arXiv:0806.1734v1 [astro-ph] 10 Jun 2008
Measurement of the Spin-Orbit Angle of Exoplanet HAT-P-1b1
John Asher Johnson2,3, Joshua N. Winn4, Norio Narita5, Keigo Enya6, Peter K. G.
Williams2, Geoffrey W. Marcy2, Bun’ei Sato7, Yasuhiro Ohta8, Atsushi Taruya8, Yasushi
Suto8, Edwin L. Turner9, Gaspar Bakos10, R. Paul Butler11, Steven S. Vogt12, Wako Aoki5,
Motohide Tamura5, Toru Yamada13, Yuzuru Yoshii14, Marton Hidas15
1Based on observations obtained at the Keck Observatory, which is operated as a scientific partnership
among the California Institute of Technology, the University of California, and the National Aeronautics and
Space Administration; the Subaru Telescope, which is operated by the National Astronomical Observatory
of Japan; and the Lick Observatory, which is operated by the University of California.
2Department of Astronomy, University of California, Mail Code 3411, Berkeley, CA 94720
3Current Address: Institute for Astronomy, University of Hawaii, Honolulu, HI 96822; NSF Postdoctoral
4Department of Physics, and Kavli Institute for Astrophysics and Space Research, Massachusetts Institute
of Technology, Cambridge, MA 02139
5National Astronomical Observatory of Japan, 2-21-1 Osawa, Mitaka, Tokyo 181–8588, Japan
6Department of Infrared Astrophysics, Institute of Space and Astronautical Science, Japan Aerospace
Exploration Agency, 3-1-1 Yoshinodai, Sagamihara, Kanagawa 229–8510, Japan
7Global Edge Institute, Tokyo Institute of Technology, 2-12-1 Okayama, Meguro, Tokyo 152-8550, Japan
8Department of Physics, School of Science, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo
9Princeton University Observatory, Peyton Hall, Princeton, NJ 08544, USA
10Harvard-Smithsonian Center for Astrophysics, 60 Garden Street, Cambridge, MA 02138; NSF Postdoc-
11Department of Terrestrial Magnetism, Carnegie Institution of Washington DC, 5241 Broad Branch Rd.
NW, Washington DC, 20015-1305
12UCO/Lick Observatory, University of California at Santa Cruz, Santa Cruz, CA 95064
13Astronomical Institute, Tohoku University, Aramaki, Aoba, Sendai, 980-8578, Japan
14Institute of Astronomy, School of Science, The University of Tokyo, 2-21-1 Osawa, Mitaka, Tokyo 181–
15Las Cumbres Observatory, 6740 Cortona Dr. Suite 102, Santa Barbara, CA 93117
– 2 –
We present new spectroscopic and photometric observations of the HAT-P-1
planetary system. Spectra obtained during three transits exhibit the Rossiter-
McLaughlin effect, allowing us to measure the angle between the sky projections
of the stellar spin axis and orbit normal, λ = 3.◦7 ±2.◦1. The small value of λ for
this and other systems suggests that the dominant planet migration mechanism
preserves spin-orbit alignment. Using two new transit light curves, we refine the
transit ephemeris and reduce the uncertainty in the orbital period by an order
of magnitude. We find a upper limit on the orbital eccentricity of 0.067, with
99% confidence, by combining our new radial-velocity measurements with those
Subject headings: techniques: radial velocities—planetary systems: formation—
stars: individual (HAT-P-1, ADS16402A)
Prior to 1995, it was expected that Jovian planets around other stars would inhabit
wide, circular orbits similar to the Solar System gas giants. It was therefore a surprise when
the first exoplanet was discovered with a minimum mass of 0.468 MJupand a semimajor axis
of only 0.05 AU (Mayor & Queloz 1995). Since then, 85 “hot Jupiters”—Jovian planets with
periods ≤10 days—have been detected around Sun-like stars (Butler et al. 2006; Torres et al.
2008). It is unlikely that these planets formed in situ due to the low surface densities and
high temperatures of the inner regions of circumstellar disks (Lin et al. 1996). A more likely
scenario is that these massive planets formed at a distance of several astronomical units, and
then migrated inward to their current locations.
Theories for the inward migration of planets can be divided into two broad categories.
The first category involves tidal interactions between the planet and a remaining gaseous disk
(Lin et al. 1996; Moorhead & Adams 2008). The second category involves few-body grav-
itational dynamics, such as planet–planet scattering (Rasio & Ford 1996; Chatterjee et al.
2008), dynamical relaxation (Papaloizou & Terquem 2001; Adams & Laughlin 2003), and
Kozai cycles accompanied by tidal friction (Holman et al. 1997; Fabrycky & Tremaine 2007;
Wu et al. 2007; Nagasawa et al. 2008). One possible way to distinguish between these cat-
egories is to examine the present-day alignment between the stellar rotation axis and the
planetary orbital axis. Assuming that these axes were initially well aligned, disk-planet tidal
interactions would preserve this close alignment (Ward & Hahn 1994), while the second cat-
egory of theories would at least occasionally result in large misalignments. For example,
Adams & Laughlin (2003) predict a final inclination distribution for dynamically relaxed
– 3 –
planetary systems that peaks near 20◦and extends to 85◦. Likewise, Fabrycky & Tremaine
(2007) and Wu et al. (2007) simulated systems of planets with randomly aligned outer com-
panions and found that the Kozai interaction resulted in a wide distribution of final orbital
inclinations for the inner planet, with retrograde orbits (λ > 90◦) not uncommon. Similar
results were found by Nagasawa et al. (2008), for the case in which Kozai oscillations are
caused by an outer planet, rather than a companion star.
Spin-orbit alignment can be measured by taking advantage of the Rossiter–McLaughlin
(RM) effect that occurs during a planetary transit. As the planet blocks portions of the
rotating stellar surface, the star’s rotational broadening kernel becomes asymmetric and its
spectrum appears to be anomalously Doppler-shifted. The RM effect has previously been
observed and modeled for eight transiting planetary systems (Queloz et al. 2000; Winn et al.
2005, 2006, 2007c; Wolf et al. 2007; Narita et al. 2007a,b; Bouchy et al. 2008; Loeillet et al.
2008; Winn et al. 2008). In this work, we add HAT-P-1 to this sample.
HAT-P-1 (ADS16402B) is a member of a G0V/G0V visual binary and harbors a short–
period, Jovian planet. The transits of HAT-P-1b were discoverd by Bakos et al. (2007) as
part of the Hungarian-made Automated Telescope Network (HATNet). The planet has a
4.465 day orbital period, a mass of 0.53 MJup, a radius RP = 1.20 RJup(Bakos et al. 2007;
Winn et al. 2007a). We have monitored HAT-P-1 using precise radial velocity (RV) and
photometric measurements made both in and out of transit in order to measure the RM
effect and improve the precision with which the system’s orbital parameters are known. In
the following section we describe our observations and data reduction procedures. In §3 we
present the transit model that we fit to our observations, and in § 4 we present our results,
and we conclude in § 5 with a brief discussion.
2. Observations and Data Reduction
2.1. Radial Velocity Measurements
We observed the optical spectrum of HAT-P-1 using the High Resolution Echelle Spec-
trometer (HIRES, Vogt et al. 1994) on the Keck I 10m telescope and the High Dispersion
Spectrograph (HDS, Noguchi et al. 2002) on the Subaru 8m telescope. We set up the HIRES
spectrometer in the same manner that has been used consistently for the California-Carnegie
planet search (Butler et al. 1996; Marcy et al. 2005). This is also the same setup that was
used to gather the 9 Keck/HIRES spectra reported by Bakos et al. (2007). Specifically, we
employed the red cross-disperser and used the I2absorption cell to calibrate the instrumental
response and the wavelength scale. The slit width was set by the 0.′′85 B5 decker, and the
– 4 –
typical exposure times ranged from 3–5 min, giving a resolution of about 60,000 at 5500˚ A
and a signal-to-noise ratio (SNR) of approximately 120 pixel−1. We gathered 3 spectra on
several nights when transits were not occurring, in order to refine the parameters of the
spectroscopic orbit. In addition we gathered a dense time series of spectra on each of two
nights, UT 2007 July 6 and UT 2007 September 2, when transits were predicted to occur. On
each night we attempted to observe the star for many hours bracketing the predicted transit
midpoint, but there were interruptions due to clouds and pointing failures. However, both
nights of data provide good phase coverage of the entire transit event. In total we obtained
79 new Keck/HIRES spectra, of which 49 were observed while a transit was happening.
For our Subaru/HDS spectra we employed the standard I2a setup of the HDS, covering
the wavelength range 4940–6180˚ A with the I2absorption cell. The slit width of 0.′′8 yielded
a spectral resolution of ∼45,000. The typical exposure time was 10 min resulting in a SNR
of 120 pixel−1. Our Subaru observations took place on 3 different nights spread out over 2
months. Two of the nights were not transit nights; we gathered 8 spectra on those nights in
order to refine the parameters of the spectroscopic orbit. The last night, UT 2007 Septem-
ber 20, was a transit night, and we gathered 25 spectra over 7.3 hr bracketing the predicted
transit midpoint, of which 16 were gathered during the transit.
We performed the Doppler analysis with the algorithm of Butler et al. (1996). For the
Subaru data we used a version of this algorithm customized for HDS by Sato et al. (2002).
We estimated the measurement error in the Doppler shift derived from a given spectrum
based on the weighted standard deviation of the mean among the solutions for individual
2˚ A spectral segments. The typical measurement error was 3 m s−1for the Keck data and
7 m s−1for the Subaru data. The data are given in Table 1 and plotted in Figs. 1 and 4.
Also given in that table, and shown in those figures, are data based on the 9 Keck/HIRES
spectra and 4 Subaru/HDS spectra obtained previously by Bakos et al. (2007). We note
that the RV timestamps reported by Bakos et al. (2007) are incorrect. They were said to
be Heliocentric Julian dates, but they are actually Julian dates. We provide the corrected
dates in Table 1.
We obtained photometric measurements of HAT-P-1 during the transit of UT 2007 Oct 8
using the Nickel 1m telescope at Lick Observatory on Mount Hamilton, California. We
used the Nickel Direct Imaging Camera, which is a thinned Loral 20482CCD with a 6.3′
– 5 –
square field of view1. We observed through a Gunn Z filter, and used 2 × 2 binning for an
effective pixel scale of 0.′′37 pixel−1. The exposure times varied depending upon conditions
but were typically 10-12 s, with a readout and setup time between exposures of 34 s. The
conditions were clear for most of the transit with ∼ 1.′′0 seeing. However, observations
during ingress were partially obscured by clouds and the data from that time period proved
to be significantly noisier than the rest; we have excluded those data from our analysis. We
determined the instrumental magnitude of HAT-P-1 relative to two comparison stars using
an aperture with an 11 pixel radius and a sky background annulus extending from 15 to 18
We observed the transit of UT 2007 September 20 with the MAGNUM 2m telescope on
Haleakala, in Hawaii (Kobayashi et al. 1998; Yoshii 2002; Yoshii et al. 2003). The MAGNUM
photometric observations were conducted on the same night as the Subaru/HDS transit
observations described in § 2.1. We employed the Multicolor Imaging Photometer (MIP),
using a 10242SITe CCD with a pixel scale of 0.′′277 pixel−1. The camera’s field of view is
1.′5, which is much smaller than the field of view of the detector. During each exposure, the
field was shifted on the detector along a 3 × 3 grid, which allowed us to increase the duty
cycle since the chip was read out only once for every 9 exposures. Observations were made
through a Johnson V -band filter, and the exposure times were 10 s, with 40 s per exposure
for readout and setup. The MIP images were reduced with the standard pipeline described
by Minezaki et al. (2004). We determined the instrumental magnitude of HAT-P-1 relative
to its visual binary companion, ADS16402A, using an aperture radius of 15 pixels, and
estimated the sky background level with an annulus from 20 to 25 pixels.
The photometric data are given in Table 2 and plotted in Fig. 2. In the final light
curves, the root-mean-squared (rms) relative flux, outside of transits, is 0.0019 for the Nickel
data and 0.0016 for the MAGNUM data.
3. The Model
3.1. An Updated Ephemeris
The extended time baseline of our new photometric measurements allows us to refine
the transit ephemeris. We first computed midtransit times from the light curves using the
method described by Winn et al. (2007a). In particular, to assign proper weights to the
1This is the same camera used by Winn et al. (2007a), which they mistakenly described as a 20482
Lawrence Labs CCD with a 6.′1 × 6.′1 field of view.
– 25 –
Fig. 1.— Top: Relative radial velocity measurements of HAT-P-1, from this work and
from Bakos et al. (2007). The solid line is the best-fitting model. The typical measurement
uncertainties are illustrated as points with error bars in the lower left corner. A detailed view
near midtransit is shown in Figure 4. Bottom: Residual radial velocities after subtracting
the best-fitting model.
– 26 –
Fig. 2.— Photometry of transits of HAT-P-1, using the Nickel 1m telescope and a Z-band
filter (top), and the MAGNUM 2m telescope and a V -band filter (bottom). The solid lines
show the best-fitting model.
– 27 –
Fig. 3.— Residuals from a linear ephemeris that was fitted to the transit times in Table 3.
The best–fitting ephemeris has a period P = 4.4652934 ± 0.0000093 days and midtransit
time Tc= 2454363.94656 ± 0.00072 (HJD).
– 28 –
Fig. 4.— Top: Relative radial velocity measurements of HAT-P-1, centered on the midtran-
sit time. Bottom: Composite, time-binned transit light curve of HAT-P-1, based on the Z
and z-band data from Winn et al. (2007) and this work. The rms scatter of the residuals is
0.00057. This composite light curve was fitted simultaneously with the radial-velocity data.