HAT-P-16b: A 4  planet transiting a bright star on an eccentric orbit 1 2


We report the discovery of HAT-P-16b, a transiting extrasolar planet orbiting the F8 dwarf GSC 2792-01700, with a period , transit epoch (BJD27 ), and transit duration  d. The host star has a mass of , radius of , effective temperature  K, and metallicity . The planetary companion has a mass of , and radius of  yielding a mean density of . Comparing these observed characteristics with recent theoretical models, we find that HAT-P-16b is consistent with a 1 Gyr H/He-dominated gas giant planet. HAT-P-16b resides in a sparsely populated region of the mass–radius diagram and has a non-zero eccentricity of with a significance of .

Subject headings:
planetary systems — stars: individual (HAT-P-16, GSC 2792-01700) techniques: spectroscopic, photometric

1. Introduction

Planets that transit their host stars play a special role in our understanding of the characteristics of exoplanets: their transit allows us to accurately determine the radius and the orbital inclination of the planet from the photometric light curve so that an actual mass can be derived from a spectroscopic orbit of the host star. The mass and radius enable us to infer a bulk composition of the planet, and although there are degeneracies associated with the bulk composition, it allows us to put constraints on models of planetary structure and formation theories. The incredible diversity of the over 60 discovered transiting planets, ranging from dense planets with a higher mean density than that of copper to strongly irradiated puffed-up planets with a mean density comparable to that of corkwood, have baffled the exoplanet community, and no unified theory has been established to explain all the systems consistently. Transiting extrasolar planet (TEP) discoveries are primarily the result of dedicated ground–based searches, such as SuperWASP (Pollacco et al., 2006), HATNet (Bakos et al., 2004), TrES (Alonso et al., 2004) and XO (McCullough et al., 2005), and space-borne searches, such as CoRoT (Baglin et al., 2006) and Kepler (Borucki et al., 2010).

Since its commissioning in 2003, the Hungarian-made Automated Telescope Network (HATNet; Bakos et al., 2004) survey has been one of the major contributors to the discoveries of TEPs. HATNet has discovered over a dozen TEPs since 2006 by surveying bright stars ( mag) in the Northern hemisphere and has now covered approximately 11% of the Northern sky. HATNet consists of six wide field automated telescopes; four of these are located at the Fred Lawrence Whipple Observatory (FLWO) in Arizona, and two on the roof of the Submillimeter Array hangar (SMA) of the Smithsonian Astrophysical Observatory (SAO) in Hawaii. In this paper we report a new TEP discovery of HATNet, called HAT-P-16b.

The structure of the paper is as follows. In Section 2 we summarize the observations, including the photometric detection, and follow-up observations. In Section 3 we describe the analysis of the data, such as the stellar parameter determination (Section 3.1), ruling out blend scenarios (Section 3.2), and global modeling of the data (Section 3.3). We discuss our findings in Section 4.

2. Observations

2.1. Photometric detection

Figure 1.— Unbinned light curve of HAT-P-16 including all 12,552 instrumental  band and  band 5.5 minute cadence measurements obtained with the HAT-6, HAT-7, HAT-8 and HAT-9 telescopes (see the text for details), and folded with the period of  days (which is the result of the fit described in Section 3). The  band and  band data are merged in this figure. The solid line shows the transit model fit to the light curve (Section 3.3). In the lower panel, the large solid circles represent the binned average of the photometric measurements with a bin size of 0.005 days.

The transits of HAT-P-16b were detected with the HAT-6 and HAT-7 telescopes in Arizona and the HAT-8 and HAT-9 telescopes in Hawaii. The star GSC 2792-01700 lies in the intersection of 3 separate HATNet fields internally labeled as 123, 163 and 164. Field 123 was observed with an filter and a 2K2K CCD, while fields 163 and 164 were observed through filters with 4K4K CCDs. All three fields were observed with a 5 minute exposure time and at a 5.5 minute cadence in the period between July 2007 and September 2008 during which over 12,000 exposures were gathered for the three fields. Each image contains between 27,000 and 76,000 stars down to a magnitude of for field 123 and for fields 163 and 164, yielding a photometric precision for the brightest stars in the field of  mmag for field 123 and  mmag for fields 163 and 164.

Frame calibration, astrometry and aperture photometry were done in an identical way to recent HATNet TEP discoveries, as described in Bakos et al. (2009) and Pál (2009). The resulting light curves were decorrelated (cleaned of trends) using the External Parameter Decorrelation technique in “constant” mode (EPD; see Bakos et al., 2009) and the Trend Filtering Algorithm (TFA; see Kovács et al., 2005). The light curves were searched for periodic box-like signals using the Box Least-Squares method (BLS; see Kovács et al., 2002). A significant signal was detected in the light curve of GSC 2792-01700 (also known as 2MASS 00381756+4227470; , ; J2000; ; Droege et al., 2006), with a depth of  mmag, and a period of  days. The dip had a relative duration (first to last contact) of , corresponding to a total duration of  hr (see Figure 1).

2.2. Reconnaissance Spectroscopy

Transiting planet candidates found by ground-based, wide-field photometric surveys must undergo a rigorous vetting process to eliminate the many astrophysical systems mimicking transiting planets (called false positives), the rate of which has proved to be much higher than the occurrence of true planets (10 to 20 times higher). Low signal to noise ratio (S/N) high-resolution reconnaissance spectra are used to extract stellar parameters such as effective temperature, gravity, metallicity and rotational and radial velocities to rule out these false positives. Examples of false positives which are discarded are eclipsing binaries and triple systems. The latter can be either hierarchical or chance alignment systems where the light of the eclipsing pair of stars is diluted by the light of a third brighter star. Rapidly rotating and/or hot host stars whose spectrum is unsuitable for high precision velocity work are also discarded.

We acquired  reconnaissance spectra with the CfA Digital Speedometer (DS; Latham, 1992) mounted on the FLWO 1.5 m Tillinghast Reflector between December 2008 and January 2009. The extracted modest precision radial velocities gave a mean absolute RV=  with an rms of , which is consistent with no detectable RV variation. The stellar parameters derived from the spectra (Torres, Neuhäuser, & Guenther, 2002), including the effective temperature , surface gravity (log cgs) and projected rotational velocity , correspond to a F8 dwarf.

2.3. High resolution, high S/N spectroscopy

The high significance of the transit detection by HATNet, together with the stellar spectral type and small RV variations encouraged us to gather high–resolution, high S/N spectra to determine the orbit of the system. We have taken 21 spectra between August and October 2009 using the FIbre–fed Échelle Spectrograph (FIES) at the 2.5 m Nordic Optical Telescope (NOT) at La Palma, Spain (Djupvik & Andersen, 2009). We used the medium and the high–resolution fibers ( projected diameter) with resolving powers of and , respectively, giving a wavelength coverage of . Recently, FIES has also been used to obtain a spectroscopic orbit for one of the first Kepler planets, namely Kepler-7b (Latham et al., 2010).

We also used the HIRES instrument (Vogt et al., 1994) on the Keck I telescope located on Mauna Kea, Hawaii. We obtained 6 exposures with an iodine cell, plus 1 iodine-free template with Keck-I/HIRES. The observations were made on 6 nights between 2009 July 3 and 2009 October 29. The width of the spectrometer slit used on HIRES was , resulting in a resolving power of , with a wavelength coverage of  Å. The iodine gas absorption cell was used to superimpose a dense forest of lines on the stellar spectrum and establish an accurate wavelength fiducial (see Marcy & Butler, 1992). Relative RVs in the Solar System barycentric frame were derived as described by Butler et al. (1996), incorporating full modeling of the spatial and temporal variations of the instrumental profile.

The final RV data and their errors (for both instruments) are listed in Table 1. The folded data are plotted in Figure 2. The systemic gamma velocities (that are the result of the global modeling, as laid out in Section 3) have been subtracted to ensure a common zero–point. The best orbital fit (see Section 3) is superimposed in the figure.

Figure 2.— Top: RV measurements from the Nordic Optical Telescope and Keck for HAT-P-16, along with an orbital fit, shown as a function of orbital phase, using our best-fit period (see Section 3). Velocities from NOT using the medium and high resolution fiber are shown as squares and triangles, respectively, and the Keck velocities are shown as circles. Zero phase is defined by the transit midpoint. The center-of-mass velocity has been subtracted. Note that the error bars include the stellar jitter, added in quadrature to the formal errors given by the spectrum reduction pipeline. Second panel: phased residuals after subtracting the orbital fit (also see Section 3). The rms variation of the residuals is about . Third panel: bisector span variations (BS) including the template spectrum (Section 3.2). Bottom: relative S activity index values which have only been calculated for the Keck spectra. Note the different vertical scales of the panels.

Since this is the first time the NOT has been used to determine the orbit of a HATNet planet, we describe briefly the extraction of the radial velocities. We have built a custom pipeline to rectify and cross correlate spectra from Échelle spectrographs with the goal of providing very precise radial velocities. The first step in the extraction process is to remove the bias level and crop the raw images. To effectively remove cosmic rays, the observation is divided into three separate exposures, which enables us to combine the raw images using median filtering, removing virtually all cosmic rays. We use a flat field to trace the Échelle orders and we rectify the spectra by using the “optimal extraction” algorithm (Hewett et al., 1985; Horne, 1986). The blaze function is determined by fitting cubic splines to a combined high S/N flat field exposure and is saved separately in order to preserve the original flux in the stellar exposure. By dividing the normalized blaze function into the rectified flat field spectrum, we can determine the pixel to pixel variations and correct for these. The scattered light in the two–dimensional raw image is determined and removed by masking out the signal in the Échelle orders and fitting the inter–order scattered light flux with a two–dimensional polynomial.

Thorium argon (ThAr) calibration images are obtained through the science fiber before and after the stellar observation. The two calibration images are combined to form the basis for the fiducial wavelength calibration. We have determined that the best wavelength solution is achieved by choosing an exposure time that saturates the strong argon lines but enhances the forest of weaker thorium lines.

FIES is not a vacuum spectrograph and is only temperature controlled to . Consequently, the radial velocity errors are dominated by our limited ability to correct for shifts due to pressure, humidity and temperature variations. In order to successfully remove these large variations (), it is critical that the ThAr light travels through the same light path as the stellar light and thus acts as an effective proxy to remove these variations. We have therefore chosen to interleave the stellar observation between two ThAr exposures instead of using the simultaneous ThAr technique, which may not exactly describe the induced shifts in the science fiber. The centers of the ThAr lines are found by fitting a Gaussian function to the line profiles and a two– dimensional fifth order Legendre polynomial is used to describe the wavelength solution.

Once the spectra have been extracted, a multi–order cross correlation is performed order by order. First, the spectra are interpolated to a common oversampled log wavelength scale with the same monotonic wavelength increment in all orders. A high and low pass filter is applied in the Fourier domain and the ends of the spectra are apodized with a cosine bell function. The orders are cross correlated using a Fast Fourier Transform (FFT) and the cross correlation function (CCF) for each order is co–added. This automatically weights each order by its flux, giving more weight to the orders with more photons. This CCF peak is fitted with a Gaussian function to determine its center. The internal precision is estimated by finding the radial velocity for each order and the RV error is thus , where is the RV of the individual orders and is the number of orders. For the final radial velocities, a template spectrum is constructed by shifting and co–adding all the observed spectra, and the individual spectra are cross correlated against this co–added template spectrum to minimize the contribution of noise in the template.

BJD RV28 BS Inst.
    (2,455,000) () () () ()

Note. – For the iodine-free template exposure we do not measure the RV but do measure the BS. Such template exposures can be distinguished by the missing RV value. The Inst. column refers to the instrument used, i.e. the FIES spectrograph at the NOT using the medium and high resolution fibers or the HIRES spectrograph at Keck I. and are formal statistical errors.

Table 1 Relative radial velocity and bisector span variation measurements of HAT-P-16.

2.4. Photometric follow-up observations

Figure 3.— Unbinned instrumental  band transit light curves, acquired with KeplerCam at the FLWO 1.2 m telescope. Superimposed are the best-fit transit model light curves. At the bottom of the figure we show the residuals from the fit. Error bars represent the photon and background shot noise, plus the readout noise.

To confirm the transit signal and obtain high-precision light curves for modeling the system, we conducted photometric follow-up observations with the KeplerCam CCD on the FLWO 1.2 m telescope. We observed 4 transit events of HAT-P-16 on the nights of 2009 September 10, September 21, October 19 and October 30 (Figure 3). On the four nights, and frames were acquired with a cadence of and seconds ( and seconds of exposure time) in the Sloan band, respectively.

The reduction of the images, including frame calibration, astrometry, and photometry were performed as described in Bakos et al. (2009). We also performed EPD and TFA to remove trends simultaneously with the light curve modeling (for more details, see Section 3). The final light curves are shown in the upper plots of Figure 3, superimposed with our best-fit transit light curve models (see also Section 3); the photometry is provided in Table 2.

BJD Mag29 Mag(orig)30 Filter

Note. – This table is available in a machine-readable form in the online journal. A portion is shown here for guidance regarding its form and content.

Table 2Follow-up photometry of HAT-P-16

3. Analysis

3.1. Properties of the parent star

The stellar atmospheric parameters were derived from the template spectrum obtained with the Keck/HIRES instrument. We used the Spectroscopy Made Easy (SME) analysis package of Valenti & Piskunov (1996), along with the atomic-line database of Valenti & Fischer (2005). This yielded the following initial values and uncertainties (which we have conservatively increased to include our estimates of the systematic errors): effective temperature  K, stellar surface gravity  (cgs), metallicity  dex, and projected rotational velocity .

We could now use the effective temperature and the surface gravity found by SME to determine other stellar parameters, such as the mass, , and radius, , using model isochrones. However, there may be strong correlations between effective temperature, gravity and metallicity in the spectroscopic determination of these parameters. Also, the effect of  on the spectral line shapes is typically subtle, and as a result  is generally a rather poor luminosity indicator. Instead, we used the , the normalized semimajor axis (related to , the mean stellar density), which can be derived directly from the light curves (Sozzetti et al., 2007).

The stellar parameters were determined simultaneously with the modeling of the light curves and radial velocities, as described next. We began by adopting the values of , , and  from the SME analysis to fix the quadratic limb-darkening coefficients from the tabulations by Claret (2004), which are needed to model the light curves (, ). This modeling yields the probability distribution of  via a Monte Carlo approach, which is described fully in Section 3.3. We then used the  distribution together with Gaussian distributions for  and , with 1 uncertainties as reported previously, to estimate  and  by comparison with the stellar evolution models of Yi et al. (2001). This was done for each of 15,000 simulations. Parameter combinations corresponding to unphysical locations in the H-R diagram (4% of the trials) were ignored, and replaced with another randomly drawn parameter set. The resulting stellar parameters and their uncertainties were determined from the a posteriori distributions obtained in this way.

In particular, the resulting surface gravity of is somewhat different from that derived in the initial SME analysis, which is not surprising in view of the possible strong correlations among , , and the surface gravity. Therefore, in a second iteration we adopted this value of  and held it fixed in a new SME analysis, adjusting only , , and . This gave  K, , and , which we adopt as final atmospheric values for the star. Finally, we repeated the calculation of the stellar mass and radius with these values and the stellar evolution models, and obtained =, = and =, along with other parameters summarized in Table 3.

The model isochrones from Yi et al. (2001) for metallicity = are plotted in Figure 4, with the final choice of effective temperature and  marked, and encircled by the 1 and 2 confidence ellipsoids.

Figure 4.— Stellar isochrones from Yi et al. (2001) for metallicity = and ages 0.2, 0.4, 0.6, 0.8, 1.0, 1.5, 2.0, 2.5 and 3.0 Gyr (left to right). The final choice of and  are marked and encircled by the 1 and 2 confidence ellipsoids. The initial value of  and  from the first SME analysis and subsequent light curve analysis is marked with a triangle that is barely offset from the final value (indicated by a filled circle).

The stellar evolution modeling provides color indices that may be compared against the measured values as a sanity check. The best available measurements are the near-infrared magnitudes from the 2MASS Catalogue (Skrutskie et al., 2006), , and ; which we have converted to the photometric system of the models (ESO) using the transformations by Carpenter (2001). The resulting measured color index is . This is within 1 of the predicted value from the isochrones of . The distance to the object may be computed from the absolute magnitude from the models () and the 2MASS magnitude, which has the advantage of being less affected by extinction than optical magnitudes. The result is  pc, where the uncertainty excludes possible systematics in the model isochrones that are difficult to quantify.

Parameter Value Source
(K) SME31
(dex) SME
() SME
() SME
() SME
() DS
() YY++SME33
() YY++SME
(cgs) YY++SME
() YY++SME
(mag) 10.812 TASS
(mag) YY++SME
(mag,ESO) 2MASS+Carpenter34
(mag,ESO) YY++SME
Age (Gyr) YY++SME
Distance (pc) YY++SME
Table 3 Stellar parameters for HAT-P-16

3.2. Excluding blend scenarios

We performed a line bisector analysis of the observed spectra following Torres et al. (2007) to explore the possibility that the main cause of the radial velocity variations is actually distortions in the spectral line profiles due to stellar activity or a nearby unresolved eclipsing binary. The bisector analysis was performed on the Keck spectra as described in § 5 of Bakos et al. (2007a); the spectra from the NOT were analyzed in a similar manner.

The resulting bisector span variations are plotted in Figure 2 and have a low amplitude compared to the orbital semi–amplitude and show no correlation with the radial velocities. We therefore conclude that the velocity variations are real, and that the star is transited by a close-in giant planet.

3.3. Global modeling of the data

Our model for the follow-up light curves used analytic formulae based on Mandel & Agol (2002) for the eclipse of a star by a planet, where the stellar flux is described by quadratic limb-darkening. The limb darkening coefficients are based on the SME results (Section 3.1) and the tables provided by Claret (2004) for the band. The transit shape was parametrized by the normalized planetary radius , the square of the impact parameter , and the reciprocal of the half mid-ingress to mid-egress duration of the transit . We chose these parameters because of their simple geometric meanings and the fact that they show negligible correlations (see Bakos et al., 2009). Our model for the HATNet data was the simplified “P1P3” version of the Mandel & Agol (2002) analytic functions (an expansion by Legendre polynomials), for the reasons described in Bakos et al. (2009). Following the formalism presented by Pál (2009), the RV curve was modeled as an eccentric Keplerian orbit with semiamplitude , and Lagrangian orbital elements

We assumed that there is a strict periodicity in the individual transit times. In practice, we assigned the transit number to the first high quality follow-up light curve gathered on 2009 September 10. The adjusted parameters in the fit were the first transit center observed with HATNet, , and the last high quality transit center observed with the FLWO 1.2 m telescope, . The transit center times for the intermediate transits were interpolated using these two epochs and the transit number of the actual event (Bakos et al., 2007b). The model for the RV data contains the ephemeris information through the and variables. Altogether, the 8 parameters describing the physical model were , , , , , , , . Nine additional parameters were related to the instrumental configuration. These are the three instrumental blend factors of HATNet (one for each of the three fields), which account for possible dilution of the transit in the HATNet light curve due to the wide (20″ wide FWHM) PSF and possible crowding, the three HATNet out-of-transit magnitudes, , and three relative RV zero–points (one each for the Keck, high-resolution FIES, and medium-resolution FIES observations).

We extended our physical model with an instrumental model that describes the systematic variations of the data. This was done in a similar fashion to the analysis presented in Bakos et al. (2009). The HATNet photometry has already been EPD- and TFA-corrected before the global modeling, so we only considered systematic correction of the follow-up light curves. We chose the “ELTG” method, i.e. EPD was performed in “local” mode with EPD coefficients defined for each night, and TFA was performed in “global” mode using the same set of stars and TFA coefficients for all nights. The underlying physical model was based on the Mandel & Agol (2002) analytic formulae, as described earlier. The five EPD parameters were the hour angle (characterizing a monotonic trend that linearly changes over time), the square of the hour angle, and the stellar profile parameters (equivalent to FWHM, elongation, position angle). The functional form of the above parameters contained six coefficients, including the auxiliary out-of-transit magnitude of the individual events. The EPD parameters were independent for all four nights, implying 24 additional coefficients in the global fit. For the global TFA analysis we chose 20 template stars that had good quality measurements for all nights and on all frames, implying an additional 20 parameters in the fit. Thus, the total number of fitted parameters was 17 (physical model) + 24 (local EPD) + 20 (global TFA) = 61.

The joint fit was performed as described in Bakos et al. (2009). We minimized  in the parameter space by using a hybrid algorithm, combining the downhill simplex method (AMOEBA; see Press et al., 1992) with the classical linear least squares algorithm. Uncertainties for the parameters were derived using the Markov Chain Monte-Carlo method (MCMC, see Ford, 2006; Pál, 2009).

The eccentricity of the system appeared as significantly non-zero (, ). The best-fit results for the relevant physical parameters are summarized in Table 4. We also list the RV “jitter,” which is a component of assumed astrophysical noise intrinsic to the star that we add in quadrature to the RV measurement uncertainties in order to have from the RV data for the global fit. In addition, some auxiliary parameters (not listed in the table) are:  (BJD),  (BJD), , (FIES, high resolution), (FIES, medium resolution), , , . Note that these gamma velocities do not correspond to the true center of mass RV of the system, but are only relative offsets. The true systemic velocity of the system is RV=found by cross–correlating the observed spectrum against a library template spectrum. The planetary parameters and their uncertainties can be derived by the direct combination of the a posteriori distributions of the light curve, RV and stellar parameters. We found that the planet is fairly massive with , and compact with radius of , yielding a mean density of . The final planetary parameters are summarized at the bottom of Table 4.

               Parameter Value
Light curve parameters
    (days) 35
    (days) 36
RV parameters
   RV jitter ()
   RV rms from fit ()
Secondary eclipse parameters
Planetary parameters
    () 41
Table 4Orbital and planetary parameters

4. Discussion

We present the discovery of HAT-P-16b with a period of , a mass of and a radius of . Currently, there are only a handful of planets residing near HAT-P-16b in the mass–radius diagram. CoRoT-2b (Alonso et al., 2008) and CoRoT-6b (Fridlund et al., 2010) have similar masses and periods, but both have eccentricities consistent with zero. HD80606 (Naef et al., 2001) and HD17156 (Fischer et al., 2007) are also similar in mass, but both planets have long periods and very eccentric orbits. WASP-10b ( d, , see Christian et al. 2009; Johnson et al. 2009) is probably the transiting planet which most resembles HAT-P-16b  with a similar period and eccentricity, albeit slightly lower mass and radius. At , HAT-P-16 is one of the brightest stars known to host a planet.

We have compared HAT-P-16b to the theoretical models from Fortney et al. (2007) by interpolating the models to the solar-equivalent semi–major axis of , the result of which can be seen overplotted in Figure 5. We find that the mass and radius of HAT-P-16b are quite consistent with the 1 Gyr model and conclude that HAT-P-16b is most likely a H/He-dominated planet.

The radial velocity dataset used to characterize HAT-P-16b consist of three different velocity sets from two different telescopes (FIES medium and high resolution spectra from the NOT and HIRES spectra from Keck-I), requiring additional free parameters in the global fit to account for the arbitrary velocity offset between the measurements. Nevertheless, the different observations yield very similar semi-amplitude and we find an rms from the best fit model of only . The rich dataset consisting of 27 RV observations allows us to accurately determine the small non-zero eccentricity of the orbit as with a high significance of , with at and at . The eccentricity is consistent and clearly evident even when fitting separate orbits to the three different velocity datasets.

Most short–period TEPs have eccentricities consistent with zero, as is expected because of circularization driven by tidal dissipation within the planet with a typical timescale substantially less than 1 Gyr (see Jackson, Greenberg, & Barnes 2008; Mardling 2007). In fact, the estimated circularization timescale for HAT-P-16b is (when the tidal dissipation parameter is assumed to be ), which is much less than the estimated age of the system. On the other hand, Matsumura et al. (2008) argue that circularization timescales could be significantly higher and the planets with eccentric orbits may simply be in the process of being circularized. Adams & Laughlin (2006) argue that for multiple systems with a hot Jupiter as the inner planet, secular excitation of the eccentricity by companion planets could explain the non–zero eccentricity of these systems. Thus the origin of the eccentricity of HAT-P-16b remains unclear.

With a stellar rotation of  and an impact parameter of , HAT-P-16 is a favorable target for measuring the Rossiter–McLaughlin (RM) effect. This effect can be used to determine the alignment of the planet’s orbital angular momentum vector with the stellar spin axis (Winn et al., 2009). The estimated amplitude of the RM effect is for HAT-P-16b.

Figure 5.— Mass–radius diagram of currently known TEPs. HATNet planets are show in green and HAT-P-16b is labeled explicitly. The Solar system planets are shown as filled gray squares. Isodensity curves (in ) are plotted as dotted lines. Overlaid are planetary 1.0 Gyr (brown, the upper set of lines) and 4.5 Gyr (green, the lower set of lines) isochrones from Fortney et al. (2007) for H/He-dominated planets with core mases of (solid) and (dashed), interpolated to the solar equivalent semimajor axis of HAT-P-16b.
HATNet operations have been funded by NASA grants NNG04GN74G, NNX08AF23G and SAO IR&D grants. Work of G.Á.B. and J. Johnson were supported by the Postdoctoral Fellowship of the NSF Astronomy and Astrophysics Program (AST-0702843 and AST-0702821, respectively). GT acknowledges partial support from NASA grant NNX09AF59G. We acknowledge partial support also from the Kepler Mission under NASA Cooperative Agreement NCC2-1390 (D.W.L., PI). G.K. thanks the Hungarian Scientific Research Foundation (OTKA) for support through grant K-81373. Based in part on observations made with the Nordic Optical Telescope, operated on the island of La Palma jointly by Denmark, Finland, Iceland, Norway, and Sweden, in the Spanish Observatorio del Roque de los Muchachos of the Instituto de Astrofisica de Canarias. This research has made use of Keck telescope time granted through NASA (N018Hr).


  1. affiliation: Based in part on observations made with the Nordic Optical Telescope, operated on the island of La Palma jointly by Denmark, Finland, Iceland, Norway, and Sweden, in the Spanish Observatorio del Roque de los Muchachos of the Instituto de Astrofisica de Canarias
  2. affiliation: Based in part on observations obtained at the W. M. Keck Observatory, which is operated by the University of California and the California Institute of Technology. Keck time has been granted by NASA (N018Hr).
  3. affiliation: Harvard-Smithsonian Center for Astrophysics, Cambridge, MA
  4. affiliation: Niels Bohr Institute, Copenhagen University, Denmark
  5. affiliation: Harvard-Smithsonian Center for Astrophysics, Cambridge, MA
  6. affiliation: NSF Fellow
  7. affiliation: Harvard-Smithsonian Center for Astrophysics, Cambridge, MA
  8. affiliation: Harvard-Smithsonian Center for Astrophysics, Cambridge, MA
  9. affiliation: Konkoly Observatory, Budapest, Hungary
  10. affiliation: Harvard-Smithsonian Center for Astrophysics, Cambridge, MA
  11. affiliation: Harvard-Smithsonian Center for Astrophysics, Cambridge, MA
  12. affiliation: Harvard-Smithsonian Center for Astrophysics, Cambridge, MA
  13. affiliation: Department of Astronomy, University of California, Berkeley, CA
  14. affiliation: Department of Astronomy, University of California, Berkeley, CA
  15. affiliation: Department of Astronomy, Yale University, New Haven, CT 06511
  16. affiliation: Department of Astrophysics, California Institute of Technology, MC 249-17, Pasadena, CA 91125, USA
  17. affiliation: Niels Bohr Institute, Copenhagen University, Denmark
  18. affiliation: Nordic Optical Telescope Scientific Association, La Palma, Canarias, Spain
  19. affiliation: Harvard-Smithsonian Center for Astrophysics, Cambridge, MA
  20. affiliation: Harvard-Smithsonian Center for Astrophysics, Cambridge, MA
  21. affiliation: Harvard-Smithsonian Center for Astrophysics, Cambridge, MA
  22. affiliation: Harvard-Smithsonian Center for Astrophysics, Cambridge, MA
  23. affiliation: Harvard-Smithsonian Center for Astrophysics, Cambridge, MA
  24. affiliation: Hungarian Astronomical Association, Budapest, Hungary
  25. affiliation: Hungarian Astronomical Association, Budapest, Hungary
  26. affiliation: Hungarian Astronomical Association, Budapest, Hungary
  27. affiliation: All BJD values given in this paper are UTC-based (e.g. Torres et al. (2010))
  28. The fitted zero-point that is on an arbitrary scale (denoted as in Section 3.3) has not been subtracted from the velocities.
  29. The out-of-transit level has been subtracted. These magnitudes have been subjected to the EPD and TFA procedures, carried out simultaneously with the transit fit.
  30. These are raw magnitude values without application of the EPD and TFA procedures.
  31. SME = “Spectroscopy Made Easy” package for analysis of high-resolution spectra Valenti & Piskunov (1996). These parameters depend primarily on SME, with a small dependence on the iterative analysis incorporating the isochrone search and global modeling of the data, as described in the text.
  32. SME+Claret = Based on the SME analysis and tables of quadratic limb-darkening coefficients from Claret (2004).
  33. YY++SME = YY2 isochrones (Yi et al., 2001),  luminosity indicator, and SME results.
  34. Based on the relations from Carpenter (2001).
  35. : total transit duration, time between first to last contact; : ingress/egress time, time between first and second, or third and fourth contact.
  36. : total transit duration, time between first to last contact; : ingress/egress time, time between first and second, or third and fourth contact.
  37. Lagrangian orbital parameters derived from the global modeling, and primarily determined by the RV data.
  38. Lagrangian orbital parameters derived from the global modeling, and primarily determined by the RV data.
  39. Correlation coefficient between the planetary mass  and radius .
  40. The Safronov number is given by (see Hansen & Barman, 2007).
  41. Incoming flux per unit surface area.


  1. Adams, F. C., & Laughlin, G. 2006, ApJ, 649, 992
  2. Alonso, R., Brown, T. M., Torres, G., Latham, D. W., Sozzetti, A., Mandushev, G., Belmonte, J. A., Charbonneau, D., Deeg, H. J., Dunham, E. W., O’Donovan, F. T., & Stefanik, R. P. 2004, ApJ, 613, L153
  3. Alonso, R., et al. 2008, A&A, 482, L21
  4. Bakos, G. Á., Noyes, R. W., Kovács, G., Stanek, K. Z., Sasselov, D. D., & Domsa, I. 2004, PASP, 116, 266
  5. Bakos, G. Á., et al. 2007a, ApJ, 670, 826
  6. Bakos, G. Á., et al. 2007b, ApJ, 670, 826
  7. Bakos, G. Á., et al. 2009, arXiv:0901.0282
  8. Baglin, A., Auvergne, M., Boisnard, L., Lam-Trong, T., Barge, P., Catala, C., Deleuil, M., Michel, E., & Weiss, W. 2006, 36th COSPAR Scientific Assembly, 36, 3749
  9. Barbieri, M., Alonso, R., Laughlin, G., Almenara, J. M., Bissinger, R., Davies, D., Gasparri, D., Guido, E., Lopresti, C., Manzini, F., & Sostero, G. 2007, A&A, 476, L13
  10. Barbieri, M., et al. 2007, A&A, 476, L13
  11. Borucki, W. J., et al. 2010, Science, 327, 977
  12. Butler, R. P. et al. 1996, PASP, 108, 500
  13. Carpenter, J. M. 2001, AJ, 121, 2851
  14. Christian, D. J., et al. 2009, MNRAS, 392, 1585
  15. Claret, A. 2004, A&A, 428, 1001
  16. Djupvik, A. A., & Andersen, J. 2009, arXiv:0901.4015
  17. Droege, T. F., Richmond, M. W., & Sallman, M. 2006, PASP, 118, 1666
  18. Fischer, D. A., Vogt, S. S., Marcy, G. W., Butler, R. P., Sato, B., Henry, G. W., Robinson, S., Laughlin, G., Ida, S., Toyota, E., Omiya, M., Driscoll, P., Takeda, G., Wright, J. T., & Johnson, J. A. 2007, ApJ, 669, 1336
  19. Ford, E. 2006, ApJ, 642, 505
  20. Fortney, J. J., Marley, M. S., & Barnes, J. W. 2007, ApJ, 659, 1661
  21. Fridlund, M., et al. 2010, arXiv:1001.1426
  22. Hansen, B. M. S., & Barman, T. 2007, ApJ, 671, 861
  23. Hewett, P. C., Irwin, M. J., Bunclark, P., Bridgeland, M. T., Kibblewhite, E. J., He, X. T., Smith, M. G. 1985, MNRAS, 213, 971
  24. Horne, K. 1986, PASP, 98, 609
  25. Jackson, B., Greenberg, R., & Barnes, R. 2008, ApJ, 678, 1396
  26. Johnson, J. A., Winn, J. N., Cabrera, N. E., & Carter, J. A. 2009, ApJ, 692, L100
  27. Kovács, G., Zucker, S., & Mazeh, T. 2002, A&A, 391, 369
  28. Kovács, G., Bakos, G. Á., & Noyes, R. W. 2005, MNRAS, 356, 557
  29. Latham, D. W. 1992, in IAU Coll. 135, Complementary Approaches to Double and Multiple Star Research, ASP Conf. Ser. 32, eds. H. A. McAlister & W. I. Hartkopf (San Francisco: ASP), 110
  30. Latham, D. W. Borucki, W. J., Koch, D. G., Brown, T. M., Buchhave, L. A., Basri, G., Batalha, N. M.,Caldwell, D. A., Cochran, W. D., Dunham, E. W., Furesz, G, Gautier III, T. N., Geary, J. C., Gilliland, R. L., Howell, S. B., et al. 2010, arXiv:1001.0190
  31. Mardling, R. A. 2007, MNRAS, 382, 1768
  32. Matsumura, S., Takeda, G., & Rasio, F. A. 2008, ApJ, 686, L29
  33. Naef, D., Latham, D. W., Mayor, M., Mazeh, T., Beuzit, J. L., Drukier, G. A., Perrier-Bellet, C., Queloz, D., Sivan, J. P., Torres, G., Udry, S., Zucker, S. 2001, A&A375
  34. Pál, A., & Bakos, G. Á. 2006, PASP, 118, 1474
  35. Pál, A. 2009, MNRAS, 396, 1737
  36. Press, W. H., Teukolsky, S. A., Vetterling, W. T. & Flannery, B. P., 1992, Numerical Recipes in C: the art of scientific computing, Second Edition, Cambridge University Press
  37. Mandel, K., & Agol, E. 2002, ApJ, 580, L171
  38. Marcy, G. W., & Butler, R. P. 1992, PASP, 104, 270
  39. McCullough, P. R., Stys, J. E., Valenti, J. A., Fleming, S. W., Janes, K. A., & Heasley, J. N. 2005, PASP, 117, 783
  40. Pollacco, D. L., et al. 2006, PASP, 118, 1407
  41. Skrutskie, M. F., et al. 2006, AJ, 131, 1163
  42. Sozzetti, A. et al. 2007, ApJ, 664, 1190
  43. Torres, G., Neuhäuser, R., & Guenther, E. W. 2002, AJ, 123, 1701
  44. Torres, G. et al. 2007, ApJ, 666, 121
  45. Torres, G. et al. 2010, in preparation
  46. Valenti, J. A., & Fischer, D. A. 2005, ApJS, 159, 141
  47. Valenti, J. A., & Piskunov, N. 1996, A&AS, 118, 595
  48. Vogt, S. S. et al. 1994, Proc. SPIE, 2198, 362
  49. Yi, S. K. et al. 2001, ApJS, 136, 417
  50. Winn, J. N., Johnson, J. A., Albrecht, S., Howard, A. W., Marcy, G. W., Crossfield, I. J., & Holman, M. J. 2009, ApJ, 703, L99
This is a comment super asjknd jkasnjk adsnkj
The feedback cannot be empty
Comments 0
The feedback cannot be empty
Add comment

You’re adding your first comment!
How to quickly get a good reply:
  • Offer a constructive comment on the author work.
  • Add helpful links to code implementation or project page.