VALIDATING KEPLER'S HABITABLE-ZONE SUPER-EARTHS WITH THE SPITZER SPACE TELESCOPE
The NASA Kepler Mission will soon have completed two years of science observations, the minumum baseline sufficient to identify candidate transiting planets orbiting with the habitable zones of Sun-like stars. The Kepler Team has already identified planets orbiting within the habitable zones of stars less massive that the Sun, and it will continue to uncover habitable zone planets as the mission baseline increases. The principal task that lies ahead is to reject from this sample the astrophysical false positives (blends of eclipsing binaries that precisely mimic the signal of a transiting exoplanet), and to confirm the planetary nature of the remaining candidates. For planets more massive than Neptune, the direct confirmation of their planetary status can be accomplished by radial- velocity measurements. However, such planets possess primordial envelopes of hydrogen and helium that make them unsuitable to life as we know it. The most exciting candidates-- and the ones that Kepler is specifically tasked with finding -- are super-Earth and Earth-sized candidates orbiting within their stellar habitable zones. While the Kepler Team has developed powerful tools to weed out impostors, the Spitzer Space Telescope possesses the unique capability to provide the final validation of these candidates as planets, namely by measuring the depth of the transit at infrared wavelengths. By combining the infrared and optical measurements of the transit depth with models of hypothetical stellar blends, we can definitively test the stellar-blend hypothesis. We have submitted an Exploration Science proposal to Spitzer requesting 600 hours of observations with the IRAC camera. We propose to use this time to observe the transits of 20 candidate habitable-zone super-Earths identified, or to be identified, by the Kepler Mission. For these investigations to succeed, they must be developed in direct partnership with the Kepler Science Team and planned with the use of the Kepler photometry, and the results from Spitzer must be shared directly with the Kepler Follow- On Program, all of which motivates the current Kepler Mission Participating Scientist proposal. The results from this proposal will be twofold: First, we will definitively validate the first potentially habitable planets ever identified. Second, we will determine the rate of occurrence of impostors. This rate of false positives can then be applied to the much larger sample of candidates identified by Kepler, to deduce the true rate of planetary companions. The proposed investigation will directly address the Kepler Mission goals to "provide a statistically significant value for the frequency of Earth-size and larger planets in and near the habitable zone of their host stars" and "charaterize the size and orbital distributions of planets around other stars", and the NASA Research Objective to "generate a census of extra-solar planets and measure their properties".
CHARACTERIZING THE ORBITAL ECCENTRICITIES OF EARTH-SIZE PLANETS AND ARCHITECTURES OF PLANETARY SYSTEMS WITH KEPLER
University of Florida
The discovery of over 400 extrasolar planets with the radial velocity technique revealed that many giant planets have large eccentricities, in striking contrast with most of the planets in the solar system and prior theories of planet formation. The realization that many giant planets have large eccentricities raises a fundamental question: ``Do terrestrial-size planets of other stars typically have significantly eccentric orbits or nearly circular orbits like the Earth?'' Theorists have proposed numerous mechanisms that could excite orbital eccentricities which make different predictions for the eccentricities for low-mass planets (as well as their mutual inclinations and frequency of multiple planet systems). We propose to use Kepler data (and follow-up observations when available) to characterize the orbital eccentricities of the 1235 transiting planet candidates found by Kepler. We will improve upon our previous results during Cycle 1 of the Kepler Participating Scientist Program by: 1) incorporating more accurate stellar parameters based on high-resolution spectroscopy and (when possible) astroseismology; 2) analyzing planet candidates in and near the habitable zone (already 54, and likely to grow with increased time baseline); 3) obtaining more precise parameters thank to Kepler photometry spanning an order of magnitude more time; 4) analyzing short-cadence data for favorable cases than enable an accurate measurement of the impact parameter; and 5) incorporating dynamical stability constraints for multiple planet systems. One third of the Kepler planet candidates are in systems with multiple transiting planet candidates. In some cases, the planets can be confirmed based on transit timing variations (TTVs), without the need for expensive Doppler observations. Indeed, two systems (7 planets) have been confirmed based on transit timing variations. We identified over 60 candidates likely to have transit timing variations based on the first four months of Kepler data. With increasing timespan of observations of these and other systems, we expect roughly five times as many systems to exhibit measurable TTVs, as well as much stronger constraints on masses and orbital eccentricities. We propose to continue our critical roles in the identification, analysis and interpretation of these systems. The relative frequency of multiply transiting systems (along with orbital separations and stellar radii) constrains the distribution of mutual orbital inclinations and rates of multiplicity. Multiplicity, eccentricity and inclination are all linked to the formation history. We will combine analyses of individual systems and the ensemble of systems with theoretical models and simulations to improve our understanding the origin and history of solar systems, and particularly Earth-size planets. Since a significant eccentricity will cause the incident stellar flux to vary, a planet's eccentricity affects its climate and potentially its habitability. Thus, this research would contribute to NASA's goal of searching for Earth-like planets and the results could influence the design of future space missions that will attempt to characterize Earth-like planets and search for signs of life.
APPLYING THE METHOD OF TRANSIT TIMING VARIATIONS TO KEPLER: PHASE 2
Smithsonian Astrophysical Observatory
Measuring the masses of Kepler planet candidates presents a daunting challenge. Most of the target stars are relatively faint. Most of the planets have small radii and thus are expected to have low masses. And some of the candidate planets have distant, long- period orbits. Thus, the radial velocity signatures are small and difficult to measure. Although some of planetary masses can be measured with high-precision spectrographs on large aperture telescopes, given ample amounts of observing time, in most cases an alternative means of establishing the planetary nature of these candidates must be used. The method of transit timing variations (TTVs) is just such a technique. For a single transiting planet, the variation in the interval between transits, produced by gravitational interactions with additional planets, allows for the detection of those perturbing planets. And detailed observations of those variations can, in principle, allow the orbital period and mass of the perturbing planets to be determined from transit observations alone (Holman & Murray 2005, Agol et al. 2005). Holman and Murray (2005) predicted that systems with multiple planets transiting the same star would be observed by Kepler. They noted that for some such systems the masses of all the planets could be measured from their mutual timing variations alone. The Kepler-9 (Holman et al. 2010) and Kepler-11 (Lissauer et al. 2011a) systems confirmed this prediction. In the case of Kepler-9, the larger two of its three known planets show substantial, anticorrelated TTVs. We measured the masses of Kepler-9b and Kepler-9c by modeling the transit times, along with six Keck HIRES RV observations (Holman et al. 2010). Kepler-9d was validated using a BLENDER analysis (Torres et al.2011). In the case of Kepler-11, with its closely packed configuration, the masses of five of its six transiting planets could be estimated from their TTVs, and the sixth planet was again validated with BLENDER (Lissauer et al. 2011a). Kepler has discovered an abundance of systems with multiple transiting planet candidates (Borucki et al. 2010, Steffen et al. 2010, Ragozzine and Holman 2010, Borucki et al. 2011, Lissauer et al. 2011b, Ford et al. 2011). A significant number of these are amenable to TTV analyses that will yield mass estimates with future Kepler data (Ford et al. 2011). In addition, some single transiting systems show significant TTVs; future data may permit us to establish the masses and orbits of the perturbers in these systems. I propose to continue contributing to the theoretical and analytical interpretation of the transit timing data from the Kepler mission. The specific scientific objectives of this proposal are: 1) to apply the TTV method to Kepler candidate systems that have not yet been confirmed; 2) to re-analyze currently known multiple planet systems, such as Kepler-9 and Kepler-11, as new data become available; and 3) to place upper limits on the masses of possible additional planets in systems with single and multiple confirmed planets or planet candidates.
DISCOVERY OF SHORT-PERIOD NON-TRANSITING PLANETS/BROWN-DWARF/LOW-MASS-STELLAR COMPANIONS THROUGH THE BEAMING (DOPPLER BOOSTING) EFFECT IN KEPLER LIGHTCURVES
Tel Aviv University
We developed a new algorithm, BEER, to search for a combination of the BEaming, Ellipsoidal and the Reflection/heating periodic effects induced by short-period non- transiting low-mass companions. The phases and amplitudes of the three effects form a periodic modulation with a specific signature, which is used by BEER to identify stellar candidates that have small non-transiting companions. The algorithm is constructed to work on the Kepler data and utilize the superb accuracy of its lightcurves. A paper describing the algorithm was submitted on Jan. 2nd , 2011 to MNRAS. BEER has already discovered in the Kepler data two very low-mass stellar companions, with masses, up to sine of the orbital inclination, of 88 and 73 Jupiter masses, on the border line between the stellar and the brown-dwarf mass ranges, with orbital periods of 5.6 and 3.5 days, respectively. The two detections were confirmed by radial-velocity observations.
Although we used only data from Q1, spanned over only 33 days, BEER could find periodic effects with amplitudes as small as one part in 10,000 of the stellar flux. With access to the full Kepler accumulating dataset we will be able to detect amplitudes induced by most planets with masses larger than 7-10 Jupiter masses, with orbital periods shorter than 10 days. The goal is to discover most of the massive-planet/brown- dwarf/low-mass-stellar companions with short orbital periods, and not only the transiting ones. The radial-velocity follow-up will be relatively easy, as the expected amplitudes of the proposed detections are on the order of 1-10 km/s. The proposed search will map in details the famous brown-dwarf desert (BDD), which lies between 10-20 Jupiter masses on one side and 60-80 Jupiter masses on the other side. The large sample of detections will enable us to find the low-mass end of the distribution of stellar companions and the high-mass end of planets. It will eventually help to settle the long-lasting heated discussion about the definition of a planet and its mass upper bound. The new detections will enable us to find the dependence of the BDD on the stellar mass and metallicity, shading more light on the planetary formation and dynamical evolution processes.
The success of such a search depends on access to the data as soon as they are ready. This is so, because comparing the phase of the follow-up RV observations with those of the photometric (ellipsoidal, beaming and reflection) modulations can solve the main problem of detecting orbital modulation through the beaming effect - how to exclude other interpretations of periodic modulation, stellar pulsation in particular. If the three or four observed modulations --- ellipsoidal, beaming, and possibly reflection effects, together with the RV variation, are all phased together as expected by the companion assumption, the attribution of those observations to a small companion is secure. Comparing the phase of the radial-velocity observations and the photometric modulation can be performed only if the RV observations are done not too late after the lightcurve was obtained; otherwise the knowledge of the photometric phase is lost. I therefore apply
for being a Participating Scientist in the next two years. The application is without a budget. The whole analysis will be done in Israel supported by other sources. The proposed project, when performed, will be equivalent to a radial-velocity Doppler continuous monitoring of tens of thousands of stars with a precision on the order of 1 km/s. It will change substantially our view of the mass distribution of massive planets and low-mass secondaries in short orbital periods and the brown-dwarf desert between the two populations.
THOROUGH UNDERSTANDING OF FOREGROUND AND BACKGROUND ECLIPSING BINARY STAR POPULATIONS OBSERVED BY KEPLER
The proposal emphesizes the importance of eclipsing binary stars observed by Kepler. Background eclipsing binaries are the leading source of false positives and a thorough analysis is required for planet transit validation. Studying the statistical distributions of the eclipsing binary periods, amplitudes and principal parameters allows a direct comparison with planet candidate distributions and hints to similarities and fundamental differences between the two types of systems. I build a case for the continued detection, classification and characterization of binaries in the Kepler field. In particular, I derive preliminary results from the first and the second EB catalog release and show that the observed distributions are in remarkable agreement with the distributions derived from the Besancon model of the Galaxy. I demonstrate that the published list of planet candidates by Borucki et al. (2011) is likely only marginally contaminated by false positives. I discuss the plans to look for circumstellar and circumbinary planets via transit and eclipse timing methods, and present other aspects of my involvement in the Mission that could substantially contribute to its success and be realized through this proposal.
UNIFORM MODELING OF KEPLER OBJECTS OF INTEREST
The central objective of the proposal is perform uniform state-of-the-art lightcurve modeling with Kepler's list of planetary candidates (KOIs). This process includes modeling of planetary transits, phase curves and providing orbital solutions. We will use Kepler-photometry and other groundbased observables to determine key planetary parameters such as the radius and mass. More importantly, we will also determine posterior probability distributions for the fitted parameters by employing a Markov chain Monte Carlo algorithm. Our algorithms have already been tested and published for various Kepler planet- discoveries. We wish move from modeling a single system at a time, to batch processing the entire Kepler planetary candidate catalog. Our plan to then generate a catalog of model parameters for each KOIs and make this catalog readily available to the Kepler Science Team and collaborators on an annual basis. We also wish to make the catalog public on a timescale that coincides with Kepler public data releases. By calculating detailed models of the Kepler targets, our work will help support the Kepler Mission to achieve many of its primary scientific goals. We expect to measure with uncertainties: orbital periods, planet radii, inclinations, reflection/emission from the planet, the amplitude of planet-star graviational interactions and transit timing variations. When sufficient groundbased radial velocities are available we will also model orbital solutions and planetary densities. We will also model multi-planet systems where multiple planets are seen transiting a star.
CHARACTERIZING PLANETARY SYSTEMS: CHEMICAL ABUNDANCES OF KEPLER STARS FROM KECK/HIRES FOLLOW-UP SPECTROSCOPY
Association for Universities for Research in Astronomy
A comprehensive, homogeneous chemical abundance analysis of planetary host stars discovered by Kepler is proposed. Precise stellar parameters (effective temperature and surface gravities) and abundances of up to 20 elements (Li, C, N, O, Na, Mg, Al, Si, S, K, Ca, Sc, Ti, V, Cr, Mn, Fe, Co, Ni, and Zn) will be derived for the highest priority Kepler Object of Interest (KOI) host stars using high-resolution echelle spectroscopy being obtained with the High Resolution Echelle Spectrometer (HIRES) and 10-m Keck I telescope as part of the Kepler Follow-up Observing Program (FOP). There are four main goals of this project: 1) Derive precise stellar parameters that will contribute to the characterization of KOI host stars; 2) Identify correlations between KOI host star metallicities and planetary system characteristics (mass, semi-major axis, orbital period, eccentricity, etc.); 3) Chemically characterize KOI host stars by deriving abundances of numerous elements; and 4) Search for chemical signatures of planet formation processes. The proposed abundance analysis will contribute directly to Kepler project efforts to deduce planet and stellar host properties, and the Kepler primary scientific objective to identify correlations between the presence and characteristics of planetary systems with stellar properties of the host stars. Precise stellar parameters will constrain stellar luminosity classes (dwarf, subgiant, giant) and stellar radii, the latter of which are needed to determine planet radii. Metallicities will be derived with uncertainties of approximately 0.05 dex, allowing for meaningful comparisons to planetary system characteristics. Deriving the abundances of numerous elements will allow a complete determination of host star metallicities, as opposed to assuming a scaled-solar metallicity, and if the abundances of individual elements, such as Li, are correlated with planetary system characteristics. Patterns in element-to-element relative abundances can be investigated to search for vestiges of planet formation processes.
Early results from Kepler suggest that Neptune-size and smaller planets are much more common in the Galaxy than the Jupiter-type giant planets that currently dominate the sample of known exoplanets. The proposed abundance analysis will establish the chemical signatures of stars with small planets, including those with Earth-size planets, and place critical constraints on planet formation models. This project is highly complementary to Kepler Science Team activities and will enhance the scientific value of the Kepler Mission.
DETECTING AND CHARACTERIZING PLANETARY SYSTEMS WITH TRANSIT TIMING VARIATIONS
Each transiting planet discovered by Kepler provides an opportunity to discover additional nontransiting planets in the same system, determine their masses and orbital elements, and characterize the interactions among the various objects. This is done by observing and analyzing the deviations from a constant period that arise from mutual gravitational perturbations (transit timing variations or TTVs). This method was first developed by myself, my advisor, and my present collaborator (Agol et al. 2005, Holman & Murray 2005) and I have contributed to many of the first analyses and planet discoveries using TTVs (e.g., Steffen & Agol 2006, Agol & Steffen 2007, Holman et al. 2010). TTVs are also important probes of planet formation and evolution theories as some models predict an abundance of systems in or near mean motion resonances where a TTV signal would be the largest. I propose to continue my work for the Kepler mission, where I currently chair the TTV working group, in analyzing the transit times of Kepler systems to detect and characterize the planets discovered in those systems. In particular, using my existing software I will identify additional, nontransiting planets in Kepler systems including possibly nontransiting planets in the habitable zone of the host star, measure the orbital elements of planets in multi and singly transiting systems, eliminate potential false positive signals based upon the observation or non-observation of TTVs, and constrain planet formation and evolution theories using the observed orbital architectures of planetary systems (e.g., proximity to resonance or large eccentricities where the TTV signal is particularly strong).
TESTING KEPLER ASTROMETRY WITH OUR NEAREST NEIGHBORS
Georgia State University
While the primary science goal for the Kepler mission is the detection and characterization of terrestrial and giant exoplanets through ultra-precision photometry, the telescope is capable of collecting milli-arcsecond precision astrometric data for each of the target stars. This single measurement precision when combined with the large number of observations collected by the mission each quarter (~1600) and over its 3.5 year lifetime (>22k), means Kepler should be sensitive to Jupiter-mass planets and brown dwarfs around some of the nearest stars in the input catalog. Astrometry can also be used to derive the trigonometric parallax and proper motions of Kepler stars resulting in a direct measurement of the radius of the star. Finally, astrometric measurements can rule out false positives due to blends with eclipsing binaries. With this in mind, I propose to assist the Kepler team in extracting precise astrometric information from the Kepler data sets with the final science goal of measuring proper motions, parallaxes and companion perturbations. I will use my expertise with ground and space-based precision data analysis to determine robust astrometric solutions for the Kepler data. The statistics on the frequency of stars with massive planets at separations of > 1 AU, when eventually combined with results from the WFIRST micro-lensing mission, will aid future NASA missions including JWST in determining how to efficiently directly image planets around nearby stars. Both astrometry and the detection of habitable planets were emphasized as priority goals in the 2010 Decadal Survey.
VALIDATION OF KEPLER CANDIDATE TRANSITING PLANETS WITH BLENDER
Smithsonian Astrophysical Observatory
In its quest to determine the frequency of Earth-size planets in the habitable zone of Sun- like stars, the Kepler Mission has identified hundreds of stars displaying small and periodic drops in brightness that appear to be due to a transiting planet, some possibly as small as the Earth. However, experience has shown that a significant fraction of these are due to phenomena other than a planet, such as a faint background eclipsing binary along the same line of sight (a "blend") that is contaminating the photometry. The Kepler Team is now faced with the difficult task of sorting through the candidates to rule out these false positives, in order to confirm a transiting planet. This is an essential step for establishing the frequency of such objects. Some of the most troublesome cases are those with shallow signals that might correspond to Earth-size planets in the habitable zone of their parent stars. These are precisely among the most valuable and exciting candidates the Kepler Mission has found, yet confirmation by the usual spectroscopic methods is often beyond reach. This is because the reflex radial-velocity motion they impose on the star is typically too small to detect with current instrumentation, or because the star is too faint, chromospherically active, or rotating too rapidly. Other follow-up observations such as near-infrared spectroscopy, Spitzer observations, or high-resolution imaging, along with an analysis of the motion of the image centroids, can help to rule out many of the false positive scenarios, but not all. The most effective way of exploring the remaining scenarios that are the most difficult to exclude is to model the Kepler light curves directly in terms of a blend. Once these cases are identified, validation of a candidate must come through a careful statistical assessment of the likelihood of such blends compared to the likelihood of a true planet. This is a proposal to perform this computationally intensive modeling for the most interesting candidates that the Kepler Mission is not able to validate in any other way. We propose to make use of BLENDER, a sophisticated light-curve analysis technique developed by the PI specifically for this purpose. BLENDER is able to systematically explore the vast space of parameters for blends and pose tight constraints on the configurations that can mimic the photometric signal. These constraints are complementary to those placed by other observations, and drastically reduce the overall number of blends that remain possible. The statistical estimate of the a priori likelihood of those blends, which is also part of this proposal, will incorporate information on the mean number density of stars around the target, as well as the expected frequencies of background (or foreground) eclipsing binaries, hierarchical triples, and other configurations. The end result of the analysis is a confidence level for the statement that the candidate is a planet rather than a false positive. BLENDER is a powerful new tool in the field of exoplanet discovery that has already had a significant impact on a number ground-based surveys, and has been applied successfully by the PI in a pilot study to several of the most interesting Kepler candidates. The ability to perform these critical studies will enable the Kepler Team to validate small planets that cannot be confirmed by the usual means, and thus clearly supports one of the main science objectives of the Mission.
DETAILED MODELING OF EXOPLANET TRANSITS
San Diego State University
The P.I. proposes to continue to use state-of-the-art transit modeling software to measure the properties of Kepler discovered exoplanets. The P.I. will provide the Kepler Team with the following products:
(1) accurate system parameters for the high signal-to-noise planets and host stars (e.g.~planet masses and radii);
(2) precise transit timings to be used by the Kepler Science Team to confirm planets, measure their masses, and detect other non-transiting bodies; and
(3) characterization of stellar activity and its impact on exoplanet science (e.g.~how starspots affect transit timing).
The PI has an outstanding record as a Participating Scientist, having generated products that have been used in several key discovery papers (most notably Kepler-9 and Kepler- 11). The P.I. has also lead two significant discovery papers (ellipsoidal variation in HAT- P-7, and the remarkable system KOI-54). The proposed work is a refined version of the PI's original PSP activities, but refined and tailored to the specific needs of the post- launch era.
The P.I. will analyze a subset of the Kepler observations at a level that is impossible to do for a large set of systems. These short-period systems will be scrutinized with painstaking attention to detail. For the larger planets with very high signal-to-noise ratios, the investigation will emphasize accuracy not just precision, and will focus on measuring subtle but important astrophysical effects. For the smaller planets, the focus will be more on robust determination of the planetary characteristics. This narrow focus adds a strong complement to the broader Kepler Mission objectives.
The eclipse modeling software ``ELC'' will be the primary tool employed by the P.I. and will be used to measure planetary radii, masses, temperatures, reflections/albedo, orbital elements, transit timing variations, and other characteristics of the planet-star system. As part of this project, ELC will be improved to take full advantage of the unprecedented level of precision of the Kepler light curves, and the high occurrence of multi-transiting systems in the Kepler data. Thus a fourth product will be generated:
(4) the enhanced ELC code will be made available to the Kepler Science Team.
The PI of this proposal is well--versed in skills relevant to this research, as evidenced by the Kepler papers he has led and contributed to, and the many Working Groups he actively participates in. The proposed investigation is directly related to the objectives of the Kepler Mission, and to NASA interests as a whole, as the work will help the Kepler Team achieve the principle goals of the Mission.
OBLIQUITIES OF KEPLER STARS: CLUES TO PLANET MIGRATION
Massachusetts Institute of Technology
In the Solar system, the Sun's spin axis and the planets' orbital axes are nearly aligned. Although exoplanetary systems were also expected to be well-aligned, recent observations have uncovered systems with strongly misaligned orbits, and even retrograde orbits. These discoveries have stimulated progress on the most enduring problem of exoplanetary science: the existence of close-in planets. The results have been marshaled as evidence against the standard scenario of disk migration, and in favor of orbit-shrinking mechanisms involving few-body dynamics and tidal dissipation. However, all the previous work pertains to "hot Jupiters," leaving many questions unanswered. Do smaller planets also show a diversity of orbital orientations? What about planets with longer periods, which do not experience strong tides? And, in systems with multiple planets, are the orbits aligned with one another and with the central star? Kepler provides the first opportunity to answer these questions and study spin-orbit alignment for small planets, long-period planets, and multiple-planet systems. The PI proposes to study the obliquities of Kepler planet-hosting stars, using three complementary techniques:
(1) Observations of the Rossiter-McLaughlin effect, a spectroscopic anomaly that reveals the sky-projected stellar obliquity for individual systems. The methodology is also useful for planet validation: making sure a transit-like signal is caused by a planet and not by a conspiracy of eclipsing stars.
(2) Rotation rate statistics, which permit an assessment of spin-orbit alignment in an ensemble of stars with transiting planets. The distribution of projected stellar rotation rates is compared to that of a control sample of stars without transits.
(3) Starspot anomalies, a new method by which a stars obliquity is revealed by timing the occultations of starspots by a transiting planet.
Questions concerning Kepler's science opportunities and open programs, public archive or community tools? Contact us via the