUSSION AND CONCLUSIONS
The spectroscopic analysis of the host star Qatar-1 indicatesthat it is a slowly-rotating, slightly metal-rich dwarf star ofspectral type K3V. The MCMC analysis of the transit du-ration and impact parameter yields a direct estimate of thestellar density. When compared with evolutionary tracks andisochrones for [Fe/H]=0.2 in the (ρ?/ρ�)1/3 versus Teff plane(Fig. 5), the spectroscopically-measured effective tempera-ture indicates a mass between 0.76 and 0.87 M�, in goodagreement with the MCMC estimate using the calibrationof Enoch et al. (2010). The stellar density derived from theMCMC analysis is substantially lower than would be ex-pected for a star of this age on the zero-age main sequence,giving a lower limit on the stellar age of 4 Gyr. The slow stel-lar rotation rate derived from the spectra is also consistentwith a spin-down age > 4 Gyr.
The planet Qatar-1b is 10% more massive than Jupiter,and has a radius 16% greater than Jupiter’s. It orbits its pri-mary every 34 hours, making it one of the shortest-periodplanets yet found orbiting a star less massive than theSun. The blackbody equilibrium temperature given in Ta-ble 4 is calculated assuming a planetary albedo of zero andisotropic re-radiation of the power received from the hoststar. A more general estimate of the dayside temperatureis T 4
eql = T 4? (R?/2a)2((1 − A)/F ), where A is the planet’s
Bond albedo and F is the fraction of the stellar surfacethat re-radiates at Teql. Measurements of starlight reflectedat optical wavelengths from the dayside hemispheres of thehot Jupiters HD 209458b (Rowe et al. 2008) and CoRoT-2b(Alonso et al. 2009) indicate that their albedos are typicallyof order a few percent. The blackbody value of 1399 K thusconstitutes a reasonable lower limit on the planet’s daysidetemperature.
Qatar-1b was one of the very first batch of transit can-didates from the Alsubai Project to be subjected to spec-troscopic reconnaissance and radial-velocity follow-up. The
Figure 5. The position of Qatar-1 in the (ρ∗/ρ�)−1/3 planecompared to theoretical evolutionary tracks and isochrones inter-polated from Yi, Kim, & Demarque (2003) to [Fe/H]=0.2. Theisochrones displayed represent 2.0, 4.0, 6.0, 8.0, 10.0, 12.0, 14.0,
16.0, 18.0 and 20.0 Gyr. The evolutionary tracks represent 0.7,0.8, 0.9 and 1.0 M�.
rapidity of the discovery, and the fact that the planet orbitsa mid-K dwarf, confirm that the instrument is well-suited tothe efficient discovery of planets around lower main-sequencestars.
REFERENCES
Alonso R., Guillot T., Mazeh T., Aigrain S., Alapini A.,Barge P., Hatzes A., Pont F., 2009, A&A, 501, L23
Alsubai K., et al., 2011, PASP, submitted.
c© 2010 RAS, MNRAS 000, 1–8
8 K. Alsubai et al.
Bakos G. A., et al., 2010, ApJ, 710, 1724Borucki W. J., for the Kepler Team, 2010, arXiv,arXiv:1006.2799
Bramich D. M., 2008, MNRAS, 386, L77Claret A., 2004, A&A, 428, 1001Collier Cameron A., et al., 2006, MNRAS, 373, 799Collier Cameron A., et al., 2007, MNRAS, 380, 1230Enoch B., Collier Cameron A., Parley N. R., Hebb L., 2010,A&A, 516, A33
Ford E. B., 2006, ApJ, 642, 505Kovacs G., Zucker S., Mazeh T., 2002, A&A, 391, 369Kovacs G., Bakos G., Noyes R. W., 2005, MNRAS, 356,557
Kurucz, R. L. 1992, in The Stellar Populations of Galax-ies, IAU Symp. No. 149, ed. B. Barbuy and A. Renzini(Kluwer Acad. Publ.: Dordrecht), 225
Latham D. W., et al., 2009, ApJ, 704, 1107Lucy L. B., Sweeney M. A., 1971, AJ, 76, 544Mandel K., Agol E., 2002, ApJ, 580, L171Mordasini C., Alibert Y., Benz W., Naef D., 2009, A&A,501, 1161
Pollacco D. L., et al., 2006, PASP, 118, 1407Pollacco D., et al., 2008, MNRAS, 385, 1576Queloz D., et al., 2001, A&A, 379, 279Rowe J. F., et al., 2008, ApJ, 689, 1345Tamuz O., Mazeh T., Zucker S., 2005, MNRAS, 356, 1466Torres G., Neuhauser R., Guenther E. W., 2002, AJ, 123,1701
Yi S. K., Kim Y.-C., Demarque P., 2003, ApJS, 144, 259
This paper has been typeset from a TEX/ LATEX file preparedby the author.
c© 2010 RAS, MNRAS 000, 1–8